Top Banner
Meteoritics & Planetary Science 40, Nr 6, 817–840 (2005) Abstract available online at http://meteoritics.org 817 © The Meteoritical Society, 2005. Printed in USA. Earth Impact Effects Program: A Web-based computer program for calculating the regional environmental consequences of a meteoroid impact on Earth Gareth S. COLLINS, 1* H. Jay MELOSH, 2 and Robert A. MARCUS 2 1 Impacts and Astromaterials Research Centre, Department of Earth Science and Engineering, Imperial College London, South Kensington Campus, London, SW7 2AZ, UK 2 Lunar and Planetary Laboratory, University of Arizona, 1629 East University Boulevard, Tucson, Arizona 85721–0092, USA * Corresponding author. E-mail: [email protected] (Received 29 July 2004; revision accepted 14 April 2005) Abstract–We have developed a Web-based program for quickly estimating the regional environmental consequences of a comet or asteroid impact on Earth (www.lpl.arizona.edu/ impacteffects). This paper details the observations, assumptions and equations upon which the program is based. It describes our approach to quantifying the principal impact processes that might affect the people, buildings, and landscape in the vicinity of an impact event and discusses the uncertainty in our predictions. The program requires six inputs: impactor diameter, impactor density, impact velocity before atmospheric entry, impact angle, the distance from the impact at which the environmental effects are to be calculated, and the target type (sedimentary rock, crystalline rock, or a water layer above rock). The program includes novel algorithms for estimating the fate of the impactor during atmospheric traverse, the thermal radiation emitted by the impact-generated vapor plume (fireball), and the intensity of seismic shaking. The program also approximates various dimensions of the impact crater and ejecta deposit, as well as estimating the severity of the air blast in both crater-forming and airburst impacts. We illustrate the utility of our program by examining the predicted environmental consequences across the United States of hypothetical impact scenarios occurring in Los Angeles. We find that the most wide-reaching environmental consequence is seismic shaking: both ejecta deposit thickness and air-blast pressure decay much more rapidly with distance than with seismic ground motion. Close to the impact site the most devastating effect is from thermal radiation; however, the curvature of the Earth implies that distant localities are shielded from direct thermal radiation because the fireball is below the horizon. INTRODUCTION Asteroid and comet impacts have played a major role in the geological and biological history of the Earth. It is widely accepted that one such event, 65 million years ago, perturbed the global environment so catastrophically that a major biological extinction ensued (Alvarez 1980). As a result, both the scientific community and the general populace are increasingly interested in both the threat to civilization and the potential environmental consequences of impacts. Previous papers have examined, in detail, the natural hazard associated with the major environmental perturbations caused by impact events (Toon et al. 1994, 1997). To provide a quick and straightforward method for estimating the severity of several of these environmental effects, we have developed a free-of-charge, easy-to-use Web page maintained by the University of Arizona, which is located at: www.lpl.arizona.edu/impacteffects. Our program focuses on the consequences of an impact event for the regional environment; that is, from the impact location to a few thousand km away. The purpose of this paper is to present and justify the algorithm behind our program so that it may be applied more specifically to important terrestrial impact events and its reliability and limitations may be understood. Before describing our program in detail, we will briefly review the impact process and the related environmental consequences. The impact of an extraterrestrial object on Earth begins when the impactor enters the tenuous upper atmosphere. At this moment, the impactor is traveling at a speed of between 11 and 72 km s 1 on a trajectory anywhere between normal incidence (90° to the Earth’s surface) and a grazing impact, parallel to the Earth’s surface. The most likely impact angle is 45° (Shoemaker 1962). The impactor’s
24
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Documentation

Meteoritics amp Planetary Science 40 Nr 6 817ndash840 (2005)Abstract available online at httpmeteoriticsorg

817 copy The Meteoritical Society 2005 Printed in USA

Earth Impact Effects Program A Web-based computer program for calculating the regional environmental consequences of a meteoroid impact on Earth

Gareth S COLLINS1 H Jay MELOSH2 and Robert A MARCUS2

1Impacts and Astromaterials Research Centre Department of Earth Science and Engineering Imperial College LondonSouth Kensington Campus London SW7 2AZ UK

2Lunar and Planetary Laboratory University of Arizona 1629 East University Boulevard Tucson Arizona 85721ndash0092 USACorresponding author E-mail gcollinsimperialacuk

(Received 29 July 2004 revision accepted 14 April 2005)

AbstractndashWe have developed a Web-based program for quickly estimating the regionalenvironmental consequences of a comet or asteroid impact on Earth (wwwlplarizonaeduimpacteffects) This paper details the observations assumptions and equations upon which theprogram is based It describes our approach to quantifying the principal impact processes that mightaffect the people buildings and landscape in the vicinity of an impact event and discusses theuncertainty in our predictions The program requires six inputs impactor diameter impactor densityimpact velocity before atmospheric entry impact angle the distance from the impact at which theenvironmental effects are to be calculated and the target type (sedimentary rock crystalline rock ora water layer above rock) The program includes novel algorithms for estimating the fate of theimpactor during atmospheric traverse the thermal radiation emitted by the impact-generated vaporplume (fireball) and the intensity of seismic shaking The program also approximates variousdimensions of the impact crater and ejecta deposit as well as estimating the severity of the air blastin both crater-forming and airburst impacts We illustrate the utility of our program by examining thepredicted environmental consequences across the United States of hypothetical impact scenariosoccurring in Los Angeles We find that the most wide-reaching environmental consequence is seismicshaking both ejecta deposit thickness and air-blast pressure decay much more rapidly with distancethan with seismic ground motion Close to the impact site the most devastating effect is from thermalradiation however the curvature of the Earth implies that distant localities are shielded from directthermal radiation because the fireball is below the horizon

INTRODUCTION

Asteroid and comet impacts have played a major role inthe geological and biological history of the Earth It iswidely accepted that one such event 65 million years agoperturbed the global environment so catastrophically that amajor biological extinction ensued (Alvarez 1980) As aresult both the scientific community and the generalpopulace are increasingly interested in both the threat tocivilization and the potential environmental consequences ofimpacts Previous papers have examined in detail thenatural hazard associated with the major environmentalperturbations caused by impact events (Toon et al 19941997) To provide a quick and straightforward method forestimating the severity of several of these environmentaleffects we have developed a free-of-charge easy-to-useWeb page maintained by the University of Arizona which is

located at wwwlplarizonaeduimpacteffects Our programfocuses on the consequences of an impact event for theregional environment that is from the impact location to afew thousand km away The purpose of this paper is topresent and justify the algorithm behind our program so thatit may be applied more specifically to important terrestrialimpact events and its reliability and limitations may beunderstood

Before describing our program in detail we will brieflyreview the impact process and the related environmentalconsequences The impact of an extraterrestrial object onEarth begins when the impactor enters the tenuous upperatmosphere At this moment the impactor is traveling at aspeed of between 11 and 72 km sminus1 on a trajectory anywherebetween normal incidence (90deg to the Earthrsquos surface) and agrazing impact parallel to the Earthrsquos surface The most likelyimpact angle is 45deg (Shoemaker 1962) The impactorrsquos

818 G S Collins et al

traverse of the atmosphere may disrupt and decelerate theimpactor significantlymdasha process that greatly affects theenvironmental consequences of the collision Small impactorsare disrupted entirely during their atmospheric traversedepositing their kinetic energy well above the surface andforming no crater Larger objects however retain sufficientmomentum through the atmosphere to strike the Earth withenough energy to excavate a large crater and initiate severalprocesses that affect the local regional and even globalenvironment

The formation of an impact crater is an extremelycomplicated and dynamic process (Melosh 1989) The abruptdeceleration of a comet or asteroid as it collides with the Earthtransfers an immense amount of kinetic energy from theimpacting body to the target As a result the target andimpactor are rapidly compressed to very high pressures andheated to enormous temperatures Between the compressedand uncompressed material a shock wave is created thatpropagates away from the point of impact In the wake of theexpanding shock wave the target is comprehensivelyfractured shock-heated shaken and set in motionmdashleadingto the excavation of a cavity many times larger than theimpactor itself This temporary cavity (often termed thetransient crater Dence et al 1977) subsequently collapsesunder the influence of gravity to produce the final crater formAs the crater grows and collapses large volumes of rockdebris are ejected onto the surface of the Earth surroundingthe crater Close to the crater rim this ldquoejecta depositrdquo formsa continuous blanket smothering the underlying terrainfurther out the ejecta lands as a scattered assortment of fine-grained dust and larger bombs that may themselves formsmall secondary craters

In addition to cratering the surface of the earth animpact event initiates several other processes that may havesevere environmental consequences During an impact thekinetic energy of the impactor is ultimately converted intothermal energy (in the impactor and target) seismic energyand kinetic energy of the target and atmosphere The increasein thermal energy melts and vaporizes the entire impactor andsome of the target rocks The hot plume of impact-generatedvapor that expands away from the impact site (referred to asthe ldquofireballrdquo) radiates thermal energy that may ignite firesand scorch wildlife within sight of the fireball As the impact-generated shock wave propagates through the target iteventually decays into elastic waves that travel greatdistances and cause violent ground shaking several craterradii away In addition the atmosphere is disturbed in asimilar manner to the target rocks a shock wave propagatesaway from the impact site compressing the air to highpressures that can pulverize animals and demolish buildingsvehicles and infrastructure particularly where constructionalquality is poor Immediately behind the high-pressure frontviolent winds ensue that may flatten forests and scatterdebris

All of these impact-related processes combine and interactin an extremely complicated way that requires detailedobservation laboratory experiments or computer models tofully simulate and understand However with certainsimplifying assumptions we can derive reasonable estimatesof their consequences for the terrestrial environment In thefollowing sections we describe each of the steps that allow usto achieve this in the Earth Impact Effects Program We discusshow our program estimates 1) the impact energy and averagetime interval between impacts of the same energy somewhereon Earth 2) the consequences of atmospheric entry 3) forcrater forming events the resulting crater size and volume ofthe melt produced 4) the thermal radiation damage from thefireball 5) the impact-induced seismic shaking 6) the extentand nature of the ejecta deposit and 7) the damage caused bythe blast wave To clearly identify our algorithm in thefollowing discussion all of the equations that we implement inthe code are labeled with an asterisk ()

To make the program accessible to the broadest range ofusers it was written with as few input parameters as possibleThe program requests six descriptors which are illustratedschematically in Fig 1 the diameter of the impactor L0 (we usethe term impactor to denote the asteroid comet or otherextraterrestrial object considered) the impactor density ρi theimpact velocity v0 the angle that the trajectory of the impactorsubtends with the surface of the Earth at the impact point θ thetarget type and the distance away from the impact at which theuser wishes to calculate the environmental consequences rThree target types are possible sedimentary rock for which weassign a target density of ρt = 2500 kg mminus3 crystalline rock (ρt= 2750 kg mminus3) or a marine target for which the programrequests a water-layer depth dw and assigns a density of ρw =1000 kg mminus3 for the water and a target density of ρt = 2700 kgmminus3 for the rock layer below The program offers the user avariety of options for units however in this paper the units forall variables are the SI units (mks) unless otherwise stated

IMPACT ENERGY AND RECURRENCE INTERVAL

The most fundamental quantity in assessing theenvironmental consequences of the impact is the energyreleased during the impact which is related to the kineticenergy of the impactor E before atmospheric entry begins Atnormal solar system impact speeds E is approximately givenas one half times the impactor mass mi times the square of theimpactor velocity v0 which can be rewritten in terms of themeteoroidrsquos density ρi and diameter L0 assuming that themeteoroid is approximately spherical

(1)

In fact the program uses the relativistic energy equationto accommodate the requests of several science fictionwriters The program does not limit the impact velocity to

E 12---miv0

2 π12------ρiL0

3v02= =

Earth Impact Effects Program 819

72 km sminus1 the maximum possible for an impactor bound tothe Sun however we have limited the maximum velocity tothe speed of light in response to attempts of a few users toinsert supra-light velocities

Natural objects that encounter the Earth are eitherasteroids or comets Asteroids are made of rock (ρi ~2000ndash3000 kg m3 Hilton 2002) or iron (ρi ~8000 kg m3) andtypically collide with the Earthrsquos atmosphere at velocities of12ndash20 km sminus1 (Bottke et al 1994) Detailed knowledge of thecomposition of comets is currently lacking however they areof much lower density (ρi ~500ndash1500 kg m3) and are composedmainly of ice (Chapman and Brandt 2004) Typical velocitiesat which comets might encounter the Earthrsquos atmosphere are inthe range of 30ndash70 km sminus1 (Marsden and Steel 1994) Thus anasteroid or comet typically has 4ndash20 times the energy per unitmass of TNT at the moment atmospheric entry beginsTherefore impact events have much in common with chemicaland nuclear explosions a fact that we will rely on later in ourestimates of the environmental effects of an impact

Observations of near-Earth objects made by severaltelescopic search programs show that the number of near-Earth asteroids with a diameter greater than Lkm (in km) maybe expressed approximately by the power law (Near-EarthObject Science Definition Team 2003)

N(gtL) asymp 1148Lkmminus2354 (2)

These data may also be represented in terms of therecurrence interval TRE in years versus the impact energy EMtin megatons of TNT by assuming a probability of a single-object collision with Earth (~16 times 10minus9 yrminus1 Near-Earth Object

Science Definition Team 2003 their Fig 23) and multiplyingby the number of asteroids of a given potential impact energythat are estimated to be circling the sun with potentiallyhazardous Earth-crossing orbits We found that a simplepower-law relationship adequately represents these data

TRE asymp 109EMt078 (3)

Thus for a given set of user-input impact parameters (L0v0 ρi ρt and θ) the program computes the kinetic energy(EMt in megatons 1 Mt = 418 times 1015 J) possessed by theimpacting body when it hits the upper atmosphere and definesan average time interval between impacts of that energysomewhere on the Earth Furthermore we estimate therecurrence interval TRL for impacts of this same energy withina certain specified distance r of the impact This is simply theproduct of the recurrence interval for the whole Earth and thefraction of the Earthrsquos surface area that is within the distance r

(4)

where ∆ is the epicentral angle from the impact point to arange r (given in radians by ∆ = rRE where RE is the radiusof the Earth Fig 1)

Currently the relative importance of comets to the Earth-crossing impactor flux is not well-constrained The Near-EarthObject Science Definition Team (2003) suggests that cometscomprise only about 1 of the estimated population of smallNEOs however there is evidence to suggest that at largersizes comets may comprise a significantly larger proportion ofthe impactor flux (Shoemaker et al 1990) Of the asteroids thatcollide with the Earthrsquos atmosphere the current best estimateis that approximately 2ndash10 are iron asteroids (Bland andArtemieva 2003) based on NEO and main-belt asteroidspectroscopy (Bus et al 2002 Binzel et al 2003) meteoritecomposition and the impactor types in large terrestrial craters

ATMOSPHERIC ENTRY

Atmospheric entry of asteroids has been discussed indetail by many authors (Chyba et al 1993 Ivanov et al 1997Krinov 1966 Melosh 1981 Passey and Melosh 1980 Svetsovet al 1995 Korycansky et al 2000 2002 Korycansky andZahnle 2003 2004 Bland and Artemieva 2003) and is nowunderstood to be a complex process involving interaction ofthe atmosphere and fragmenting impactor in the Earthrsquosgravitational field For the purposes of a simple program of thetype that we have created many of the refinements nowunderstood are too complex to be included Therefore wehave opted to make a number of drastic simplifications thatwe believe will still give a good description of the basicevents during atmospheric entry for most cases Of course forrefined predictions a full simulation using all of the knownprocesses and properties must be undertaken Atmosphericentry has no significant influence on the shape energy or

Fig 1 Diagram illustrating the input parameters for the Earth ImpactEffects Program L0 is the impactor diameter at the top of theatmosphere v0 is the velocity of the impactor at the top of theatmosphere ρi is the impactor density ρt is the target density and θis the angle subtended between the impactorrsquos trajectory and thetangent plane to the surface of the Earth at the impact point Thedistance r from the impact site at which the environmentalconsequences are determined is measured along the surface of theEarth the epicentral angle ∆ between the impact point and thisdistance r is given by ∆ = rRE where RE is the radius of the Earth

TRLTRE

2---------- 1 ∆cosndash( )=

820 G S Collins et al

momentum of impactors with a mass that is much larger thanthe mass of the atmosphere displaced during penetration Forthis reason the program procedure described below is appliedonly for impactors less than 1 km in diameter

For the purposes of the Earth Impact Effects Program weassume that the trajectory of the impactor is a straight linefrom the top of the atmosphere to the surface sloping at aconstant angle to the horizon given by the user Accelerationof the impactor by the Earthrsquos gravity is ignored as isdeviation of the trajectory toward the vertical in the case thatterminal velocity is reached as it may be for small impactorsThe curvature of the Earth is also ignored The atmosphere isassumed to be purely exponential with the density given by

ρ(z) = ρ0eminuszH (5)

where z is the altitude above the surface H is the scale heighttaken to be 8 km on the average Earth and ρ0 is the surfaceatmospheric density taken to be equal to 1 kgm3

During the first portion of the impactorrsquos flight its speedis decreased by atmospheric drag but the stresses are toosmall to cause fragmentation Small meteoroids are oftenablated to nothing during this phase but in the currentprogram implementation we ignore ablation on the groundsthat it seldom affects the larger impactors that reach thesurface to cause craters Thus this program should not beused to estimate the entry process of small objects that maycause visible meteors or even drop small meteorites to thesurface at terminal velocity

While the body remains intact the diameter of theincoming impactor is constant equal to the diameter L0 givenby the user The rate of change of the velocity v is given by theusual drag equation (corrected from Melosh 1989 chapter 11)

(6)

where CD is the drag coefficient taken to equal 2 and ρi is theimpactor density (an input parameter) This equation can begreatly simplified by making the replacement dt = minusdzv sinθ(justified by our assumption that the impactor travels in astraight line) and rearranging

(7)

Integration of this equation using the exponential densitydependence gives the velocity of the impactor as a function ofaltitude

(8)

where θ is the entry angle and v0 is the impact velocity at thetop of the atmosphere given by the user

As the impactor penetrates the atmosphere theatmospheric density increases and the stagnation pressure at

the leading edge of the impactor Ps = ρ(z) v(z)2 risesEventually this exceeds the strength of the impactor and itbegins to break up Observed meteoroids often undergoseveral cascades of breakup reflecting components of widelyvarying strengths The entire subject of meteoroid strength ispoorly understood as measured crushing strengths ofspecimens collected on the ground are often a factor of 10 lessthan strengths inferred from observed breakup (Svetsov et al1995) Clearly strong selection effects are at work For thepurposes of our program we decided not to embroil the userin the ill-defined guesswork of estimating meteoroid crushingstrength Instead we found a rough correlation betweendensity and estimated strength for comets (about 15 Pa intension from the tidal breakup of SL-9 Scotti and Melosh1993) chondrites (Chyba et al 1993) and iron or stoneobjects (Petrovic 2001) Based on four simplified estimatesfor comets carbonaceous stony and iron meteorites weestablished an empirical strength-density relation for use inthe program The yield strength Yi of the impactor in Pa is thuscomputed from

(9)

where the impactor density ρi is in kg mminus3 Note that even atzero density this implies a non-zero strength of about 130 PaThus this empirical formula should not be applied too far outof the range of 1000 to 8000 kg mminus3 over which it wasestablished

Using this estimate of strength and comparing it to thestagnation pressure we can compute an altitude of breakup zby solving the transcendental equation

Yi = ρ(z)v2(z) (10)

Rather than solving this equation in the program directlyan excellent analytic approximation to the solution was foundand implemented

(11)

where If is given by

(12)

In certain specific instances (ie small strongimpactors) the impactor may reach the surface intact in thiscase If gt1 and Equation 11 does not apply The properlydecremented velocity calculated using Equation 8 is used tocompute a crater size (If this velocity happens to be less thanthe terminal velocity then the maximum of the two is usedinstead) The velocity at the top of the atmosphere and at thesurface is reported

Most often the impactor begins to break up well abovethe surface in this case If lt1 and Equation 11 is used to

dvdt------

3ρzCD4ρiL0-----------------ndash v2=

d ln vdz-------------

3ρ z( )CD4ρiL0 θsin--------------------------=

v z( ) v03ρ z( )CDH4ρiL0 θsin---------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

log10Yi 2107 00624+ ρi=

z HndashYi

ρ0vi2-----------

⎝ ⎠⎜ ⎟⎛ ⎞

ln 1308 0314If 1303 1 Ifndashndashndash+asymp

If 407CDHYi

ρiL0vi2 θsin

----------------------------=

Earth Impact Effects Program 821

compute the breakup altitude z After breakup the fragmentsbegin to disperse in a complex series of processes (Passey andMelosh 1980 Svetsov et al 1995) that require detailednumerical treatment However a simple approximation to thiscascade was found (Chyba et al 1993 Melosh 1981) calledthe pancake model that does a good job for Tunguska-classevents The basic idea of this model is that the impactor oncefractured expands laterally under the differential pressurebetween the front and back surfaces The front of the impactoris compressed at the stagnation pressure and the rear isessentially in a vacuum with zero pressure The sides squirtout at a rate determined by force balance in an inviscid fluidThis leads to a simple equation for the expansion of theimpactor diameter L now a function of time

(13)

The initial condition is that L = L0 at z = z If L does notincrease too much over the scale height H the timederivatives can be replaced with altitude derivatives (Chybaet al 1993) and a nonlinear differential equation can beconstructed that does not contain v(z)

(14)

Again we construct an analytic approximation to the fullsolution of this equation which is adequate for the purposesof the program

(15)

where the dispersion length scale l is given by

(16)

The velocity as a function of altitude is then given byinserting this expression for L(z) into the drag equation andintegrating downward from the breakup altitude z Becauseof the rapid expansion of the pancake the drag rises rapidly aswell and the velocity drops as a double exponential

(17)

The crushed impactor spreads laterally until the ratioL(z)L0 reaches a prescribed limit which we call the ldquopancakefactorrdquo fp In reality this should be no larger than 2 to 4(Ivanov et al 1997) after which the fragments are sufficientlyseparated that they follow independent flight paths and may

suffer one or more further pancake fragmentation eventsHowever Chyba et al (1993) obtained good agreement withTunguska-class events using pancake factors as large as 5ndash10In this work we experimented with different factors andsettled on a value of 7 to terminate the dispersion of theimpactor The altitude at which this dispersion is obtained iscalled the ldquoairburst altituderdquo (zb see Fig 2a) it is given bysubstituting fp = L(z)L0 into Equation 15 and rearranging

(18)

If the airburst occurs above the surface (Fig 2a) most ofthe energy is dissipated in the air We report the airburstaltitude zb and the residual velocity of the swarm which iscomputed using Equation 17 In this case the integral in theexponent evaluated from the airburst altitude to thedisruption altitude is given by

(19)

with the definition The surface impact velocityof the remnants from the airburst vi is also reported as themaximum of the terminal velocity of a fragment half thediameter of the original impactor or the velocity of theswarm as a whole The spreading velocity at airburstmultiplied by the time to impact is added to the breadth ofthe swarm to estimate the dispersion of what will be a strewnfield on the surface The principal environmentalconsequence of such an event is a strong blast wave in theatmosphere (see below)

On the other hand if the pancake does not spread to thelimiting size before it reaches the ground (zb le0 inEquation 19 Fig 2b) the swarm velocity at the moment ofimpact is computed using Equation 17 In this case theintegral in the exponent evaluated from the surface (z = 0) tothe disruption altitude is given by

(20)

The dispersion of the swarm at impact is compared to theestimated transient crater size (see below) and if it iscomparable or larger then the formation of a crater field isreported similar to that actually observed at HenburyAustralia Otherwise we assume the impact to be a crater-

d2Ldt2---------

CDPsρiL

-------------CDρ z( )v2 z( )

ρiL--------------------------------= =

Ld2Ldz2---------

CDρ z( )

ρisin2θ-------------------=

L z( ) L0 1 2Hl

-------⎝ ⎠⎛ ⎞ 2 z zndash

2H-------------

⎩ ⎭⎨ ⎬⎧ ⎫

exp 1ndash⎝ ⎠⎜ ⎟⎛ ⎞

2

+=

l L0 θρi

CDρ z( )---------------------sin=

v z( ) v z( ) 34---ndash

CDρ z( )

ρiL03 θsin

---------------------- ez zndash( ) Hfrasl

z

z

int L2 z( )dz

⎩ ⎭⎪ ⎪⎨ ⎬⎪ ⎪⎧ ⎫

exp=

zb z 2H 1 l2H------- fp

2 1ndash+lnndash=

ez zndash( ) Hfrasl

zburst

z

int L2 z( )dz

lL0

2

24--------α 8 3 α2+( ) 3α l

H---- 2 α2+( )+

=

α fp2 1ndashequiv

ez zndash( ) Hfrasl

0

z

int L2 z( )dz H3L0

2

3l2------------- 34 lH----⎝ ⎠

⎛ ⎞ 2+ e

z Hfrasl

6e2z Hfrasl

16e3z 2Hfrasl

3ndashndash

+

lH----⎝ ⎠

⎛ ⎞ 22ndash

=

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 2: Documentation

818 G S Collins et al

traverse of the atmosphere may disrupt and decelerate theimpactor significantlymdasha process that greatly affects theenvironmental consequences of the collision Small impactorsare disrupted entirely during their atmospheric traversedepositing their kinetic energy well above the surface andforming no crater Larger objects however retain sufficientmomentum through the atmosphere to strike the Earth withenough energy to excavate a large crater and initiate severalprocesses that affect the local regional and even globalenvironment

The formation of an impact crater is an extremelycomplicated and dynamic process (Melosh 1989) The abruptdeceleration of a comet or asteroid as it collides with the Earthtransfers an immense amount of kinetic energy from theimpacting body to the target As a result the target andimpactor are rapidly compressed to very high pressures andheated to enormous temperatures Between the compressedand uncompressed material a shock wave is created thatpropagates away from the point of impact In the wake of theexpanding shock wave the target is comprehensivelyfractured shock-heated shaken and set in motionmdashleadingto the excavation of a cavity many times larger than theimpactor itself This temporary cavity (often termed thetransient crater Dence et al 1977) subsequently collapsesunder the influence of gravity to produce the final crater formAs the crater grows and collapses large volumes of rockdebris are ejected onto the surface of the Earth surroundingthe crater Close to the crater rim this ldquoejecta depositrdquo formsa continuous blanket smothering the underlying terrainfurther out the ejecta lands as a scattered assortment of fine-grained dust and larger bombs that may themselves formsmall secondary craters

In addition to cratering the surface of the earth animpact event initiates several other processes that may havesevere environmental consequences During an impact thekinetic energy of the impactor is ultimately converted intothermal energy (in the impactor and target) seismic energyand kinetic energy of the target and atmosphere The increasein thermal energy melts and vaporizes the entire impactor andsome of the target rocks The hot plume of impact-generatedvapor that expands away from the impact site (referred to asthe ldquofireballrdquo) radiates thermal energy that may ignite firesand scorch wildlife within sight of the fireball As the impact-generated shock wave propagates through the target iteventually decays into elastic waves that travel greatdistances and cause violent ground shaking several craterradii away In addition the atmosphere is disturbed in asimilar manner to the target rocks a shock wave propagatesaway from the impact site compressing the air to highpressures that can pulverize animals and demolish buildingsvehicles and infrastructure particularly where constructionalquality is poor Immediately behind the high-pressure frontviolent winds ensue that may flatten forests and scatterdebris

All of these impact-related processes combine and interactin an extremely complicated way that requires detailedobservation laboratory experiments or computer models tofully simulate and understand However with certainsimplifying assumptions we can derive reasonable estimatesof their consequences for the terrestrial environment In thefollowing sections we describe each of the steps that allow usto achieve this in the Earth Impact Effects Program We discusshow our program estimates 1) the impact energy and averagetime interval between impacts of the same energy somewhereon Earth 2) the consequences of atmospheric entry 3) forcrater forming events the resulting crater size and volume ofthe melt produced 4) the thermal radiation damage from thefireball 5) the impact-induced seismic shaking 6) the extentand nature of the ejecta deposit and 7) the damage caused bythe blast wave To clearly identify our algorithm in thefollowing discussion all of the equations that we implement inthe code are labeled with an asterisk ()

To make the program accessible to the broadest range ofusers it was written with as few input parameters as possibleThe program requests six descriptors which are illustratedschematically in Fig 1 the diameter of the impactor L0 (we usethe term impactor to denote the asteroid comet or otherextraterrestrial object considered) the impactor density ρi theimpact velocity v0 the angle that the trajectory of the impactorsubtends with the surface of the Earth at the impact point θ thetarget type and the distance away from the impact at which theuser wishes to calculate the environmental consequences rThree target types are possible sedimentary rock for which weassign a target density of ρt = 2500 kg mminus3 crystalline rock (ρt= 2750 kg mminus3) or a marine target for which the programrequests a water-layer depth dw and assigns a density of ρw =1000 kg mminus3 for the water and a target density of ρt = 2700 kgmminus3 for the rock layer below The program offers the user avariety of options for units however in this paper the units forall variables are the SI units (mks) unless otherwise stated

IMPACT ENERGY AND RECURRENCE INTERVAL

The most fundamental quantity in assessing theenvironmental consequences of the impact is the energyreleased during the impact which is related to the kineticenergy of the impactor E before atmospheric entry begins Atnormal solar system impact speeds E is approximately givenas one half times the impactor mass mi times the square of theimpactor velocity v0 which can be rewritten in terms of themeteoroidrsquos density ρi and diameter L0 assuming that themeteoroid is approximately spherical

(1)

In fact the program uses the relativistic energy equationto accommodate the requests of several science fictionwriters The program does not limit the impact velocity to

E 12---miv0

2 π12------ρiL0

3v02= =

Earth Impact Effects Program 819

72 km sminus1 the maximum possible for an impactor bound tothe Sun however we have limited the maximum velocity tothe speed of light in response to attempts of a few users toinsert supra-light velocities

Natural objects that encounter the Earth are eitherasteroids or comets Asteroids are made of rock (ρi ~2000ndash3000 kg m3 Hilton 2002) or iron (ρi ~8000 kg m3) andtypically collide with the Earthrsquos atmosphere at velocities of12ndash20 km sminus1 (Bottke et al 1994) Detailed knowledge of thecomposition of comets is currently lacking however they areof much lower density (ρi ~500ndash1500 kg m3) and are composedmainly of ice (Chapman and Brandt 2004) Typical velocitiesat which comets might encounter the Earthrsquos atmosphere are inthe range of 30ndash70 km sminus1 (Marsden and Steel 1994) Thus anasteroid or comet typically has 4ndash20 times the energy per unitmass of TNT at the moment atmospheric entry beginsTherefore impact events have much in common with chemicaland nuclear explosions a fact that we will rely on later in ourestimates of the environmental effects of an impact

Observations of near-Earth objects made by severaltelescopic search programs show that the number of near-Earth asteroids with a diameter greater than Lkm (in km) maybe expressed approximately by the power law (Near-EarthObject Science Definition Team 2003)

N(gtL) asymp 1148Lkmminus2354 (2)

These data may also be represented in terms of therecurrence interval TRE in years versus the impact energy EMtin megatons of TNT by assuming a probability of a single-object collision with Earth (~16 times 10minus9 yrminus1 Near-Earth Object

Science Definition Team 2003 their Fig 23) and multiplyingby the number of asteroids of a given potential impact energythat are estimated to be circling the sun with potentiallyhazardous Earth-crossing orbits We found that a simplepower-law relationship adequately represents these data

TRE asymp 109EMt078 (3)

Thus for a given set of user-input impact parameters (L0v0 ρi ρt and θ) the program computes the kinetic energy(EMt in megatons 1 Mt = 418 times 1015 J) possessed by theimpacting body when it hits the upper atmosphere and definesan average time interval between impacts of that energysomewhere on the Earth Furthermore we estimate therecurrence interval TRL for impacts of this same energy withina certain specified distance r of the impact This is simply theproduct of the recurrence interval for the whole Earth and thefraction of the Earthrsquos surface area that is within the distance r

(4)

where ∆ is the epicentral angle from the impact point to arange r (given in radians by ∆ = rRE where RE is the radiusof the Earth Fig 1)

Currently the relative importance of comets to the Earth-crossing impactor flux is not well-constrained The Near-EarthObject Science Definition Team (2003) suggests that cometscomprise only about 1 of the estimated population of smallNEOs however there is evidence to suggest that at largersizes comets may comprise a significantly larger proportion ofthe impactor flux (Shoemaker et al 1990) Of the asteroids thatcollide with the Earthrsquos atmosphere the current best estimateis that approximately 2ndash10 are iron asteroids (Bland andArtemieva 2003) based on NEO and main-belt asteroidspectroscopy (Bus et al 2002 Binzel et al 2003) meteoritecomposition and the impactor types in large terrestrial craters

ATMOSPHERIC ENTRY

Atmospheric entry of asteroids has been discussed indetail by many authors (Chyba et al 1993 Ivanov et al 1997Krinov 1966 Melosh 1981 Passey and Melosh 1980 Svetsovet al 1995 Korycansky et al 2000 2002 Korycansky andZahnle 2003 2004 Bland and Artemieva 2003) and is nowunderstood to be a complex process involving interaction ofthe atmosphere and fragmenting impactor in the Earthrsquosgravitational field For the purposes of a simple program of thetype that we have created many of the refinements nowunderstood are too complex to be included Therefore wehave opted to make a number of drastic simplifications thatwe believe will still give a good description of the basicevents during atmospheric entry for most cases Of course forrefined predictions a full simulation using all of the knownprocesses and properties must be undertaken Atmosphericentry has no significant influence on the shape energy or

Fig 1 Diagram illustrating the input parameters for the Earth ImpactEffects Program L0 is the impactor diameter at the top of theatmosphere v0 is the velocity of the impactor at the top of theatmosphere ρi is the impactor density ρt is the target density and θis the angle subtended between the impactorrsquos trajectory and thetangent plane to the surface of the Earth at the impact point Thedistance r from the impact site at which the environmentalconsequences are determined is measured along the surface of theEarth the epicentral angle ∆ between the impact point and thisdistance r is given by ∆ = rRE where RE is the radius of the Earth

TRLTRE

2---------- 1 ∆cosndash( )=

820 G S Collins et al

momentum of impactors with a mass that is much larger thanthe mass of the atmosphere displaced during penetration Forthis reason the program procedure described below is appliedonly for impactors less than 1 km in diameter

For the purposes of the Earth Impact Effects Program weassume that the trajectory of the impactor is a straight linefrom the top of the atmosphere to the surface sloping at aconstant angle to the horizon given by the user Accelerationof the impactor by the Earthrsquos gravity is ignored as isdeviation of the trajectory toward the vertical in the case thatterminal velocity is reached as it may be for small impactorsThe curvature of the Earth is also ignored The atmosphere isassumed to be purely exponential with the density given by

ρ(z) = ρ0eminuszH (5)

where z is the altitude above the surface H is the scale heighttaken to be 8 km on the average Earth and ρ0 is the surfaceatmospheric density taken to be equal to 1 kgm3

During the first portion of the impactorrsquos flight its speedis decreased by atmospheric drag but the stresses are toosmall to cause fragmentation Small meteoroids are oftenablated to nothing during this phase but in the currentprogram implementation we ignore ablation on the groundsthat it seldom affects the larger impactors that reach thesurface to cause craters Thus this program should not beused to estimate the entry process of small objects that maycause visible meteors or even drop small meteorites to thesurface at terminal velocity

While the body remains intact the diameter of theincoming impactor is constant equal to the diameter L0 givenby the user The rate of change of the velocity v is given by theusual drag equation (corrected from Melosh 1989 chapter 11)

(6)

where CD is the drag coefficient taken to equal 2 and ρi is theimpactor density (an input parameter) This equation can begreatly simplified by making the replacement dt = minusdzv sinθ(justified by our assumption that the impactor travels in astraight line) and rearranging

(7)

Integration of this equation using the exponential densitydependence gives the velocity of the impactor as a function ofaltitude

(8)

where θ is the entry angle and v0 is the impact velocity at thetop of the atmosphere given by the user

As the impactor penetrates the atmosphere theatmospheric density increases and the stagnation pressure at

the leading edge of the impactor Ps = ρ(z) v(z)2 risesEventually this exceeds the strength of the impactor and itbegins to break up Observed meteoroids often undergoseveral cascades of breakup reflecting components of widelyvarying strengths The entire subject of meteoroid strength ispoorly understood as measured crushing strengths ofspecimens collected on the ground are often a factor of 10 lessthan strengths inferred from observed breakup (Svetsov et al1995) Clearly strong selection effects are at work For thepurposes of our program we decided not to embroil the userin the ill-defined guesswork of estimating meteoroid crushingstrength Instead we found a rough correlation betweendensity and estimated strength for comets (about 15 Pa intension from the tidal breakup of SL-9 Scotti and Melosh1993) chondrites (Chyba et al 1993) and iron or stoneobjects (Petrovic 2001) Based on four simplified estimatesfor comets carbonaceous stony and iron meteorites weestablished an empirical strength-density relation for use inthe program The yield strength Yi of the impactor in Pa is thuscomputed from

(9)

where the impactor density ρi is in kg mminus3 Note that even atzero density this implies a non-zero strength of about 130 PaThus this empirical formula should not be applied too far outof the range of 1000 to 8000 kg mminus3 over which it wasestablished

Using this estimate of strength and comparing it to thestagnation pressure we can compute an altitude of breakup zby solving the transcendental equation

Yi = ρ(z)v2(z) (10)

Rather than solving this equation in the program directlyan excellent analytic approximation to the solution was foundand implemented

(11)

where If is given by

(12)

In certain specific instances (ie small strongimpactors) the impactor may reach the surface intact in thiscase If gt1 and Equation 11 does not apply The properlydecremented velocity calculated using Equation 8 is used tocompute a crater size (If this velocity happens to be less thanthe terminal velocity then the maximum of the two is usedinstead) The velocity at the top of the atmosphere and at thesurface is reported

Most often the impactor begins to break up well abovethe surface in this case If lt1 and Equation 11 is used to

dvdt------

3ρzCD4ρiL0-----------------ndash v2=

d ln vdz-------------

3ρ z( )CD4ρiL0 θsin--------------------------=

v z( ) v03ρ z( )CDH4ρiL0 θsin---------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

log10Yi 2107 00624+ ρi=

z HndashYi

ρ0vi2-----------

⎝ ⎠⎜ ⎟⎛ ⎞

ln 1308 0314If 1303 1 Ifndashndashndash+asymp

If 407CDHYi

ρiL0vi2 θsin

----------------------------=

Earth Impact Effects Program 821

compute the breakup altitude z After breakup the fragmentsbegin to disperse in a complex series of processes (Passey andMelosh 1980 Svetsov et al 1995) that require detailednumerical treatment However a simple approximation to thiscascade was found (Chyba et al 1993 Melosh 1981) calledthe pancake model that does a good job for Tunguska-classevents The basic idea of this model is that the impactor oncefractured expands laterally under the differential pressurebetween the front and back surfaces The front of the impactoris compressed at the stagnation pressure and the rear isessentially in a vacuum with zero pressure The sides squirtout at a rate determined by force balance in an inviscid fluidThis leads to a simple equation for the expansion of theimpactor diameter L now a function of time

(13)

The initial condition is that L = L0 at z = z If L does notincrease too much over the scale height H the timederivatives can be replaced with altitude derivatives (Chybaet al 1993) and a nonlinear differential equation can beconstructed that does not contain v(z)

(14)

Again we construct an analytic approximation to the fullsolution of this equation which is adequate for the purposesof the program

(15)

where the dispersion length scale l is given by

(16)

The velocity as a function of altitude is then given byinserting this expression for L(z) into the drag equation andintegrating downward from the breakup altitude z Becauseof the rapid expansion of the pancake the drag rises rapidly aswell and the velocity drops as a double exponential

(17)

The crushed impactor spreads laterally until the ratioL(z)L0 reaches a prescribed limit which we call the ldquopancakefactorrdquo fp In reality this should be no larger than 2 to 4(Ivanov et al 1997) after which the fragments are sufficientlyseparated that they follow independent flight paths and may

suffer one or more further pancake fragmentation eventsHowever Chyba et al (1993) obtained good agreement withTunguska-class events using pancake factors as large as 5ndash10In this work we experimented with different factors andsettled on a value of 7 to terminate the dispersion of theimpactor The altitude at which this dispersion is obtained iscalled the ldquoairburst altituderdquo (zb see Fig 2a) it is given bysubstituting fp = L(z)L0 into Equation 15 and rearranging

(18)

If the airburst occurs above the surface (Fig 2a) most ofthe energy is dissipated in the air We report the airburstaltitude zb and the residual velocity of the swarm which iscomputed using Equation 17 In this case the integral in theexponent evaluated from the airburst altitude to thedisruption altitude is given by

(19)

with the definition The surface impact velocityof the remnants from the airburst vi is also reported as themaximum of the terminal velocity of a fragment half thediameter of the original impactor or the velocity of theswarm as a whole The spreading velocity at airburstmultiplied by the time to impact is added to the breadth ofthe swarm to estimate the dispersion of what will be a strewnfield on the surface The principal environmentalconsequence of such an event is a strong blast wave in theatmosphere (see below)

On the other hand if the pancake does not spread to thelimiting size before it reaches the ground (zb le0 inEquation 19 Fig 2b) the swarm velocity at the moment ofimpact is computed using Equation 17 In this case theintegral in the exponent evaluated from the surface (z = 0) tothe disruption altitude is given by

(20)

The dispersion of the swarm at impact is compared to theestimated transient crater size (see below) and if it iscomparable or larger then the formation of a crater field isreported similar to that actually observed at HenburyAustralia Otherwise we assume the impact to be a crater-

d2Ldt2---------

CDPsρiL

-------------CDρ z( )v2 z( )

ρiL--------------------------------= =

Ld2Ldz2---------

CDρ z( )

ρisin2θ-------------------=

L z( ) L0 1 2Hl

-------⎝ ⎠⎛ ⎞ 2 z zndash

2H-------------

⎩ ⎭⎨ ⎬⎧ ⎫

exp 1ndash⎝ ⎠⎜ ⎟⎛ ⎞

2

+=

l L0 θρi

CDρ z( )---------------------sin=

v z( ) v z( ) 34---ndash

CDρ z( )

ρiL03 θsin

---------------------- ez zndash( ) Hfrasl

z

z

int L2 z( )dz

⎩ ⎭⎪ ⎪⎨ ⎬⎪ ⎪⎧ ⎫

exp=

zb z 2H 1 l2H------- fp

2 1ndash+lnndash=

ez zndash( ) Hfrasl

zburst

z

int L2 z( )dz

lL0

2

24--------α 8 3 α2+( ) 3α l

H---- 2 α2+( )+

=

α fp2 1ndashequiv

ez zndash( ) Hfrasl

0

z

int L2 z( )dz H3L0

2

3l2------------- 34 lH----⎝ ⎠

⎛ ⎞ 2+ e

z Hfrasl

6e2z Hfrasl

16e3z 2Hfrasl

3ndashndash

+

lH----⎝ ⎠

⎛ ⎞ 22ndash

=

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 3: Documentation

Earth Impact Effects Program 819

72 km sminus1 the maximum possible for an impactor bound tothe Sun however we have limited the maximum velocity tothe speed of light in response to attempts of a few users toinsert supra-light velocities

Natural objects that encounter the Earth are eitherasteroids or comets Asteroids are made of rock (ρi ~2000ndash3000 kg m3 Hilton 2002) or iron (ρi ~8000 kg m3) andtypically collide with the Earthrsquos atmosphere at velocities of12ndash20 km sminus1 (Bottke et al 1994) Detailed knowledge of thecomposition of comets is currently lacking however they areof much lower density (ρi ~500ndash1500 kg m3) and are composedmainly of ice (Chapman and Brandt 2004) Typical velocitiesat which comets might encounter the Earthrsquos atmosphere are inthe range of 30ndash70 km sminus1 (Marsden and Steel 1994) Thus anasteroid or comet typically has 4ndash20 times the energy per unitmass of TNT at the moment atmospheric entry beginsTherefore impact events have much in common with chemicaland nuclear explosions a fact that we will rely on later in ourestimates of the environmental effects of an impact

Observations of near-Earth objects made by severaltelescopic search programs show that the number of near-Earth asteroids with a diameter greater than Lkm (in km) maybe expressed approximately by the power law (Near-EarthObject Science Definition Team 2003)

N(gtL) asymp 1148Lkmminus2354 (2)

These data may also be represented in terms of therecurrence interval TRE in years versus the impact energy EMtin megatons of TNT by assuming a probability of a single-object collision with Earth (~16 times 10minus9 yrminus1 Near-Earth Object

Science Definition Team 2003 their Fig 23) and multiplyingby the number of asteroids of a given potential impact energythat are estimated to be circling the sun with potentiallyhazardous Earth-crossing orbits We found that a simplepower-law relationship adequately represents these data

TRE asymp 109EMt078 (3)

Thus for a given set of user-input impact parameters (L0v0 ρi ρt and θ) the program computes the kinetic energy(EMt in megatons 1 Mt = 418 times 1015 J) possessed by theimpacting body when it hits the upper atmosphere and definesan average time interval between impacts of that energysomewhere on the Earth Furthermore we estimate therecurrence interval TRL for impacts of this same energy withina certain specified distance r of the impact This is simply theproduct of the recurrence interval for the whole Earth and thefraction of the Earthrsquos surface area that is within the distance r

(4)

where ∆ is the epicentral angle from the impact point to arange r (given in radians by ∆ = rRE where RE is the radiusof the Earth Fig 1)

Currently the relative importance of comets to the Earth-crossing impactor flux is not well-constrained The Near-EarthObject Science Definition Team (2003) suggests that cometscomprise only about 1 of the estimated population of smallNEOs however there is evidence to suggest that at largersizes comets may comprise a significantly larger proportion ofthe impactor flux (Shoemaker et al 1990) Of the asteroids thatcollide with the Earthrsquos atmosphere the current best estimateis that approximately 2ndash10 are iron asteroids (Bland andArtemieva 2003) based on NEO and main-belt asteroidspectroscopy (Bus et al 2002 Binzel et al 2003) meteoritecomposition and the impactor types in large terrestrial craters

ATMOSPHERIC ENTRY

Atmospheric entry of asteroids has been discussed indetail by many authors (Chyba et al 1993 Ivanov et al 1997Krinov 1966 Melosh 1981 Passey and Melosh 1980 Svetsovet al 1995 Korycansky et al 2000 2002 Korycansky andZahnle 2003 2004 Bland and Artemieva 2003) and is nowunderstood to be a complex process involving interaction ofthe atmosphere and fragmenting impactor in the Earthrsquosgravitational field For the purposes of a simple program of thetype that we have created many of the refinements nowunderstood are too complex to be included Therefore wehave opted to make a number of drastic simplifications thatwe believe will still give a good description of the basicevents during atmospheric entry for most cases Of course forrefined predictions a full simulation using all of the knownprocesses and properties must be undertaken Atmosphericentry has no significant influence on the shape energy or

Fig 1 Diagram illustrating the input parameters for the Earth ImpactEffects Program L0 is the impactor diameter at the top of theatmosphere v0 is the velocity of the impactor at the top of theatmosphere ρi is the impactor density ρt is the target density and θis the angle subtended between the impactorrsquos trajectory and thetangent plane to the surface of the Earth at the impact point Thedistance r from the impact site at which the environmentalconsequences are determined is measured along the surface of theEarth the epicentral angle ∆ between the impact point and thisdistance r is given by ∆ = rRE where RE is the radius of the Earth

TRLTRE

2---------- 1 ∆cosndash( )=

820 G S Collins et al

momentum of impactors with a mass that is much larger thanthe mass of the atmosphere displaced during penetration Forthis reason the program procedure described below is appliedonly for impactors less than 1 km in diameter

For the purposes of the Earth Impact Effects Program weassume that the trajectory of the impactor is a straight linefrom the top of the atmosphere to the surface sloping at aconstant angle to the horizon given by the user Accelerationof the impactor by the Earthrsquos gravity is ignored as isdeviation of the trajectory toward the vertical in the case thatterminal velocity is reached as it may be for small impactorsThe curvature of the Earth is also ignored The atmosphere isassumed to be purely exponential with the density given by

ρ(z) = ρ0eminuszH (5)

where z is the altitude above the surface H is the scale heighttaken to be 8 km on the average Earth and ρ0 is the surfaceatmospheric density taken to be equal to 1 kgm3

During the first portion of the impactorrsquos flight its speedis decreased by atmospheric drag but the stresses are toosmall to cause fragmentation Small meteoroids are oftenablated to nothing during this phase but in the currentprogram implementation we ignore ablation on the groundsthat it seldom affects the larger impactors that reach thesurface to cause craters Thus this program should not beused to estimate the entry process of small objects that maycause visible meteors or even drop small meteorites to thesurface at terminal velocity

While the body remains intact the diameter of theincoming impactor is constant equal to the diameter L0 givenby the user The rate of change of the velocity v is given by theusual drag equation (corrected from Melosh 1989 chapter 11)

(6)

where CD is the drag coefficient taken to equal 2 and ρi is theimpactor density (an input parameter) This equation can begreatly simplified by making the replacement dt = minusdzv sinθ(justified by our assumption that the impactor travels in astraight line) and rearranging

(7)

Integration of this equation using the exponential densitydependence gives the velocity of the impactor as a function ofaltitude

(8)

where θ is the entry angle and v0 is the impact velocity at thetop of the atmosphere given by the user

As the impactor penetrates the atmosphere theatmospheric density increases and the stagnation pressure at

the leading edge of the impactor Ps = ρ(z) v(z)2 risesEventually this exceeds the strength of the impactor and itbegins to break up Observed meteoroids often undergoseveral cascades of breakup reflecting components of widelyvarying strengths The entire subject of meteoroid strength ispoorly understood as measured crushing strengths ofspecimens collected on the ground are often a factor of 10 lessthan strengths inferred from observed breakup (Svetsov et al1995) Clearly strong selection effects are at work For thepurposes of our program we decided not to embroil the userin the ill-defined guesswork of estimating meteoroid crushingstrength Instead we found a rough correlation betweendensity and estimated strength for comets (about 15 Pa intension from the tidal breakup of SL-9 Scotti and Melosh1993) chondrites (Chyba et al 1993) and iron or stoneobjects (Petrovic 2001) Based on four simplified estimatesfor comets carbonaceous stony and iron meteorites weestablished an empirical strength-density relation for use inthe program The yield strength Yi of the impactor in Pa is thuscomputed from

(9)

where the impactor density ρi is in kg mminus3 Note that even atzero density this implies a non-zero strength of about 130 PaThus this empirical formula should not be applied too far outof the range of 1000 to 8000 kg mminus3 over which it wasestablished

Using this estimate of strength and comparing it to thestagnation pressure we can compute an altitude of breakup zby solving the transcendental equation

Yi = ρ(z)v2(z) (10)

Rather than solving this equation in the program directlyan excellent analytic approximation to the solution was foundand implemented

(11)

where If is given by

(12)

In certain specific instances (ie small strongimpactors) the impactor may reach the surface intact in thiscase If gt1 and Equation 11 does not apply The properlydecremented velocity calculated using Equation 8 is used tocompute a crater size (If this velocity happens to be less thanthe terminal velocity then the maximum of the two is usedinstead) The velocity at the top of the atmosphere and at thesurface is reported

Most often the impactor begins to break up well abovethe surface in this case If lt1 and Equation 11 is used to

dvdt------

3ρzCD4ρiL0-----------------ndash v2=

d ln vdz-------------

3ρ z( )CD4ρiL0 θsin--------------------------=

v z( ) v03ρ z( )CDH4ρiL0 θsin---------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

log10Yi 2107 00624+ ρi=

z HndashYi

ρ0vi2-----------

⎝ ⎠⎜ ⎟⎛ ⎞

ln 1308 0314If 1303 1 Ifndashndashndash+asymp

If 407CDHYi

ρiL0vi2 θsin

----------------------------=

Earth Impact Effects Program 821

compute the breakup altitude z After breakup the fragmentsbegin to disperse in a complex series of processes (Passey andMelosh 1980 Svetsov et al 1995) that require detailednumerical treatment However a simple approximation to thiscascade was found (Chyba et al 1993 Melosh 1981) calledthe pancake model that does a good job for Tunguska-classevents The basic idea of this model is that the impactor oncefractured expands laterally under the differential pressurebetween the front and back surfaces The front of the impactoris compressed at the stagnation pressure and the rear isessentially in a vacuum with zero pressure The sides squirtout at a rate determined by force balance in an inviscid fluidThis leads to a simple equation for the expansion of theimpactor diameter L now a function of time

(13)

The initial condition is that L = L0 at z = z If L does notincrease too much over the scale height H the timederivatives can be replaced with altitude derivatives (Chybaet al 1993) and a nonlinear differential equation can beconstructed that does not contain v(z)

(14)

Again we construct an analytic approximation to the fullsolution of this equation which is adequate for the purposesof the program

(15)

where the dispersion length scale l is given by

(16)

The velocity as a function of altitude is then given byinserting this expression for L(z) into the drag equation andintegrating downward from the breakup altitude z Becauseof the rapid expansion of the pancake the drag rises rapidly aswell and the velocity drops as a double exponential

(17)

The crushed impactor spreads laterally until the ratioL(z)L0 reaches a prescribed limit which we call the ldquopancakefactorrdquo fp In reality this should be no larger than 2 to 4(Ivanov et al 1997) after which the fragments are sufficientlyseparated that they follow independent flight paths and may

suffer one or more further pancake fragmentation eventsHowever Chyba et al (1993) obtained good agreement withTunguska-class events using pancake factors as large as 5ndash10In this work we experimented with different factors andsettled on a value of 7 to terminate the dispersion of theimpactor The altitude at which this dispersion is obtained iscalled the ldquoairburst altituderdquo (zb see Fig 2a) it is given bysubstituting fp = L(z)L0 into Equation 15 and rearranging

(18)

If the airburst occurs above the surface (Fig 2a) most ofthe energy is dissipated in the air We report the airburstaltitude zb and the residual velocity of the swarm which iscomputed using Equation 17 In this case the integral in theexponent evaluated from the airburst altitude to thedisruption altitude is given by

(19)

with the definition The surface impact velocityof the remnants from the airburst vi is also reported as themaximum of the terminal velocity of a fragment half thediameter of the original impactor or the velocity of theswarm as a whole The spreading velocity at airburstmultiplied by the time to impact is added to the breadth ofthe swarm to estimate the dispersion of what will be a strewnfield on the surface The principal environmentalconsequence of such an event is a strong blast wave in theatmosphere (see below)

On the other hand if the pancake does not spread to thelimiting size before it reaches the ground (zb le0 inEquation 19 Fig 2b) the swarm velocity at the moment ofimpact is computed using Equation 17 In this case theintegral in the exponent evaluated from the surface (z = 0) tothe disruption altitude is given by

(20)

The dispersion of the swarm at impact is compared to theestimated transient crater size (see below) and if it iscomparable or larger then the formation of a crater field isreported similar to that actually observed at HenburyAustralia Otherwise we assume the impact to be a crater-

d2Ldt2---------

CDPsρiL

-------------CDρ z( )v2 z( )

ρiL--------------------------------= =

Ld2Ldz2---------

CDρ z( )

ρisin2θ-------------------=

L z( ) L0 1 2Hl

-------⎝ ⎠⎛ ⎞ 2 z zndash

2H-------------

⎩ ⎭⎨ ⎬⎧ ⎫

exp 1ndash⎝ ⎠⎜ ⎟⎛ ⎞

2

+=

l L0 θρi

CDρ z( )---------------------sin=

v z( ) v z( ) 34---ndash

CDρ z( )

ρiL03 θsin

---------------------- ez zndash( ) Hfrasl

z

z

int L2 z( )dz

⎩ ⎭⎪ ⎪⎨ ⎬⎪ ⎪⎧ ⎫

exp=

zb z 2H 1 l2H------- fp

2 1ndash+lnndash=

ez zndash( ) Hfrasl

zburst

z

int L2 z( )dz

lL0

2

24--------α 8 3 α2+( ) 3α l

H---- 2 α2+( )+

=

α fp2 1ndashequiv

ez zndash( ) Hfrasl

0

z

int L2 z( )dz H3L0

2

3l2------------- 34 lH----⎝ ⎠

⎛ ⎞ 2+ e

z Hfrasl

6e2z Hfrasl

16e3z 2Hfrasl

3ndashndash

+

lH----⎝ ⎠

⎛ ⎞ 22ndash

=

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 4: Documentation

820 G S Collins et al

momentum of impactors with a mass that is much larger thanthe mass of the atmosphere displaced during penetration Forthis reason the program procedure described below is appliedonly for impactors less than 1 km in diameter

For the purposes of the Earth Impact Effects Program weassume that the trajectory of the impactor is a straight linefrom the top of the atmosphere to the surface sloping at aconstant angle to the horizon given by the user Accelerationof the impactor by the Earthrsquos gravity is ignored as isdeviation of the trajectory toward the vertical in the case thatterminal velocity is reached as it may be for small impactorsThe curvature of the Earth is also ignored The atmosphere isassumed to be purely exponential with the density given by

ρ(z) = ρ0eminuszH (5)

where z is the altitude above the surface H is the scale heighttaken to be 8 km on the average Earth and ρ0 is the surfaceatmospheric density taken to be equal to 1 kgm3

During the first portion of the impactorrsquos flight its speedis decreased by atmospheric drag but the stresses are toosmall to cause fragmentation Small meteoroids are oftenablated to nothing during this phase but in the currentprogram implementation we ignore ablation on the groundsthat it seldom affects the larger impactors that reach thesurface to cause craters Thus this program should not beused to estimate the entry process of small objects that maycause visible meteors or even drop small meteorites to thesurface at terminal velocity

While the body remains intact the diameter of theincoming impactor is constant equal to the diameter L0 givenby the user The rate of change of the velocity v is given by theusual drag equation (corrected from Melosh 1989 chapter 11)

(6)

where CD is the drag coefficient taken to equal 2 and ρi is theimpactor density (an input parameter) This equation can begreatly simplified by making the replacement dt = minusdzv sinθ(justified by our assumption that the impactor travels in astraight line) and rearranging

(7)

Integration of this equation using the exponential densitydependence gives the velocity of the impactor as a function ofaltitude

(8)

where θ is the entry angle and v0 is the impact velocity at thetop of the atmosphere given by the user

As the impactor penetrates the atmosphere theatmospheric density increases and the stagnation pressure at

the leading edge of the impactor Ps = ρ(z) v(z)2 risesEventually this exceeds the strength of the impactor and itbegins to break up Observed meteoroids often undergoseveral cascades of breakup reflecting components of widelyvarying strengths The entire subject of meteoroid strength ispoorly understood as measured crushing strengths ofspecimens collected on the ground are often a factor of 10 lessthan strengths inferred from observed breakup (Svetsov et al1995) Clearly strong selection effects are at work For thepurposes of our program we decided not to embroil the userin the ill-defined guesswork of estimating meteoroid crushingstrength Instead we found a rough correlation betweendensity and estimated strength for comets (about 15 Pa intension from the tidal breakup of SL-9 Scotti and Melosh1993) chondrites (Chyba et al 1993) and iron or stoneobjects (Petrovic 2001) Based on four simplified estimatesfor comets carbonaceous stony and iron meteorites weestablished an empirical strength-density relation for use inthe program The yield strength Yi of the impactor in Pa is thuscomputed from

(9)

where the impactor density ρi is in kg mminus3 Note that even atzero density this implies a non-zero strength of about 130 PaThus this empirical formula should not be applied too far outof the range of 1000 to 8000 kg mminus3 over which it wasestablished

Using this estimate of strength and comparing it to thestagnation pressure we can compute an altitude of breakup zby solving the transcendental equation

Yi = ρ(z)v2(z) (10)

Rather than solving this equation in the program directlyan excellent analytic approximation to the solution was foundand implemented

(11)

where If is given by

(12)

In certain specific instances (ie small strongimpactors) the impactor may reach the surface intact in thiscase If gt1 and Equation 11 does not apply The properlydecremented velocity calculated using Equation 8 is used tocompute a crater size (If this velocity happens to be less thanthe terminal velocity then the maximum of the two is usedinstead) The velocity at the top of the atmosphere and at thesurface is reported

Most often the impactor begins to break up well abovethe surface in this case If lt1 and Equation 11 is used to

dvdt------

3ρzCD4ρiL0-----------------ndash v2=

d ln vdz-------------

3ρ z( )CD4ρiL0 θsin--------------------------=

v z( ) v03ρ z( )CDH4ρiL0 θsin---------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

log10Yi 2107 00624+ ρi=

z HndashYi

ρ0vi2-----------

⎝ ⎠⎜ ⎟⎛ ⎞

ln 1308 0314If 1303 1 Ifndashndashndash+asymp

If 407CDHYi

ρiL0vi2 θsin

----------------------------=

Earth Impact Effects Program 821

compute the breakup altitude z After breakup the fragmentsbegin to disperse in a complex series of processes (Passey andMelosh 1980 Svetsov et al 1995) that require detailednumerical treatment However a simple approximation to thiscascade was found (Chyba et al 1993 Melosh 1981) calledthe pancake model that does a good job for Tunguska-classevents The basic idea of this model is that the impactor oncefractured expands laterally under the differential pressurebetween the front and back surfaces The front of the impactoris compressed at the stagnation pressure and the rear isessentially in a vacuum with zero pressure The sides squirtout at a rate determined by force balance in an inviscid fluidThis leads to a simple equation for the expansion of theimpactor diameter L now a function of time

(13)

The initial condition is that L = L0 at z = z If L does notincrease too much over the scale height H the timederivatives can be replaced with altitude derivatives (Chybaet al 1993) and a nonlinear differential equation can beconstructed that does not contain v(z)

(14)

Again we construct an analytic approximation to the fullsolution of this equation which is adequate for the purposesof the program

(15)

where the dispersion length scale l is given by

(16)

The velocity as a function of altitude is then given byinserting this expression for L(z) into the drag equation andintegrating downward from the breakup altitude z Becauseof the rapid expansion of the pancake the drag rises rapidly aswell and the velocity drops as a double exponential

(17)

The crushed impactor spreads laterally until the ratioL(z)L0 reaches a prescribed limit which we call the ldquopancakefactorrdquo fp In reality this should be no larger than 2 to 4(Ivanov et al 1997) after which the fragments are sufficientlyseparated that they follow independent flight paths and may

suffer one or more further pancake fragmentation eventsHowever Chyba et al (1993) obtained good agreement withTunguska-class events using pancake factors as large as 5ndash10In this work we experimented with different factors andsettled on a value of 7 to terminate the dispersion of theimpactor The altitude at which this dispersion is obtained iscalled the ldquoairburst altituderdquo (zb see Fig 2a) it is given bysubstituting fp = L(z)L0 into Equation 15 and rearranging

(18)

If the airburst occurs above the surface (Fig 2a) most ofthe energy is dissipated in the air We report the airburstaltitude zb and the residual velocity of the swarm which iscomputed using Equation 17 In this case the integral in theexponent evaluated from the airburst altitude to thedisruption altitude is given by

(19)

with the definition The surface impact velocityof the remnants from the airburst vi is also reported as themaximum of the terminal velocity of a fragment half thediameter of the original impactor or the velocity of theswarm as a whole The spreading velocity at airburstmultiplied by the time to impact is added to the breadth ofthe swarm to estimate the dispersion of what will be a strewnfield on the surface The principal environmentalconsequence of such an event is a strong blast wave in theatmosphere (see below)

On the other hand if the pancake does not spread to thelimiting size before it reaches the ground (zb le0 inEquation 19 Fig 2b) the swarm velocity at the moment ofimpact is computed using Equation 17 In this case theintegral in the exponent evaluated from the surface (z = 0) tothe disruption altitude is given by

(20)

The dispersion of the swarm at impact is compared to theestimated transient crater size (see below) and if it iscomparable or larger then the formation of a crater field isreported similar to that actually observed at HenburyAustralia Otherwise we assume the impact to be a crater-

d2Ldt2---------

CDPsρiL

-------------CDρ z( )v2 z( )

ρiL--------------------------------= =

Ld2Ldz2---------

CDρ z( )

ρisin2θ-------------------=

L z( ) L0 1 2Hl

-------⎝ ⎠⎛ ⎞ 2 z zndash

2H-------------

⎩ ⎭⎨ ⎬⎧ ⎫

exp 1ndash⎝ ⎠⎜ ⎟⎛ ⎞

2

+=

l L0 θρi

CDρ z( )---------------------sin=

v z( ) v z( ) 34---ndash

CDρ z( )

ρiL03 θsin

---------------------- ez zndash( ) Hfrasl

z

z

int L2 z( )dz

⎩ ⎭⎪ ⎪⎨ ⎬⎪ ⎪⎧ ⎫

exp=

zb z 2H 1 l2H------- fp

2 1ndash+lnndash=

ez zndash( ) Hfrasl

zburst

z

int L2 z( )dz

lL0

2

24--------α 8 3 α2+( ) 3α l

H---- 2 α2+( )+

=

α fp2 1ndashequiv

ez zndash( ) Hfrasl

0

z

int L2 z( )dz H3L0

2

3l2------------- 34 lH----⎝ ⎠

⎛ ⎞ 2+ e

z Hfrasl

6e2z Hfrasl

16e3z 2Hfrasl

3ndashndash

+

lH----⎝ ⎠

⎛ ⎞ 22ndash

=

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 5: Documentation

Earth Impact Effects Program 821

compute the breakup altitude z After breakup the fragmentsbegin to disperse in a complex series of processes (Passey andMelosh 1980 Svetsov et al 1995) that require detailednumerical treatment However a simple approximation to thiscascade was found (Chyba et al 1993 Melosh 1981) calledthe pancake model that does a good job for Tunguska-classevents The basic idea of this model is that the impactor oncefractured expands laterally under the differential pressurebetween the front and back surfaces The front of the impactoris compressed at the stagnation pressure and the rear isessentially in a vacuum with zero pressure The sides squirtout at a rate determined by force balance in an inviscid fluidThis leads to a simple equation for the expansion of theimpactor diameter L now a function of time

(13)

The initial condition is that L = L0 at z = z If L does notincrease too much over the scale height H the timederivatives can be replaced with altitude derivatives (Chybaet al 1993) and a nonlinear differential equation can beconstructed that does not contain v(z)

(14)

Again we construct an analytic approximation to the fullsolution of this equation which is adequate for the purposesof the program

(15)

where the dispersion length scale l is given by

(16)

The velocity as a function of altitude is then given byinserting this expression for L(z) into the drag equation andintegrating downward from the breakup altitude z Becauseof the rapid expansion of the pancake the drag rises rapidly aswell and the velocity drops as a double exponential

(17)

The crushed impactor spreads laterally until the ratioL(z)L0 reaches a prescribed limit which we call the ldquopancakefactorrdquo fp In reality this should be no larger than 2 to 4(Ivanov et al 1997) after which the fragments are sufficientlyseparated that they follow independent flight paths and may

suffer one or more further pancake fragmentation eventsHowever Chyba et al (1993) obtained good agreement withTunguska-class events using pancake factors as large as 5ndash10In this work we experimented with different factors andsettled on a value of 7 to terminate the dispersion of theimpactor The altitude at which this dispersion is obtained iscalled the ldquoairburst altituderdquo (zb see Fig 2a) it is given bysubstituting fp = L(z)L0 into Equation 15 and rearranging

(18)

If the airburst occurs above the surface (Fig 2a) most ofthe energy is dissipated in the air We report the airburstaltitude zb and the residual velocity of the swarm which iscomputed using Equation 17 In this case the integral in theexponent evaluated from the airburst altitude to thedisruption altitude is given by

(19)

with the definition The surface impact velocityof the remnants from the airburst vi is also reported as themaximum of the terminal velocity of a fragment half thediameter of the original impactor or the velocity of theswarm as a whole The spreading velocity at airburstmultiplied by the time to impact is added to the breadth ofthe swarm to estimate the dispersion of what will be a strewnfield on the surface The principal environmentalconsequence of such an event is a strong blast wave in theatmosphere (see below)

On the other hand if the pancake does not spread to thelimiting size before it reaches the ground (zb le0 inEquation 19 Fig 2b) the swarm velocity at the moment ofimpact is computed using Equation 17 In this case theintegral in the exponent evaluated from the surface (z = 0) tothe disruption altitude is given by

(20)

The dispersion of the swarm at impact is compared to theestimated transient crater size (see below) and if it iscomparable or larger then the formation of a crater field isreported similar to that actually observed at HenburyAustralia Otherwise we assume the impact to be a crater-

d2Ldt2---------

CDPsρiL

-------------CDρ z( )v2 z( )

ρiL--------------------------------= =

Ld2Ldz2---------

CDρ z( )

ρisin2θ-------------------=

L z( ) L0 1 2Hl

-------⎝ ⎠⎛ ⎞ 2 z zndash

2H-------------

⎩ ⎭⎨ ⎬⎧ ⎫

exp 1ndash⎝ ⎠⎜ ⎟⎛ ⎞

2

+=

l L0 θρi

CDρ z( )---------------------sin=

v z( ) v z( ) 34---ndash

CDρ z( )

ρiL03 θsin

---------------------- ez zndash( ) Hfrasl

z

z

int L2 z( )dz

⎩ ⎭⎪ ⎪⎨ ⎬⎪ ⎪⎧ ⎫

exp=

zb z 2H 1 l2H------- fp

2 1ndash+lnndash=

ez zndash( ) Hfrasl

zburst

z

int L2 z( )dz

lL0

2

24--------α 8 3 α2+( ) 3α l

H---- 2 α2+( )+

=

α fp2 1ndashequiv

ez zndash( ) Hfrasl

0

z

int L2 z( )dz H3L0

2

3l2------------- 34 lH----⎝ ⎠

⎛ ⎞ 2+ e

z Hfrasl

6e2z Hfrasl

16e3z 2Hfrasl

3ndashndash

+

lH----⎝ ⎠

⎛ ⎞ 22ndash

=

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 6: Documentation

822 G S Collins et al

forming event and use the velocity at the surface to computea crater size In either case the environmental consequencesof these events are calculated based on an impact energyequal to the total kinetic energy of the swarm at the moment itstrikes the surface

Although simple we have found the prescription aboveto give a fairly reasonable account of atmospheric entry overa wide range of impactor sizes and compositions Asmentioned above a much more complex treatment must bemade on a case-by-case basis if more exact results are neededIn particular our program is not capable of providing a mass-or velocity-distribution for fragmented impactors andtherefore cannot be used to model production of terrestrialcrater fields where the size of the largest crater is related to thelargest surviving fragment

CRATER DIMENSIONS AND MELT PRODUCTION

Determining the size of the final crater from a givenimpactor size density velocity and angle of incidence is not

a trivial task The central difficulty in deriving an accurateestimate of the final crater diameter is that no observational orexperimental data exist for impact craters larger than a fewtens of meters in diameter Perhaps the best approach is to usesophisticated numerical models capable of simulating thepropagation of shock waves the excavation of the transientcrater and its subsequent collapse however this method isbeyond the scope of our simple program Instead we use a setof scaling laws that extrapolate the results of small-scaleexperimental data to scales of interest or extend observationsof cratering on other planets to the Earth The first scaling lawwe apply is based on the work of Holsapple and Schmidt(1982) Schmidt and Housen (1987) and Gault (1974) andcombines a wide range of experimental cratering data (forexample small-scale hypervelocity experiments and nuclearexplosion experiments) The equation relates the density ofthe target ρt and impactor ρi (in kg mminus3) the impactordiameter after atmospheric entry L (in m) the impact velocityat the surface vi (in m sminus1) the angle of impact θ (measured tothe horizontal) and the Earthrsquos surface gravity gE (in m sminus2)

Fig 2 Schematic illustration of two atmospheric entry scenarios considered in the Earth Impact Effects Program a) the impactor (initialdiameter L0) begins to break up at an altitude z from this point the impactor spreads perpendicular to the trajectory due to the differentpressures on the front and back face We define the airburst altitude zb to be the height above the surface at which the impactor diameter L(z)= 7L0 All the impact energy is assumed to be deposited at this altitude no crater is formed but the effects of the blast wave are estimated b)the impactor breaks up but the critical impactor diameter is not reached before the fragmented impactor strikes the surface (z gt0 zb lt0) Thecluster of fragments impacts the target surface with a velocity vi forming a single crater or crater field depending on the lateral spread of thecluster L(z = 0)sinθ

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 7: Documentation

Earth Impact Effects Program 823

to the diameter of the transient crater Dtc (in m) as measuredat the pre-impact target surface (Fig 3a)

(21)

This equation applies for impacts into solid rock targetswhere gravity is the predominant arresting influence in cratergrowth which is the case for all terrestrial impacts larger thana couple of hundred meters in diameter For impacts intowater the constant 1161 must be replaced by 1365 (Schmidtand Housen 1987) In reality these constants are not known tothree decimal places the values quoted serve as a bestestimate within a range of 08 to 15

The transient crater is only an intermediate step in thedevelopment of the final crater (Fig 3) To estimate the finalcrater diameter we must consider the effect of the transient

craterrsquos collapse using another scaling law For craterssmaller than ~32 km in diameter on Earth (classified byDence [1965] as ldquosimplerdquo based on their intuitivemorphology) the collapse process is well-understoodhighly brecciated and molten rocks that were originallypushed out of the opening crater slide back down the steeptransient cavity walls forming a melt-and-breccia lens at thebase of the crater (Grieve et al 1977 Fig 3a) To derive anestimate of the final crater diameter for simple craters weapplied an analytical model for the collapse of simplecraters originally developed by Grieve and Garvin (1984) totwo terrestrial craters for which good observational data onbreccia-lens volume and final crater dimensions exist Inmatching the observational data to model predictions wefound that an excellent first order approximation is that thefinal rim-to-rim diameter Dfr for a simple crater is givenapproximately by

Fig 3 Symbols used in the text to denote the various dimensions of an impact crater a) Transient crater dimensions Dtc is the transient craterdiameter measured at the pre-impact surface Dtr is the diameter of the transient crater measured from rim crest to rim crest htr is the rim heightof the transient crater measured from the pre-impact surface dtc is the depth of the transient crater measured from the pre-impact surface (weassume that Dtc = 2 dtc) b) simple crater dimensions (the transient crater outline is shown by the dotted line) Dfr is the rim-to-rim diameterhfr is the rim height above the pre-impact surface tbr is the breccia lens thickness dfr is the crater depth measured from the crater floor (abovethe breccia lens) to the rim crest We assume that the base of the breccia lens coincides with the floor of the transient crater at a depth of dtcbelow the pre-impact surface therefore dfr = dtc + hfr minus tbr c) complex crater dimensions Dfr is the rim-to-rim diameter hfr is the rim heightabove the pre-impact surface tm is the melt sheet thickness dfr is the crater depth measured from the crater floor (above the melt sheet) to therim crest

2

Dtc 1161ρiρt----⎝ ⎠

⎛ ⎞1 3frasl

L078vi044gE

022ndash θ1 3fraslsin=

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 8: Documentation

824 G S Collins et al

Dfr asymp 125Dtc (22)

if the unbulked breccia lens volume Vbr (ie the observedvolume of the breccia lens multiplied by a 90ndash95 bulkingcorrection factor Grieve and Garvin 1984) is assumed to berelated to the final crater diameter by

Vbr asymp 0032Dfr3 (23)

This approximate relationship is based on estimates ofunbulked breccia-lens volumes at Meteor Crater and BrentCrater (Grieve and Garvin 1984)

The model may also be used to estimate the thicknessof the breccia lens the depth to the base of the breccia lensand the final depth of the crater Assuming that the topsurface of the breccia lens is parabolic and that thebrecciation process increases the bulk volume of thismaterial by 10 the thickness of the breccia lens tbr isgiven approximately by

(24)

where dtc is the transient crater depth (below the originalground plane) and hfr is the rim height (above the originalground plane) of the final crater (see the section below onejecta deposits) The depth to the base of the breccia lens istaken to be the same as the transient crater depth dtc which weassume is given by

(25)

based on observations by Dence et al (1977) The depth ofthe final crater from the rim to the crater floor dfr is thensimply (see Fig 3b)

dfr = dtc + hfr minus tbr (26)

For craters larger than 32 km on Earth (termed complexbecause of their unintuitive morphology after Dence [1965])the collapse process is less well-understood and involves thecomplicated competition between gravitational forcestending to close the transient crater and the strengthproperties of the post-impact target rocks Several scalinglaws exist for estimating the rim-to-rim diameter of acomplex crater from the transient crater diameter or viceversa based on reconstruction of the transient craters oflunar complex craters (see for example Croft 1985McKinnon and Schenk 1985 Holsapple 1993) We use thefunctional form

(27)

established by McKinnon and Schenk (1985) which liesintermediate between the estimates of Croft (1985) and

Holsapple (1993) In this equation Dc is the diameter atwhich the transition from simple to complex crater occurs(taken to be 32 km on Earth) both Dtc and Dfr are in km (SeeFig 3b) If the transient crater diameter is greater than256 km we apply Equation 27 to determine the final craterdiameter and report that a ldquocomplexrdquo crater is formedotherwise we apply Equation 22 and report that a ldquosimplerdquocrater is formed It is worth emphasizing that the final craterdiameter that the program reports is the diameter of the freshcrater measured from rim crest to rim crest (see Figs 3b and3c) The topographic rim is likely to be strongly affected bypost-impact erosion Furthermore multiple concentric zonesof structural deformation are often observable at terrestrialimpact structuresmdasha fact that has led to uncertainty in therelationship between the structural (apparent) andtopographic (rim-to-rim) crater diameter (Turtle et al 2005)Therefore the results of the scaling arguments above shouldbe compared with caution to apparent diameters of knownterrestrial impact structures

To estimate the average depth dfr (in km) from the rim tofloor of a complex crater of rim-to-rim diameter Dfr (in km)we use the depth-to-diameter relationship of Herrick et al(1997) for venusian craters

dfr = 04Dfr03 (28)

The similarity in surface gravity between Earth andVenus as well as the large number of fresh complex craters onVenus makes this relationship more reliable than that basedon the limited and erosion-affected data for terrestrialcomplex craters (Pike 1980 Grieve and Therriault 2004)

We also estimate the volume of melt produced duringthe impact event based on the results of numerical modelingof the early phase of the impact event (OrsquoKeefe and Ahrens1982b Pierazzo et al 1997 Pierazzo and Melosh 2000) andgeological observation at terrestrial craters (Grieve andCintala 1992) Provided that 1) the impact velocity is inexcess of ~12 km sminus1 (the threshold velocity for significanttarget melting OrsquoKeefe and Ahrens 1982b) 2) the densityof the impactor and target are comparable and 3) all impactsare vertical these data are well-fit by the simple expression

(29)

where Vm is the volume of melt produced Vi is the volume ofthe impactor and εm is the specific energy of the Rankine-Hugoniot state from which the isentropic release ends at the1 bar point on the liquidus To avoid requiring further inputparameters in our program we use εm = 52 MJkg for granite(see Pierazzo et al 1997) which we take as representative ofupper-crustal rocks and assume an impactor and targetdensity of 2700 kg mminus3 This allows us to rewrite Equation 29giving the impact melt volume Vm (in m3) in terms of just theimpact energy E (in J) Vm = 89 times 10minus12 E

To account for the effect of impact angle on impact melt

tbr 28Vbrdtc hfr+

dtcDfr2--------------------

⎝ ⎠⎜ ⎟⎛ ⎞

=

dtc Dtc 2 2( )frasl=

Dfr 117Dtc

113

Dc013------------=

Vm 025vi

2

εm------Vi=

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 9: Documentation

Earth Impact Effects Program 825

production we assume based on numerical modeling work(Pierazzo and Melosh 2000 Ivanov and Artemieva 2002) thatthe volume of impact melt is roughly proportional to thevolume of the transient crater In our program the diameterand depth of the transient crater are proportional to sin13θ(Equations 21 and 25) hence the volume of the transientcrater is proportional to sinθ The equation used in ourprogram to compute the impact melt volume is therefore

Vm = 89 times 10minus12 E sinθ (30)

This expression works well for all geologic materialsexcept ice In this case Vm is about ten times larger than forrock (Pierazzo et al 1997) Equation 30 neglects the effect ofgeothermal gradient on melt production For very largeimpacts which affect rocks deep in the Earth where ambienttemperatures are much closer to the melting point thisexpression will underestimate the volume of melt producedEquation 30 agrees well with model predictions (Pierazzo andMelosh 2000) of impact melt volume versus impact angle forimpact angles greater than ~15deg to the horizontal for impactangles of ~15deg or less Equation 30 probably overestimatesthe volume of impact melt produced by a factor of ~2

In simple craters the melt is well-mixed within thebreccia lens on the floor of the crater in larger complexcraters however the melt forms a coherent sheet whichusually has an approximately uniform thickness across thecrater floor (Grieve et al 1977) Here we assume that thecrater floor diameter is similar to the transient crater diameter(Croft 1985) Thus we estimate the average thickness of thissheet tm as the ratio of the melt volume to the area of a circleequal in diameter to the transient crater

tm = 4VmπDtc2 (31)

In extremely large terrestrial impact events (Dtcgt1500 km) the volume of melt produced as predicted byEquation 30 is larger than the volume of the crater In this casewe anticipate that the transient crater would collapse to ahydrostatic almost-featureless surface and therefore ourprogram does not quote a final crater diameter Instead of atopographically observable crater the program postulates thata large circular melt province would be formed We notehowever that no such feature has been unequivocallyidentified on Earth Our program also compares the volume ofimpact-generated melt to the volume of the Earth and reportsthe fraction of the planet that is melted in truly gigantic impacts

THERMAL RADIATION

As alluded to above the compression of the target andimpactor during the initial stages of an impact eventdrastically raises the temperature and pressure of a smallregion proximal to the impact site For impacts at a velocitygreater than ~12 km sminus1 the shock pressures are high enoughto melt the entire impactor and some target material

vaporization also occurs for impacts at velocities greater than~15 km sminus1 Any vapor produced is initially at very highpressure (gt100 GPa) and temperature (gt10000 K) and thusbegins to rapidly inflate the expanding hot vapor plume istermed the ldquofireballrdquo The high temperatures imply thatthermal radiation is an important part of the energy balance ofthe expanding plume Initially the fireball is so hot that the airis ionized and its radiation absorption properties aresubstantially increased As a result the fireball is initiallyopaque to the emitted radiation which remains bottled upwithin the ball of plasma The actual process is much morecomplex than the simple description here and we refer theinterested reader to Glasstone and Dolan (1977) for a morecomplete exposition With continued expansion the fireballcools as the temperature approaches a critical temperatureknown as the transparency temperature T (Zelrsquodovich andRaizer 1966 p 607) the opacity rapidly diminishes and thethermal radiation escapes bathing the Earthrsquos surface in heatfrom the fireball The thermal radiation lasts for a few secondsto a few minutes the radiation intensity decays as theexpanding fireball rapidly cools to the point where radiationceases For Earthrsquos atmosphere the transparency temperatureis ~2000ndash3000 K (Nemtchinov et al 1998) hence thethermal radiation is primarily in the visible and infraredwavelengthsmdashthe fireball appears as a ldquosecond sunrdquo in thesky The transparency temperature of silicate vapor is about6000 K (Melosh et al 1993) so that the limiting factor forterrestrial impacts is the transparency temperature of airsurrounding the silicate vapor fireball

Provided that the impact velocity is in excess of 15 km sminus1we estimate the fireball radius Rf at the moment thetransparency temperature is achieved which we consider to bethe time of maximum radiation Numerical simulations of vaporplume expansion (Melosh et al 1993 Nemtchinov et al 1998)predict that the fireball radius at the time of maximum radiationis 10ndash15 times the impactor diameter We use a value of 13 andassume ldquoyield scalingrdquo applies to derive a relationship betweenimpact energy E in joules and the fireball radius in meters

Rf = 0002E13 (32)

Yield scaling is the empirically derived concept thatcertain length and time scales measured for two differentexplosions (or impacts) are approximately identical if dividedby the cube root of the yield (or impact) energy Yield scalingcan be justified theoretically provided that gravity and rate-dependent processes do not strongly influence the measuredparameters (Melosh 1989 p 115) The constant inEquation 32 was found by dividing the fireball radius (givenby Rf = 13L0) by the cube root of the impact energy (given byEquation 1) for a typical impactor density (2700 kg mminus3) andterrestrial impact velocity (20 km sminus1)

The time at which thermal radiation is at a maximum Tt isestimated by assuming that the initial expansion of the fireballoccurs at approximately the same velocity as the impact

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 10: Documentation

826 G S Collins et al

(33)

To calculate the environmental effects of the thermalradiation from the fireball we consider the heating at alocation a distance r from the impact site The total amount ofthermal energy emitted as thermal radiation is some smallfraction η (known as the ldquoluminous efficiencyrdquo) of the impactenergy E The luminous efficiency for hypervelocity impactsis not presently well-constrained Numerical modeling results(Nemtchinov et al 1998) suggest that η scales as some powerlaw of impact velocity The limited experimentalobservational and numerical results that exist indicate thatfor typical asteroidal impacts with Earth η is in the range of10minus4ndash10minus2 (Ortiz et al 2000) for a first-order estimate weassume η = 3 times 10minus3 and ignore the poorly-constrainedvelocity dependence

The thermal exposure Φ quantifies the amount of heatingper unit area at our specified location Φ is given by the totalamount of thermal energy radiated ηE divided by the areaover which this energy is spread (the surface area of ahemisphere of radius r 2πr2)

(34)

The total thermal energy per unit area Φ that heats ourlocation of interest arrives over a finite time period betweenthe moment the fireball surface cools to the transparencytemperature and is unveiled to the moment when the fireballhas expanded and cooled to the point where radiation ceasesWe define this time period as the ldquoduration of irradiationrdquo τtWithout computing the hydrodynamic expansion of the vaporplume this duration may be estimated simply by dividing thetotal energy radiated per unit area (total thermal energyemitted per unit area of the fireball) by the radiant energyflux given by σT

4 where σ = 567 times 10minus8 W mminus2 Kminus4 is theStefan-Bolzmann constant In our program we use T =3000 K Then the duration of irradiation is

(35)

For situations where the specified distance away from theimpact point is so far that the curvature of the Earth implies thatpart of the fireball is below the horizon we modify the thermalexposure Φ by multiplying by the ratio f of the area of thefireball above the horizon to the total area This is given by

(36)

In this equation h is the maximum height of the fireballbelow the horizon as viewed from the point of interest givenby

h = (1 minus cos∆)RE (37)

where ∆ is the epicentral angle between the impact point andthe point of interest and RE is the radius of the Earth Ifh geRf then the fireball is entirely below the horizon in thiscase no direct thermal radiation will reach our specifiedlocation The angle δ in Equation 36 is half the angle of thesegment of the fireball visible above the horizon given byδ = cosminus1 hRf We presently ignore atmospheric refractionand extinction for rays close to the horizon (this effect isimportant only over a small range interval)

Whether a particular material catches fire as a result ofthe fireball heating depends not only on the corrected thermalexposure fΦ but also on the duration of irradiation Thethermal exposure Φignition (J mminus2) required to ignite a materialthat is to heat the surface to a particular ignition temperatureTignition is given approximately by

(38)

where ρ is the density cp is the heat capacity and κ is thethermal diffusivity of the material being heated Thisexpression equates the total radiant energy received per unitarea on the left to the heat contained in a slab of unit areaperpendicular to the fireball direction on the right Thethickness of the slab is estimated from the depth penetrated by the thermal wave during the irradiation time τtAnalysis of Equation 35 shows that τt is proportional to thethermal exposure divided by the fireball radius squaredHence the duration of irradiation is proportional to E13 andthe thermal exposure required to ignite a given material isproportional to E16 This simple relationship is supported byempirical data for the ignition of various materials by thermalradiation from nuclear explosion experiments over a range ofthree orders of magnitude in explosive yield energy(Glasstone and Dolan 1977 p 287ndash289) Thus although amore energetic impact event or explosion implies a greatertotal amount of thermal radiation this heat arrives over alonger period of time and hence there is more time for heatto be diluted by conduction through the material This resultsin a greater thermal exposure being required to ignite thesame material during a more energetic impact event

To account for the impact-energy dependence of thethermal exposure required to ignite a material (or cause skindamage) we use a simple scaling law We estimate thethermal exposure required to ignite several differentmaterials or burn skin during an impact of a given energy bymultiplying the thermal exposure required to ignite thematerial during a 1 Mt event (see Table 1 data fromGlasstone and Dolan 1977 p 287ndash289) by the impact energy(in MT) to the one-sixth power

Φignition(E) = Φignition(1 Mt)EMt16 (39)

To assess the extent of thermal radiation damage at ourlocation of interest we compute the thermal radiation

TtRfvi

--------=

Φ ηE2πr2-----------=

τtηE

2πRf2 σT

4--------------------------=

f 2π--- δ h

Rf-------- δsinndash⎝ ⎠

⎛ ⎞=

Φignition Tignitionρcp κτtasymp

κτt

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 11: Documentation

Earth Impact Effects Program 827

exposure fΦ and compare this with Φignition (calculated usingEquation 39) for each type of damage in Table 1 For thermalexposures in excess of these ignition exposures we report thatthe material ignites or burns

Our simple thermal radiation model neglects the effect ofboth atmospheric conditions (cloud fog etc) and thevariation in atmospheric absorption with altitude above thehorizon Experience from nuclear weapons testing (Glasstoneand Dolan 1977 p 279) suggests that in low visibilityconditions the reduction in direct (transmitted) radiation iscompensated for in large part by indirect scattered radiationfor distances less than about half the visibility range Thisobservation led Glasstone and Dolan (1977) to conclude thatldquoas a rough approximation the amount of thermal energyreceived at a given distance from a nuclear explosion may beassumed to be independent of the visibilityrdquo Hence althoughthe above estimate should be considered an upper estimate onthe severity of thermal heating it is probably quite reliableparticularly within half the range of visibility

SEISMIC EFFECTS

The shock wave generated by the impact expands andweakens as it propagates through the target Eventually allthat remains are elastic (seismic) waves that travel through theground and along the surface in the same way as those excitedby earthquakes although the structure of the seismic wavesinduced by these distinct sources is likely to be considerablydifferent

To calculate the seismic magnitude of an impact eventwe assume that the ldquoseismic efficiencyrdquo (the fraction of thekinetic energy of the impact that ends up as seismic waveenergy) is one part in ten thousand (1 times 10minus4) This value is themost commonly accepted figure based on experimental data(Schultz and Gault 1975) with a range between 10minus5ndash10minus3Using the classic Gutenberg-Richter magnitude energyrelation the seismic magnitude M is then

M = 067log10 E minus 587 (40)

where E is the kinetic energy of the impactor in Joules(Melosh 1989 p 67)

To estimate the extent of devastation at a given distancefrom a seismic event of this magnitude we determine theintensity of shaking I as defined by the Modified MercalliIntensity Scale (see Table 2) the most widely-used intensityscale developed over the last several hundred years toevaluate the effects of earthquakes We achieve this bydefining an ldquoeffective seismic magnituderdquo as the magnitudeof an earthquake centered at our specified distance away fromthe impact that produces the same ground motion amplitudeas would be produced by the impact-induced seismic shakingWe then use Table 3 after Richter (1958) to relate theeffective seismic magnitude to the Modified MercalliIntensity A range of intensities is associated with a givenseismic magnitude because the severity of shaking dependson the local geology and rheology of the ground and thepropagation of teleseismic waves for example damage inalluviated areas will be much more severe than on well-consolidated bed rock

The equations for effective seismic magnitude use curvesfit to empirical data of ground motion as a function of distancefrom earthquake events in California (Richter 1958 p 342)We use three functional forms to relate the effective seismicmagnitude Meff to the actual seismic magnitude M and thedistance from the impact site rkm (in km) depending on thedistance away from the impact site For rkm lt60 km

Meff = M minus 00238rkm (41a)

for 60 lerkm lt700 km

Meff = M minus 00048rkm minus 11644 (41b)

and for rkm ge700 km

Meff = M minus 166log10 ∆ minus 6399 (41c)

To compute the arrival time Ts of the most violent seismicshaking we assume that the main seismic wave energy is thatassociated with the surface waves Then Ts is simply the user-specified distance rkm (in km) divided by the typical surface-wave velocity of upper-crustal rocks (~5 km sminus1)

(42)

Table 1 Ignition factors for various materialsa

Material

Thermal exposure required to ignite material during a 1 Mt explosion (Φignition(1 Mt) MJ mminus2)

Clothing 10Plywood 067Grass 038Newspaper 033Deciduous trees 025Third degree burns 042Second degree burns 025First degree burns 013

aData extracted from Glasstone and Dolan (1977)

Table 2 Seismic magnitudeModified Mercalli IntensityaRichter magnitude Modified Mercalli Intensity

0ndash1 ndash1ndash2 I2ndash3 IndashII3ndash4 IIIndashIV4ndash5 IVndashV5ndash6 VIndashVII6ndash7 VIIndashVIII7ndash8 IXndashX8ndash9 XndashXI9+ XII

aBased on data from Richter (1958)

Tsrkm5--------=

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 12: Documentation

828 G S Collins et al

EJECTA DEPOSIT

During the excavation of the crater material originallysituated close to the target surface is either thrown out of thecrater on ballistic trajectories and subsequently lands to formthe ejecta deposit or is merely displaced upward and outwardto form part of the crater rim This uplifted portion of thecrater-rim material is significant close to the transient craterrim but decreases rapidly with distance such that outside twotransient-crater radii from the crater center the materialabove the pre-impact target surface is almost all ejectadeposit For simplicity we ignore the uplifted fraction of thecrater rim material We estimate the thickness of ejecta at agiven distance from an impact by assuming that the materiallying above the pre-impact ground surface is entirely ejectathat it has a maximum thickness te = htr at the transient craterrim and that it falls off as one over the distance from thecrater rim cubed

(43)

The power of minus3 is a good approximation of data fromexplosion experiments (McGetchin et al 1973) and asatisfactory compromise for results from numericalcalculations of impacts and shallow-buried nuclearexplosions which show that the power can vary between minus25and minus35

The ejecta thickness at the transient crater rim (assumedto be equal to the transient crater rim height htr) may becalculated from a simple volume conservation argumentwhere we equate the volume of the ejecta deposit and uplifted

transient crater rim Ve with the volume of the transient craterbelow the pre-impact surface Vtc For this simple model weassume that the transient crater is a paraboloid with a depth todiameter ratio of 12 Ve is given by

(44)

where Dtr is the diameter of the transient crater at the transientcrater rim (see Fig 3a) which is related to Dtc by

(45)

The volume of the transient crater is given by

(46)

Equating Ve with Vtc and rearranging to find the rimheight gives htr = Dtc141 Inserting this result intoEquation 43 gives the simple expression used in the program

(47)

Table 3 Abbreviated version of the Modified Mercalli Intensity scaleIntensity Description

I Not felt except by a very few under especially favorable conditionsII Felt only by a few persons at rest especially on upper floors of buildingsIII Felt quite noticeably by persons indoors especially on upper floors of buildings Many people do not recognize it as an

earthquake Standing motor cars may rock slightly Vibrations similar to the passing of a truckIV Felt indoors by many outdoors by few during the day At night some awakened Dishes windows doors disturbed walls

make cracking sound Sensation like heavy truck striking building Standing motor cars rocked noticeablyV Felt by nearly everyone many awakened Some dishes windows broken Unstable objects overturned Pendulum clocks

may stopVI Felt by all many frightened Some heavy furniture moved a few instances of fallen plaster Damage slightVII Damage negligible in buildings of good design and construction slight to moderate in well-built ordinary structures

considerable damage in poorly built or badly designed structures some chimneys broken VIII Damage slight in specially designed structures considerable damage in ordinary substantial buildings with partial collapse

Damage great in poorly built structures Fall of chimneys factory stacks columns monuments and walls Heavy furniture overturned

IX General panic Damage considerable in specially designed structures well-designed frame structures thrown out of plumb Damage great in substantial buildings with partial collapse Buildings shifted off foundations Serious damage to reservoirs Underground pipes broken Conspicuous cracks in ground In alluviated areas sand and mud ejected earthquake fountains sand craters

X Most masonry and frame structures destroyed with their foundations Some well-built wooden structures and bridges destroyed Serious damage to dams dikes and embankments Large landslides Water thrown on banks of canals rivers lakes etc Sand and mud shifted horizontally on beaches and flat land Rails bent slightly

XI As X Rails bent greatly Underground pipelines completely out of serviceXII As X Damage nearly total Large rock masses displaced Lines of sight and level distorted Objects thrown into the air

tehtr8------

dtrr------⎝ ⎠

⎛ ⎞3

=

2

VehtrDtr

3

8-------------- 2πrdrr3--------------

Dtr 2frasl

infin

int 2πrDtc 2frasl

Dtr 2frasl

int+4dtc

Dtc2----------r2 dtcndash

⎝ ⎠⎜ ⎟⎛ ⎞

dr=

π2--- htrDtr

2 dtcDtr

4 Dtc4ndash

4Dtc2----------------------

Dtr2 Dtc

2ndash2----------------------ndash+

⎝ ⎠⎜ ⎟⎛ ⎞

=

Dtr Dtcdtc htr+

dtc-------------------=

VtcπDtc

3

16 2-------------=

teDtc

4

112r3-------------=

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 13: Documentation

Earth Impact Effects Program 829

As this model ignores any ldquobulkingrdquo of the ejecta depositand entrainment of the substrate on which the ejecta lands itprovides a lower bound on the probable ejecta thickness Theuse of transient crater diameter instead of final crater diameteravoids the need for a separate rim height equation for simpleand complex craters Rim heights of complex craters as afraction of the final crater diameter are significantly smallerthan the scaled rim heights of simple craters because forcomplex craters the thickest part of the ejecta blanketcollapses back into the final crater during the late stages of thecratering process As this collapse process is not fullyunderstood we only report the ejecta thickness outside thefinal crater rim The final rim height of the crater which isrequired for our estimate of the breccia-lens thickness insimple craters (above) is found by inserting r = Dfr2 intoEquation 31

(48)

The outward flight of rock ejected from the crater occursin a transient rarefied atmosphere within the expandingfireball In large impacts (E gt200 Mt) the fireball radius iscomparable to the scale height of the atmosphere hence theejectarsquos trajectory takes it out of the dense part of theatmosphere allowing it to reach distances much in excess ofthe fireball radius For smaller impacts however the ejectarsquosoutward trajectory is ultimately stifled at the edge of thefireball where the atmospheric density returns to normal Weincorporate these considerations into our program by limitingthe spatial extent of the ejecta deposit to the range of thefireball for impact energies less than 200 Mt

The ejecta arrival time is determined using ballistic traveltime equations derived by Ahrens and OrsquoKeefe (1978) for aspherical planet Using a mean ejection angle of 45deg to theEarthrsquos surface allows us to estimate the approximate arrivaltime of the bulk of the ejecta In reality material is ejectedfrom the crater at a range of angles and consequently thearrival of ejecta at a given location does not occursimultaneously However this assumption allows us to writedown an exact (although complex) analytical expression forthe average travel time of the ejecta Te to our specifiedlocation

(49)

where RE is the radius of the Earth gE is the gravitationalacceleration at the surface of the Earth and ∆ is the epicentralangle between the impact point and the point of interest Theellipticity e of the trajectory of ejecta leaving the impact site atan angle of 45deg to the horizontal and landing at the point ofinterest is given by

(50)

where ve is the ejection velocity and e is negative when ve2

gERE le1 The semi-major axis a of the trajectory is given by

(51)

To compute the ejection velocity of material reaching thespecified range r = ∆RE we use the relation

(52)

which assumes that all ejecta is thrown out of the crater fromthe same point and at the same angle (45deg) to the horizontal

Equation 49 is valid only when ve2gERE le1 which

corresponds to distances from the impact site less than about10000 km (14 of the distance around the Earth) Fordistances greater than this a similar equation exists (Ahrensand OrsquoKeefe 1978) however we do not implement it in ourprogram because in this case the arrival time of the ejecta ismuch longer than one hour Consequently an accurateestimate of ejecta thickness at distal locations must take intoaccount the rotation of the Earth which is beyond the scope ofour simple program Furthermore ejecta traveling along thesetrajectories will be predominantly fine material thatcondensed out of the vapor plume and will be greatly affectedby reentry into the atmosphere which is also not consideredin our current model For ejecta arrival times longer than onehour therefore the program reports that ldquolittle rocky ejectareaches our point of interest fallout is dominated bycondensed vapor from the impactorrdquo

We also estimate the mean fragment size of the fineejecta at our specified location using results from a study ofparabolic ejecta deposits around venusian craters (Schallerand Melosh 1998) These ejecta deposits are thought to formby the combined effect of differential settling of fine ejectafragments through the atmosphere depending on fragmentsize (smaller particles take longer to drop through theatmosphere) and the zonal winds on Venus (Vervack andMelosh 1992) Schaller and Melosh (1998) compared atheoretical model for the formation of the parabolic ejectadeposits with radar observations and derived an empirical lawfor the mean diameter of impact ejecta d (in m) on Venus as afunction of distance from the crater center rkm (in km)

(53)

where Dfr is the final crater diameter measured from rim torim (in km) α = 265 and dc = 2400(Dfr2)minus162 This relationneglects the effects of the atmosphere and windtransportation on Earth which will be more significant for

hfr 007Dtc

4

Dfr3--------=

Te2a15

gERE2

----------------- 2 1ndash 1 endash1 e+------------ ∆

4---tan⎝ ⎠

⎛ ⎞ e 1 e2ndash ∆ 2frasl( )sin1 e ∆ 2frasl( )cos+---------------------------------------------

⎝ ⎠⎜ ⎟⎛ ⎞

ndashtan=

e2 12---

ve2

gERE------------- 1ndash

⎝ ⎠⎜ ⎟⎛ ⎞

2

1+=

ave

2

2gE 1 e2ndash( )----------------------------=

ve2 2gERE ∆ 2frasltan

1 ∆ 2frasltan+------------------------------------=

d dcDfr

2rkm-----------⎝ ⎠

⎛ ⎞α

=

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 14: Documentation

830 G S Collins et al

smaller fragment sizes and the disintegration of ejectaparticles as they land Thus the uncertainty in thesepredictions is greatest very close to the crater where ejectafragments are large and will break up significantly duringdeposition and at great distances from the impact pointwhere the predicted fragment size is small We circumventthis problem at small distances by not calculating the meanfragment size for ranges less than two crater radii whichroughly corresponds to the extent of the continuous ejectablanket observed around extra-terrestrial craters (Melosh1989 p 90) We also emphasize that the predicted fragmentsize is a rough mean value of the ejecta fragment size At anygiven location there will be a range of fragment sizes aroundthis mean including large bombs and very fine-grained dustwhich will arrive at different times depending on how easilythey traverse the atmosphere

AIR BLAST

The impact-induced shock wave in the atmosphere isreferred to as the air blast or blast wave The intensity of theblast depends on the energy released during the impact andthe height in the atmosphere at which the energy is depositedwhich is either zero for impacts where a crater is formed orthe burst altitude for airburst events The effects of the blastwave may be estimated by drawing on data from US nuclearexplosion tests (Glasstone and Dolan 1977 Toon et al 19941997 Kring 1997) The important quantities to determine arethe peak overpressure that is the maximum pressure inexcess of the ambient atmospheric pressure (1 bar = 105 Pa)and the ensuing maximum wind speed With these data tablescompiled by the US Department of Defense may be used topredict the damage to buildings and structures of varyingconstructional quality vehicles windows and trees

To estimate the peak overpressure for crater-formingimpacts we assume that the impact-generated shock wave inthe air is directly analogous to that generated by an explosivecharge detonated at the ground surface (surface burst) Wefound that the expression

(54)

is an excellent fit to empirical data on the decay of peakoverpressure p (in Pa) with distance r1 (in m) for a 1 kiloton(kt) surface burst (Glasstone and Dolan 1977 their Fig 366p 109) In this equation the pressure px at the crossover pointfrom ~1r23 behavior to ~1r behavior is 75000 Pa(075 bars) this occurs at a distance of 290 m

The peak overpressure resulting from an airburst isestimated using a similar suite of equations fit to empiricaldata on the peak overpressure experienced at differentdistances away from explosions detonated at various heightsabove the surface (Glasstone and Dolan 1977 p 113) Therelationship between peak overpressure and distance away

from ground zero (the location on the Earth directly below theairburst) is more complex than for a surface burst due to theinteraction between the blast wave direct from the source andthe wave reflected off the surface Within a certain distancefrom ground zero the delay between the arrival of the directwave and the reflected wave is sufficient for little constructiveinterference of the waves to occur this region is known as theregular reflection region Beyond this zone however the twowaves merge in what is known as the ldquoMach reflectionregionrdquo this effect can increase the overpressure at a givenlocation by as much as a factor of two (Glasstone and Dolan1977 p 38) Within the Mach region we found that Equation54 holds approximately provided that the crossover distancerx is increased slightly as a function of burst altitude (rx = 289+ 065zb) At distances inside the regular reflection region wefound that the peak overpressure decreases exponentiallywith distance from ground zero

(55)

where p0 and β are both functions of burst altitude

p0 = 314 times 1011zbminus26 (56a)

β = 3487zbminus173 (56b)

To extrapolate these relationships to explosions (impacts)of greater energy we again rely on yield scaling whichimplies that a specific peak overpressure occurs at a distancefrom an explosion that is proportional to the cube root of theyield energy In other words the ratio of the distance at whicha certain peak overpressure occurs to the cube root of theimpact energy (r(p)E13) is constant for all impactsTherefore the peak overpressure at the user-specifieddistance r away from an impact of energy Ekt (in kilotons) isthe same as that at a distance r1 away from an impact ofenergy 1 kt where r1 is given by

(57)

The equivalent burst altitude in a 1 kt explosion zb1 isrelated to the actual burst altitude by a similar equation zb1 =zbEkt

13To compute the peak overpressure we substitute the

scaled-distance r1 into Equation 54 or 55 depending onwhether the distance r1 lies within the Mach region or theregular reflection region for a 1 kt explosion The distancefrom ground zero to the inner edge of the Mach region rm1 insuch an explosion depends only on the altitude of burst zb1we found a good fit to the observational data with the simplefunction

(58)

ppxrx4r1---------- 1 3

rxr1----⎝ ⎠

⎛ ⎞13

+⎝ ⎠⎛ ⎞=

p p0eβrndash 1=

r1r

EkT1 3frasl

-----------=

rm1550zb1

12 550 zb1ndash( )-----------------------------------=

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 15: Documentation

Earth Impact Effects Program 831

Note that for surface bursts (zb1 = 0) the Mach region isassumed to begin at the impact point (rm1 = 0) for scaledburst-altitudes in excess of 550 m there is no Mach regionThe calculated peak overpressure can then be compared withdata presented in Table 4 to assess the extent of the air blastdamage

The characteristics of a blast wave in air at the shockfront are uniquely related by the Hugoniot equations whencoupled with the equation of state for air The particle velocity(or peak wind velocity) behind the shock front u is given by

(59)

where P0 is the ambient pressure (1 bar) c0 is the ambientsound speed in air (~330 m sminus1) and p is the overpressure(Glasstone and Dolan 1977 p 97) If the calculatedmaximum wind velocity is greater than 40 m sminus1 experiencefrom nuclear weapons tests suggests that ldquoabout 30 of treesare blown down the remainder have some branches andleaves blown offrdquo (Glasstone and Dolan 1977 p 225) If themaximum wind velocity is greater than 62 m sminus1 devastationis more severe ldquoUp to 90 percent of trees blown downremainder stripped of branches and leavesrdquo

The blast wave arrival time is given by

(62)

where U is the shock velocity in air given formally by

(63)

For convenience however we assume that the shockwave travels at the ambient sound speed in air c0 In this casethe air blast arrival time at our specified distance r is simply

(64)

This simplification results in large errors only very closeto the crater rim

The air blast model we use extrapolates from datarecorded after a very small explosion (in impact crateringterms) in which the atmosphere may be treated as being ofuniform density Furthermore at this scale of explosion thepeak overpressure decays to zero at distances so small (lt1km) that the curvature of the Earth may be ignored Neither ofthese assumptions applies to larger impacts thus thereliability of our predictions decreases as impact energyincreases In the future we hope to examine the effect of avariable-density atmosphere and a curved Earth on the blastwave decay using numerical modeling Such sophisticatedcalculations of the interaction between a hot ejecta plume anda realistic atmosphere by Zahnle (1990) and Toon et al(1994) which included blast wave formation are in goodagreement with our simple model in the 1ndash10000 Mt rangefor impact energies greater than this Equation 44 probablyoverestimates the blast wave effects by a factor of 2ndash5

EFFECT OF A WATER LAYER

The rationale discussed above for predicting theenvironmental consequences of an asteroid collision withEarth assumes that the impact occurs on land In fact marineimpacts are more than twice as likely to occur as land impactson Earth The influence of a water layer on the impact processhas been the subject of many recent field studies (Tsikalas et

Table 4 Air blast damageaDistance from a 1 kt explosion(d1 in m)

Over pressure (p in Pa) Description of air blast-induced damage

126 426000 Cars and trucks will be largely displaced and grossly distorted and will require rebuilding before use

133 379000 Highway girder bridges will collapse149 297000 Cars and trucks will be overturned and displaced requiring major

repairs 155 273000 Multistory steel-framed office-type buildings will suffer extreme

frame distortion incipient collapse 229 121000 Highway truss bridges will collapse 251 100000 Highway truss bridges will suffer substantial distortion of

bracing 389 42600 Multistory wall-bearing buildings will collapse411 38500 Multistory wall-bearing buildings will experience severe

cracking and interior partitions will be blown down502 26800 Wood frame buildings will almost completely collapse 549 22900 Interior partitions of wood frame buildings will be blown down

Roof will be severely damaged1160 6900 Glass windows shatter

aData extracted from Glasstone and Dolan (1977)

u 5p7P0---------

c0

1 6p+ 7P0frasl( )05---------------------------------------=

Tbdr

U r( )-----------

0

r

int=

U r( ) c0 1 6p r( )7P0

-------------+⎝ ⎠⎛ ⎞ 05

=

Tbr

c0-----=

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 16: Documentation

832 G S Collins et al

al 1998 1999 Ormouml and Lindstroumlm 2000) laboratoryexperiments (McKinnon and Goetz 1981 Gault and Sonnett1982) and numerical simulations (OrsquoKeefe and Ahrens1982a Roddy et al 1987 Ormouml and Miyamoto 2002Shuvalov et al 2002 Artemieva and Shuvalov 2002Wuumlnnemann and Lange 2002) which have led to aqualitative paradigm for submarine cratering in both the deepocean (Wuumlnnemann and Lange 2002) and shallow seas(Oberbeck et al 1993 Poag et al 2004) However like manyother aspects of impact cratering an accurate quantitativetreatment of the effect of a water layer on the crateringprocess requires complicated numerical methods beyond thescope of our program Consequently our program employsonly a rudimentary algorithm for estimating the effect of awater column on the environmental consequences of animpact We estimate the change in velocity of the impactor atthe seafloor vi|seafloor from that at the surface vi|surface byintegrating the drag equation (Equation 7) over the depth ofthe water column

(65)

In this equation dw is the thickness of the water layer Lis the diameter of the impactor after the atmospheric traverseand CD is the drag coefficient for a rigid sphere of water in thesupersonic regime which we set equal to 0877 (Landau andLifshitz 1959) This simple expression ignores both theflattening of the impactor during penetration and thepropagation of the shock wave through the water columnhowever it agrees quite favorably with numerical simulationsof deep sea impact events (Wuumlnnemann and Lange 2002)

For marine impact scenarios we calculate theapproximate kinetic energy of the impactor at the moment itstrikes the surface of the water layer Esurface and when it reachesthe seafloor Eseafloor Using Equation 16 we compute andreport two transient crater diameters one in the water layer andone in the seafloor For the transient crater diameter in thewater layer we use the impact velocity at the surface (vi =vi|surface) replace the constant 1161 with 1365 and use a targetdensity equal to the density of water (ρt = ρw = 1000 kg mminus3)For the transient crater diameter in the seafloor we assume thatthe impact velocity is that of the impactor at the seafloor (vi =vi|seafloor) and use a target density of ρt = 2700 kg mminus3

From this point the program continues as beforecalculating the dimensions of the crater in the seafloorwhether it is simple or complex the volume of the targetbelow the seafloor that is melted etc The air blast andthermal radiation calculations proceed assuming that theimpact energy is that released at the surface of the water layer(E = Esurface) the seismic shaking and ejecta calculations onthe other hand assume that the impact energy is the kineticenergy of the impactor at the moment it reaches the sea floor(E = Eseafloor) As a result our program predicts that the

thermal radiation and air blast effects are unchanged by thepresence of the water column relative to a land impact of thesame energy However a deep enough water layer couldentirely suppress the seismic shaking and excavation of rockyejecta that would occur in an impact of the same size on dryland

The current version of the program does not compute theeffects of impact-generated tsunamis for water impacts Thereare several reasons for this omission in spite of requests bymany users for this feature The first set of reasons ispractical A plausible tsunami computation requires not onlythe depth of the water at the impact site but also the depth ofthe ocean over the entire path from the impact to the observerThe observer must of course be on a coastline with anunobstructed great circle path to the impact site The observedtsunami height and run up depends on the local shorelineconfiguration and slope the presence or absence of offshorebars etc The sheer number of input parameters requiredwould daunt most potential users This sort of computationrequires a professional effort of the scale of Ward andAsphaug (2000 2003) it is far beyond the capability of oursimple program The other set of reasons centers around thecurrent uncertainty of the size of tsunamis generated byimpacts Following some initial spectacular estimates oftsunami heights heights that greatly exceed the depth of theocean itself (Hills et al 1994) a reaction occurred (Melosh2003) based on a newly-unclassified document (Van Dorn etal 1968) that suggests that impact-tsunami waves break onthe continental shelf and pose little threat to coastal locations(the ldquoVan Dornrdquo effect) The present situation with regard tothis hazard is thus confused and we decided against includingsuch an estimate in our code until the experts have sorted outthe actual size of the effect

GLOBAL EFFECTS

In addition to the regional environmental consequencesof the impact event we also compute some globalimplications of the collision We compare the linearmomentum of the impactor at the moment it strikes the targetsurface Mi = mivi with the linear momentum of the Earth ME= mEvE where mE is the mass of the Earth (583 times 1024 kg) andvE is the mean orbital velocity of the Earth (2978 km sminus1)Depending on the ratio MiME the program reports the likelyeffect of the impact on the orbit of the Earth Our choice oflimits on MiME and the corresponding degree to which theorbit changes is presented in Table 5 We compare the angularmomentum imparted by the impact Γi = miviREcosθ to theangular momentum of the Earth ΓE = 586 times 1033 kg m3 sminus1 ina similar manner Table 5 also presents the ranges of the ratioΓiΓE for which we assume certain qualitative changes to theEarthrsquos rotation period and the tilt of its axis as a result of theimpact Finally we compare the volume of the transient craterVtc with the volume of the Earth VE In the event that the ratio

vi seafloorvi surface

3ρwCDdw2ρiL θsin-------------------------ndash

⎩ ⎭⎨ ⎬⎧ ⎫

exp=

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 17: Documentation

Earth Impact Effects Program 833

VtcVE is greater than 05 we assume that the Earth iscompletely disrupted by the impact and forms a new asteroidbelt between Venus and Mars If VtcVE is in the range of 01ndash05 the program reports that the Earth is strongly disturbed bythe impact but loses little mass Otherwise the programreports that the Earth is not strongly disturbed by the impactand loses negligible mass

Currently we do not make any estimates regarding thepotentially global environmental consequences of largeimpact events In such catastrophes dust melt droplets andgas species generated during the impact event are ejected outof the Earthrsquos atmosphere and dispersed all over the globe(Alvarez 1980) Several potentially devastatingenvironmental consequences could result from the re-entryand prolonged settling though the atmosphere of this material(Toon et al 1982 1994 1997 Zahnle 1990 Kring 2000)Thermal radiation generated during the re-entry of high speedejecta may be strong enough to ignite wildfires over largeareas of the globe (Alvarez 1980 Melosh et al 1990 Toon etal 1994 1997) Dust loading in the atmosphere may block outlight and restrict photosynthesis for months after the impact(Toon et al 1982 1994 1997 Covey et al 1990 Zahnle1990) Furthermore the presence of carbonate or anhydriterocks in the sedimentary target sequence may add additionalenvironmental consequences due to the production ofclimatically active gas species (Lewis et al 1982 Prinn andFegley 1987 Zahnle 1990 Brett 1992 Pope et al 1997Pierazzo et al 1998 Kring 1999) These compounds mayproduce aerosols that further reduce the amount of light thatreaches the surface of the Earth condense with water to formacid rain react with and deplete ozone levels and causeldquogreenhouserdquo warming To make reasonable estimates of theseverity of these effects requires detailed time-consumingcomputations involving a large suite of model parameters (forexample target chemistry and mass-velocity distributions forthe ejected material Toon et al 1997) Such calculations arewell beyond the scope of our simple program we directreaders interested in these processes to the above referencesfor further information

APPLICATIONS OF THE EARTH IMPACT EFFECTS PROGRAM

We have written a computer program that estimates theenvironmental consequences of impact events both past andfuture using the analytical expressions presented above Toillustrate the utility of our program consider the hypotheticaldevastation at various locations within the United States ifasteroids of various sizes were to strike Los Angeles The firstevent worthy of consideration is the impact of a ~75-mdiameter stony asteroid (density = 2000 kg mminus3) whichoccurs somewhere on earth every 900 years on average Inthis case our program determines that the impactor wouldbegin to disrupt at an altitude of ~66 km and deposit the

majority of its kinetic energy in the atmosphere at a burstaltitude of ~5 km The air blast from this event would bestrong enough to cause substantial damage to woodenbuildings and blow down 90 of trees to a radius of ~15 kmwhich agrees well with the extent of forest damage observedafter the Tunguska airburst event in Siberia in 1908

Next let us examine the environmental consequences ofthree impact events of drastically different magnitudes at afixed distance of 200 km away from our impact site in LosAngeles which is the approximate distance from LA to SanDiego The three impacts we will consider are a 40-m diameteriron asteroid (density = 8000 kg mminus3) impacting at 20 km sminus1

into a sedimentary target (density = 2500 kg mminus3) which is theapproximate scenario of the event that formed BarringerCrater in northern Arizona a 175-km diameter stony asteroid(density = 2700 kg mminus3) impacting at 20 km sminus1 into acrystalline target (density = 2750 kg mminus3) which correspondsapproximately to the magnitude of the impact event thatformed the Ries crater in Germany and an 18-km diameterstony asteroid also impacting at 20 km sminus1 into a crystallinetarget which represents a reasonable estimate of the scale ofthe Chicxulub impact event in the Gulf of Mexico For eachimpact we assume identical impact angles (θ = 45deg) Table 6presents a comparison of the important parameters discussedin this paper for each impact event at a distance of 200 kmaway from our hypothetical impact center in Los AngelesNote the substantial variation in impact energy between eachimpact event which results in very different estimatedenvironmental effects 200 km away in San Diego The averagerecurrence interval is for the entire Earth the two largerimpact scenarios are both extremely rare events All of theseimpactors are large enough (or strong enough) to traverse theatmosphere and create a single impact crater however theBarringer-scale impactor is slowed considerably by theatmosphere

In the case of the small iron asteroid impact San Diego isa very safe place to be As little to no vapor is generatedduring this event there is no significant thermal radiationThe impact crater formed is only 12 km in diameter theatmosphere would prevent much if any ejecta thrown out of

Table 5 Global implications of an impact eventRatio Qualitative global change

MiME lt0001 No noticeable change in orbit0001 ltMiME lt001 Noticeable change in orbit001 ltMiME lt01 Substantial change in orbitMiME gt01 Totally changes orbitΓiΓE lt001 No noticeable change in rotation period

and tilt of axis001 ltΓiΓE lt01 Noticeable change in rotation period and

tilt of axis01 ltΓiΓE lt10 Substantial change in rotation period and

tilt of axisΓiΓE gt10 Totally changes rotation period and tilt of

axis

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 18: Documentation

834 G S Collins et al

the crater from reaching San Diego Furthermore the air blastwould be extremely weak at a radius of 200 km the change inatmospheric pressure would be barely discernible at a rise ofless than one part in a hundred with ensuing wind speeds ofunder a meter per second The only noticeable consequencesfrom this scale of impact would be from seismic shakingwhich would be most obvious around 40 sec after the impactoccurred The impact would be analogous to an earthquake ofRichter magnitude 49 centered in LA The ModifiedMercalli Intensity of the shaking in San Diego would be in therange of IndashII depending on the local geology meaning thatthe disturbance would be felt only in favorable circumstancesand would not cause any permanent damage

In stark contrast San Diego would not be an attractivelocation in the event that either of the two larger impactsoccurred in LA In the case of a 175-km diameter asteroidimpact the thermal exposure at a range of 200 km would besufficient to ignite most combustible materials and cause thirddegree burns to unfortunate San Diegans particularly ifvisibility was good The seismic surface waves emanating

from the impact site would arrive half a minute later andwould be violent enough to damage poorly constructedstructures topple tall chimneys factory stacks andmonuments and overturn furniture in homes and offices Arelatively thin layer of ejecta would arrive a few minutes afterthe impact and begin to rain down through the atmospherecovering the city in a few cm of ejecta fragments During thistime the air blast wave would propagate across the cityflattening any poorly constructed structure that remainedstanding and kicking up 150 ms winds capable of blowingover most trees

In the case of a Chicxulub-scale event the environmentalconsequences in San Diego would be extreme Seconds afterthe impact the fireball would engulf the city of San Diegoincinerating all combustible materials The seismic shakingthat would arrive moments later would be as violent as thatcaused by the most severe earthquake recorded on Earth Ifanything remained standing after this episode it would soonbe smothered and suffocated by the arrival of a huge amountof rock debris thrown out of the growing crater Finally a

Table 6 Comparison of environmental effects 200 km away from various impactsImpactor size (km) 004 (iron) 175 18

Percentage reduction in velocity during atmospheric entry

Equations 9 11 12 15 16 17 20

50 ndash ndash

Impact energy (J)(megatons 1 Mt = 42 times 1015 J)

Equation 1 13 times 1016

3215 times 1021

36 times 105165 times 1024

39 times 108

Recurrence interval (years whole Earth)

Equation 3 1000a 21 times 106 46 times 108

Final crater diameter (km) Equations 21 and 22 or 27

12 (Simple) 237 (Complex) 186 (Complex)

Fireball radius (km) Equation 32 ndash 23 236Time at which radiation begins (s)

Equation 33 ndash 12 ndash

Thermal exposure (MJ mminus2) Equation 34 36 37 ndash 148 ndashDuration of irradiation (s) Equation 35 ndash 300 ndashThermal radiation damage Equation 39 Table 1 No fireball created due

to low impact velocityThird degree burns many combustible materials ignited

Within the fireball radius everything incinerated

Arrival time of major seismic shaking (s)

Equation 42 40 40 40

Richter scale magnitude Equation 40 49 83 104Modified Mercalli Intensity Equation 41 Tables 2

and 3IndashII (III)b VIIndashVIII (VIII)b XndashXI (XI)b

Arrival time of bulk ejecta (s) Equations 49ndash52 Ejecta blocked by atmosphere

206 206

Average ejecta thickness (m) Equation 47 ndash 09 137Mean fragment diameter (cm) Equation 53 ndash 24 ndashArrival time of air blast (s) Equation 64 606 606 606Peak overpressure (bars) Equations 54 and 57 0004 080 77Maximum wind velocity (ms) Equation 59 096 145 2220Air blast damage Table 4 Blast pressure

insufficient to cause damage

Wooden and tall unstable buildings collapse glass windows shatter 90 trees blown down

Collapse of almost all buildings and bridges damage and overturning of vehicles 90 of trees blown down

aNote that the recurrence interval is based on impact energy alone Iron asteroids represent only ~5 of the known NEOs therefore the real recurrence intervalfor an impact of this sort is ~20 times longer

bEstimates of seismic intensity according to Toon et al (1997)

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 19: Documentation

Earth Impact Effects Program 835

strong pressure wave nearly 80 times greater than atmosphericpressure would pass through San Diego flattening anyremaining erect buildings winds over 2 km per second wouldfollow violently scattering debris and ripping up trees

The algorithm presented in this paper also allows us toextend our study of potential impact-related disasters over arange of distances away from the impact Figures 4ndash7illustrate how each of the major environmental consequencesdepends on the distance away from the impact site for thethree different scales of impact in each figure the dotted linerepresents the 40-m diameter iron asteroid impact the dashedline represents the 175-km diameter asteroid impact and thesolid line represents the 18-km diameter asteroid impact Alsomarked on the figures are the approximate locations of fourmajor US cities with respect to Los Angeles the location ofour impact site Figure 4 shows the reduction in thermalexposure with distance away from the edge of the fireballThe change in slope of the curves is caused by the curvatureof the Earth which acts to hide more and more of the fireballbelow the horizon with increasing distance away from theimpact As a result the thermal radiation damage from even aChicxulub-scale impact is restricted to a range of ~1500 kmin the event that an 18-km diameter asteroid struck LADenver would probably escape any thermal radiation damage

The horizontal positions of the grey arrows in Fig 4 denotethe radial extent of thermal radiation damage for the twolarger impacts according to Toon et al (1997) Comparingour predictions and those of Toon et al illustrates theapproximate uncertainty of both estimates Figure 5 shows theimpact ejecta thickness for each potential impact event as afunction of distance Figure 6 shows the drop in effectiveseismic magnitude with distance away from the impactwhich can be related to the intensity of shaking using Table 2The graph illustrates that impact-related seismic shakingwould be felt by all as far as Denver if a Ries-scale impactoccurred in LA and significant tremors would be felt as far-a-field as New York City following a Chicxulub-scale impactin LA The decay in peak overpressure with distance from theimpact associated with the impact air blast wave is depicted inFig 7 In the case of a 40-m diameter iron asteroid the airblast damage would be confined to a few km away from theimpact site However the blast wave from a Chicxulub-scaleimpact centered in LA may be strong enough to level steelframed buildings in San Francisco and wooden buildings asfar away as Denver For comparison the grey squares inFig 7 illustrate the approximate radial extent of airblastdamage for each impact event as predicted by Toon et al(1997) For the two larger impacts the disagreement between

Fig 4 Thermal exposure from the impact-generated fireball divided by the impact energy (in Mt) to the one-sixth power as a function ofdistance from the impact center for three hypothetical impacts in Los Angeles (Dividing fΦ by EMt

16 allows us to more easily compare theextent of thermal radiation damage for impacts of different energies Plotted in this way the scaled thermal exposure required to ignite a givenmaterial does not depend on impact energy thus values on the ordinate can be compared directly with the data in Table 1) The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid no line appearsfor the 40-m iron asteroid because little to no vapor is produced during the impact and no significant thermal radiation occurs The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities Greyarrows indicate the radial extent of fires ignited by thermal radiation from the fireball as predicted by Toon et al (1997) See the text for furtherdetails

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 20: Documentation

836 G S Collins et al

Fig 5 The effective seismic magnitude as a function of distance away from three hypothetical impacts in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impact center that correspond tothe approximate distances from LA to four major US cities See the text for further details

Fig 6 The variation in ejecta-deposit thickness with increasing distance from the impact point for three hypothetical impacts centered in LosAngeles The solid line represents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stonyasteroid the dotted line represents the impact of a 40-m diameter iron asteroid The vertical lines represent four distances from the impactcenter that correspond to the approximate distances from LA to four major US cities See the text for further details

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 21: Documentation

Earth Impact Effects Program 837

our predictions and those of Toon et al (1997) is due to ourneglect of the effects of Earth curvature and a variable densityatmosphere as discussed earlier

DISCUSSION

The Earth Impact Effects Program provides astraightforward method for estimating the regionalenvironmental consequences of the collision of extraterrestrialobjects with the Earth To implement such a program it isnecessary to make some simplifying assumptions that limit theaccuracy of any predictions Nevertheless some importantconclusions may be drawn from our simple model Of theenvironmental consequences that we consider the seismicshaking poses the most significant threat at large distancesfrom the impact site effects of ejecta fallout and the air blastdecrease much more rapidly with distance away from theimpact site Moreover the curvature of the Earth ensures thateven in the case of very rare ~20-km scale impact events thethermal radiation will be confined to a maximum range of1500 km at which point the fireball is completely hiddenbelow the horizon Closer to the impact site however the airblast thermal radiation and ejecta deposition combine toseverely affect the local environment and should all beconsidered in any hazard assessment

We believe that we have developed a valuable tool foruse both within the scientific community and the populationat large We anticipate that within the field of impact crateringour program will serve the function of providing a quickassessment of the hazard risk for potential future impactscenarios and enable those studying particular terrestrialimpact events to estimate the regional environmentalconsequences associated with the impact We welcome anysuggestions for improvements or additions to the algorithmpresented here

AcknowledgmentsndashMany members of the impact crateringcommunity and users of the Earth Impact Effects Programhave offered helpful advice for improvements to our modelIn particular we gratefully acknowledge input from BevanFrench Boris Ivanov Natasha Artemieva Ivan NemtchinovKai Wuumlnnemann Lori Styles Al Harris Alexander Reid andBlake Morlock We are indebted to the thorough andinsightful reviews of Erik Asphaug and an anonymousreviewer and the editorial handling of Elisabetta PierazzoThis work was supported by NASA grant NAG5ndash11493 Thisis IARC publication number 2005-0414

Editorial HandlingmdashDr Elisabetta Pierazzo

Fig 7 The peak airblast overpressure as a function of distance from three hypothetical impacts centered in Los Angeles The solid linerepresents an impact of an 18-km diameter stony asteroid the dashed line represents an impact of a 175-km stony asteroid the dotted linerepresents the impact of a 40-m diameter iron asteroid The dash-dotted line illustrates the decay of peak overpressure with distance away froman airburst resulting from the impact of a 75-m diameter stony object (density = 2000 kg mminus3) at 17 km sminus1 as discussed in the text The verticallines represent four distances from the impact center that correspond to the approximate distances from LA to four major US cities The greysquares show the extent of the airblast damage as predicted by Toon et al (1997) See the text for further details

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 22: Documentation

838 G S Collins et al

REFERENCES

Ahrens T J and OrsquoKeefe J D 1978 Energy and mass distributionsof impact ejecta blankets on the moon and Mercury Proceedings9th Lunar and Planetary Science Conference pp 3787ndash3802

Alvarez L W Alvarez W Asaro F and Michel H V 1980Extraterrestrial cause for the Cretaceous-Tertiary extinctionScience 2081095ndash1108

Artemieva N A and Shuvalov V V 2002 Shock metamorphism onthe ocean floor (numerical simulations) Deep Sea Research PartII Topical Studies in Oceanography 49959ndash968

Binzel R P Lupishko D F Di Martino M Whiteley R J and HahnG J 2003 Physical properties of near-Earth objects In AsteroidsIII edited by Bottke W F Cellino A Paolicchi P and Binzel PR Tucson The University of Arizona Press pp 255ndash271

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by the Earthrsquos atmosphere Nature 424288ndash291

Bottke W F Jr Nolan M C Greenberg R and Kolvoord R A1994 Collisional lifetimes and impact statistics of near-Earthasteroids In Hazards due to comets and asteroids edited byGehrels T Tucson The University of Arizona Press pp 337ndash357

Brett R 1992 The Cretaceous-Tertiary extinction A lethalmechanism involving anhydrite target rocks Geochimica etCosmochimica Acta 563603ndash3606

Bus S J and Binzel R P 2002 Phase II of the small main-beltasteroid spectroscopic survey A feature-based taxonomy Icarus158146ndash177

Chapman C R and Brandt J C 2004 Introduction to comets 2ndedition New York Cambridge University Press

Chyba C F Thomas P J and Zahnle K J 1993 The 1908 Tunguskaexplosion Atmospheric disruption of a stony asteroid Nature36140ndash44

Covey C Ghan S J Walton J J and Weissman P R 1990 Globalenvironmental effects of impact-generated aerosols Resultsfrom a general circulation model In Global catastrophes inEarth history edited by Sharpton V S and Ward P D SpecialPaper 247 Boulder Geological Society of America pp 263ndash270

Croft S K 1985 The scaling of complex craters Journal ofGeophysical Research 90C828ndashC842

Dence M R 1965 The extraterrestrial origin of Canadian cratersAnnual New York Academy of Science 123941ndash969

Dence M R Grieve R A F and Robertson P B 1977 Terrestrialimpact structures Principal characteristics and energyconsiderations In Impact and explosion cratering edited byRoddy D J Pepin R O and Merrill R B New York PergamonPress pp 247ndash275

Gault D E 1974 Impact cratering In A primer in lunar geologyedited by Greeley R and Shultz P H Moffett Field NASA AmesResearch Center pp 137ndash175

Gault D E and Sonett C P 1982 Laboratory simulation of pelagicasteroid impact Atmospheric injection benthic topography andthe surface wave radiation field In Geological implications ofimpacts of large asteroid and comets on the Earth edited bySilver L T and Schultz P H Special Paper 190 BoulderGeological Society of America pp 69ndash92

Glasstone S and Dolan P J 1977 The effects of nuclear weapons3rd edition Washington DC United States Department ofDefense and Department of Energy

Grieve R A F and Cintala M J 1992 An analysis of differentialmelt-crater scaling and implications for the terrestrial impactrecord Meteoritics 27526ndash538

Grieve R A F and Garvin J B 1984 A geometric model forexcavation and modification at terrestrial simple impact cratersJournal of Geophysical Research 8911561ndash11572

Grieve R A F Dence M R and Robertson P B 1977 Crateringprocesses As interpreted from the occurrence of impact melts InImpact and explosion cratering edited by Roddy D J Pepin RO and Merrill R B New York Pergamon Press pp 791ndash814

Grieve R A F and Therriault A M 2004 Observations at terrestrialimpact structures Their utility in constraining crater formationMeteoritics amp Planetary Science 39199ndash216

Herrick R R Sharpton V L Malin M C Lyons S N and FreelyK 1997 Morphology and morphometry of impact craters InVenus II edited by Bougher S W Hunten D M and Phillips RJ Tucson The University of Arizona Press pp 1015ndash1046

Hills J G Nemchinov I V Popov S P and Teterev A V 1994Tsunami generated by small asteroid impacts In Hazards fromcomets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 779ndash789

Hilton J L 2002 Asteroid masses and densities In Asteroids IIIedited by Bottke W F Jr Cellino A Paolicchi P and Binzel PTucson The University of Arizona Press pp 103ndash112

Holsapple K A and Schmidt R M 1982 On the scaling of craterdimensions IImdashImpact processes Journal of GeophysicalResearch 871849ndash1870

Holsapple K A 1993 The scaling of impact processes in planetarysciences Annual Review of Earth and Planetary Sciences 21333ndash373

Ivanov B A and Artemieva N A 2002 Numerical modeling of theformation of large impact craters In Catastrophic events andmass extinctions Impacts and beyond edited by Koeberl C andMacLeod K G Special Paper 356 Boulder Geological Societyof America pp 619ndash630

Ivanov B A Deniem D and Neukum G 1997 Implementation ofdynamic strength models into 2D hydrocodes Applications foratmospheric breakup and impact cratering International Journalof Impact Engineering 20411ndash430

Korycansky D G Zahnle K J and Mac Low M M 2000 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere Icarus 146387ndash403

Korycansky D G Zahnle K J and Mac Low M M 2002 High-resolution simulations of the impacts of asteroids into thevenusian atmosphere II 3D Models Icarus 1571ndash23

Korycansky D G and Zahnle K J 2003 High-resolutionsimulations of the impacts of asteroids into the venusianatmosphere III Further 3D models Icarus 161244ndash261

Korycansky D G and Zahnle K J 2004 Atmospheric impactsfragmentation and small craters on Venus Icarus 169287ndash299

Krinov E L 1966 Giant meteorites New York Pergamon Press397 p

Kring D A 1997 Air blast produced by the Meteor Crater impactevent and a reconstruction of the affected environmentMeteoritics amp Planetary Science 32517ndash530

Kring D A 1999 Ozone-depleting chlorine and bromine producedby the Chicxulub impact event Meteoritics amp Planetary Science34A67ndashA68

Kring D A 2000 Impact events and their effect on the originevolution and distribution of life GSA Today 101ndash7

Landau L D and Lifshitz E M 1959 Fluid mechanics New YorkPergamon Press 536 p

Lewis J S Watkins G H Hartman H and Prinn R G 1982Chemical consequences of major impact events on Earth InGeological implications of impacts of large asteroid and cometson the Earth edited by Silver L T and Schultz P H Special Paper190 Boulder Geological Society of America pp 215ndash221

Marsden B G and Steel D I 1994 Warning times and impactprobabilities for long-period comets In Hazards due to cometsand asteroids edited by Gehrels T Tucson The University ofArizona Press pp 221ndash239

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 23: Documentation

Earth Impact Effects Program 839

McGetchin T R Settle M and Head J W 1973 Radial thicknessvariation in impact crater ejecta Implications for lunar basindeposits Earth and Planetary Science Letters 20226ndash236

McKinnon W B and Goetz P 1981 Impact into the Earthrsquos oceanfloor during the last billion years Preliminary experimentstheoretical models and possibilities for geological detectionConference on Large Body Impacts and Terrestrial EvolutionGeological Climatological and Biological Implications pp 1ndash34

McKinnon W B and Schenk P M 1985 Ejecta blanket scaling onthe Moon and MercurymdashInferences for projectile populations(abstract) Proceedings 16th Lunar and Planetary ScienceConference pp 544ndash545

Melosh H J 1981 Atmospheric breakup of terrestrial impactors InMulti-ring basins edited by Schultz P H and Merrill R B NewYork Pergamon Press pp 29ndash35

Melosh H J 1989 Impact cratering A geologic process New YorkOxford University Press 245 p

Melosh H J 2003 Impact tsunami An over-rated hazard (abstract1338) 34th Lunar and Planetary Science Conference CD-ROM

Melosh H J Schneider N M Zahnle K J and Latham D 1990Ignition of global wildfires at the CretaceousTertiary boundaryNature 343251ndash254

Melosh H J Artemieva N A Golub A P Nemchinov I VShuvalov V V and Trubetskya I A 1993 Remote visualdetection of impacts on the lunar surface (abstract) Proceedings24th Lunar and Planetary Science Conference pp 975ndash976

Near-Earth Object Science Definition Team 2003 Study todetermine the feasibility of extending the search for near-Earthobjects to smaller limiting diameters NASA Technical Report

Nemtchinov I V Shuvalov V V Artemieva N A Ivanov B AKosarev I B and Trubetskaya I A 1998 Light flashes causedby meteoroid impacts on the lunar surface Solar SystemResearch 3299ndash114

Oberbeck V R Marshall J R and Aggarwal H 1993 Impactstillites and the breakup of Gondwanaland Journal of Geology1011ndash19

OrsquoKeefe J D and Ahrens T J 1982a The interaction of theCretaceousTertiary extinction bolide with the atmosphereocean and solid Earth In Geological implications of impacts oflarge asteroid and comets on the Earth edited by Silver L T andSchultz P H Special Paper 190 Boulder Geological Society ofAmerica pp 103ndash109

OrsquoKeefe J D and Ahrens T J 1982b Cometary and meteoriteswarm impact on planetary surfaces Journal of GeophysicalResearch 876668ndash6680

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes theseabed Geological Magazine 13767ndash80

Ormouml J and Miyamoto M 2002 Computer modeling of the waterresurge at a marine impact The Lockne crater Sweden Deep-Sea Research Part II 49983ndash994

Ortiz J L Sada P V Bellot Rubio L R Aceituno F J Aceituno JGutierrez P J and Thiele U 2000 Optical detection ofmeteoroid impacts on the Moon Nature 405921ndash923

Passey Q and Melosh H J 1980 The effects of atmospheric breakupon crater field formation Icarus 42211ndash233

Petrovic J J 2001 Mechanical properties of meteorites and theirconstituents Journal of Materials Science 361579ndash1583

Pierazzo E and Melosh H J 2000 Melt production in obliqueimpacts Icarus 145252ndash261

Pierazzo E Vickery A M and Melosh H J 1997 A re-evaluationof impact melt production Icarus 127408ndash423

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Pike R J 1980 Control of crater morphology by gravity and targettype Mars Earth Moon Proceedings 11th Lunar and PlanetaryScience Conference Geochimica et Cosmochimica Acta 32159ndash2190

Poag C W Koeberl C and Reimold W U 2004 The ChesapeakeBay CratermdashGeology and geophysics of a Late Eocenesubmarine impact structure Heidelberg Springer 522 p

Pope K O Baines K H Ocampo A C and Ivanov B A 1997Energy volatile production and climatic effects of the ChicxulubCretaceous-Tertiary impact Journal of Geophysical Research10221645ndash21654

Prinn R G and Fegley B 1987 Bolide impacts acid rain andbiosphere traumas at the Cretaceous-Tertiary boundary Earthand Planetary Science Letters 831ndash15

Richter C F 1958 Elementary seismology San Francisco W HFreeman 768 p

Roddy D J Schuster S H Rosenblatt M Grant L B Hassig P Jand Kreyenhagen K N 1987 Computer simulation of largeasteroid impacts into oceanic and continental sites-preliminaryresults on atmospheric cratering and ejecta dynamicsInternational Journal of Impact Engineering 5525ndash541

Schaller C J and Melosh H J 1998 Venusian ejecta parabolasComparing theory with observations Icarus 131123ndash137

Schmidt R M and Housen K R 1987 Some recent advances in thescaling of impact and explosion cratering International Journalof Impact Engineering 5543ndash560

Schultz P H and Gault D E 1975 Seismic effects from majorbasin formation on the Moon and Mercury The Moon 12159ndash177

Scotti J and Melosh H J 1993 Estimate of the size of cometShoemaker-Levy 9 from a tidal breakup model Nature 365733ndash735

Shoemaker E M 1962 Interpretation of lunar craters In Physics andastronomy of the Moon edited by Kopal Z New York AcademicPress pp 283ndash359

Shoemaker E M Ruth F W and Shoemaker C S 1990 Asteroidand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history edited by Sharpton V L and WardP D Special Paper 247 Boulder Geological Society of Americapp 155ndash170

Shuvalov V V Dypvik H and Tsikalas P 2002 Numericalsimulations of the Mjoslashlnir marine impact crater Journal ofGeophysical Research 107 doi1010292001JE001698

Svetsov V V Nemtchinov I V and Teterev A V 1995Disintegration of large meteoroids in the Earthrsquos atmosphereTheoretical models Icarus 116131ndash153

Toon O B Pollack J B Ackerman T P Turco R P McKay C Pand Liu M S 1982 Evolution of an impact-generated dust cloudand its effects on the atmosphere In Geological implications ofimpacts of large asteroids and comets on the Earth edited bySilver L T and Schultz P H Boulder Geological Society ofAmerica pp 187ndash200

Toon O B Zahnle K Turco R P and Covey C 1994Environmental perturbations caused by impacts In Hazards dueto comets and asteroids edited by Gehrels T Tucson TheUniversity of Arizona Press pp 791ndash826

Toon O B Zahnle K Morrison D Turco R P and Covey C 1997Environmental perturbations caused by the impacts of asteroidsand comets Reviews of Geophysics 3541ndash78

Tsikalas F Gudlaugsson S T Eldholm O and Faleide J I 1998Integrated geophysical analysis supporting the impact origin ofthe Mjoslashlnir structure Barents Sea Tectonophysics 289257ndash280

Tsikalas F Gudlaugsson S T Faleide J I and Eldholm O 1999Mjoslashlnir Structure Barents Sea A marine impact craterlaboratory In Impact cratering and planetary evolution II edited

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p

Page 24: Documentation

840 G S Collins et al

by Dressler B O and Sharpton V L Special Paper 339 BoulderGeological Society of America pp 193ndash204

Turtle E P Pierazzo E Collins G S Osinski G R Melosh H JMorgan J V and Reimold W U 2005 In Large meteoriteimpacts III edited by Kenkmann T Houmlrz F and Deutsch AGeological Society of America Special Paper 384 pp 1ndash24

Van Dorn W G LeMeacutehauteacute B and Hwang L S 1968 Handbook ofexplosion-generated water waves volume ImdashState of the artPasadena Tetra Tech

Ward S N and Asphaug E 2000 Asteroid impact tsunami Aprobabilistic hazard assessment Icarus 14564ndash78

Ward S N and Asphaug E 2003 Asteroid impact tsunami of 2880March 16 International Journal of Geophysics 153F6ndashF10

Vervack R J and Melosh H J 1992 Wind interaction with fallingejecta Origin of the parabolic features on Venus GeophysicalResearch Letters 19525ndash528

Wuumlnnemann K and Lange M A 2002 Numerical modeling ofimpact-induced modifications of the deep-sea floor Deep Sea-Research Part II 49969ndash982

Zahnle K J 1990 Atmospheric chemistry by large impacts InGlobal catastrophes in Earth history edited by Sharpton V Land Ward P D Special Paper 247 Boulder Geological Societyof America pp 271ndash288

Zelrsquodovich Ya B and Raizer Yu P 1966 Physics of shock waves andhigh-temperature hydrodynamic phenomena New YorkAcademic Press 916 p