Pent-up stress putsPent-up stress puts‘the squeeze’ on Los Angeles‘the squeeze’ on Los Angeles
Ken Hudnut
U. S. Geological Survey
Pasadena Office
USGS Public LectureCaltech
May 1, 2001
A very brief historyA very brief historyof geodesyof geodesy
Geodesy literally means measurement of the Earth (we do geodesy to study strain build-up between earthquakes)
Eratosthenes (a.k.a. ‘Beta’) measured circumference of the Earth very accurately
– Knew that sun hit bottom of well at Syene
– At same time (high noon), he measured sun’s angle at Alexandria
– 500 miles away (stadia)
so C = 25,000 miles
More recently…More recently…
Conventionalgeodesy: triangulation trilateration
Spacegeodesy:
VLBI, SLRand GPS
Elastic rebound theoryElastic rebound theoryHarry Reid – looking atgeodetic data from the1906 San Franciscoearthquake – surmisedthat strain is accumulatedand then released in anearthquake… over andover again… with similarstrain accumulation andrelease – could it predictfault behavior?
fault
fenc
e
Earthquake recurrenceEarthquake recurrence Reid’s elastic rebound
theory was too good to be true - too simple
Earthquakes do not behave this way, even in a simple system
– Brick & bungee– B & K (1967)
Time predictable, slip predictable, or non-predictable, i.e., chaotic?
We are up against friction – notoriously unpredictable
slip
time
stra
in
even
t 1
even
t 2
Slightly more complex fault model – slider blocksSimple model of an earthquake fault, the Burridge-Knopoff slider block model (1967). A scaling distribution (fractal statistics) is seen for the slider block simulations, similar to form of Gutenberg-Richter distribution of naturally occuring relationship between number of earthquakes versus magnitude
Log10 {Area}L
og
10 {
Fre
qu
ency
}
Earthquake Earthquake TerminologyTerminology
Strike-slip FaultsStrike-slip FaultsRupturesurface
Hypocenter
Hypo-center
Epicenter
Faultplane
Fault
Plate tectonic motionsPlate tectonic motions As the Farallon
plate subducted, the San Andreas fault was born
In the past 5 million years, this motion has been steady at about 5 cm/yr (that’s 5 meters per century or 50 km per million years!)
Movie by T. Atwater, UCSB
Faults & Faults & EarthquakesEarthquakes
San Andreas fault zone– North American and Pacific
plate relative motions of 56 mm/yr in a right-lateral sense
Eastern Californiashear zone
– Accomodation of right-lateral motion inboard of Sierra Nevada block
– Estimated rates of some 8-12 mm/yr (geological & space geodetic)
– Easier to go through than the Big Bend?
Big bend of San Andreas– Compression across
Los Angeles basin Ventura basin
All of Earth’s tectonic plates are constantly moving, with respect to boththe geocenter, and all of the other plates. Nothing is staying fixed, as thepieces of crust move across the surface of our sphere. Space geodesyallows us to measure absolute plate motions. We can readily estimateand then remove poles and rotation rates from our data, reducing thevelocity field so that it is referred, say, to the ‘stable’ North American plate…
… remember thatnot too long ago,baseline lengthswere re-observedto obtain precisestrain data. Spacegeodesy enabledprecise referenceframe realization,helping us to puttogether the bigpicture, allowingrigorous tests ofplate tectonics andearthquake disloc-ation theory …
Southern California Integrated GPS NetworkSouthern California Integrated GPS Network
Studying earthquakes using GPS technology
State-of-the-art network for research and earthquake response
– Software and hardware development
– New networking technology– Automated processing
systems developed Data also used widely for
surveying, GIS mapping, engineering
A few acronymsA few acronyms
Global Positioning System (GPS)
Southern California Integrated GPS Network (SCIGN)
Plate Boundary Observatory (PBO)
Southern California Earthquake Center (SCEC)
GPS – Global Positioning SystemGPS – Global Positioning System
GPS is a U.S.-built constellation of navigation satellites
Normally it is used for ‘coarse’ positioning
– Handheld GPS units (~$100-$500) C/A code only 6 meter precision (with SA turned off)
We ‘earthquake people’ do precise GPS– Top-notch GPS receivers ($11,000)
P-code and phase on both L1 & L2 Differential phase - several millimeter
precision (1000x better!) Wide range of uses for GPS data
provided by SCIGN
GPS signal GPS signal correlationcorrelation
GPS phase GPS phase differencingdifferencing
‘‘normal’ vs. ‘precise’ GPSnormal’ vs. ‘precise’ GPS
5/2/2000ended SA
normal GPSimproved byabout 10x
Continuous GPS & Continuous GPS & strainmetersstrainmeters
Best tools ever devised for highly accurate, automated, constant monitoring of crustal strain for
– long baselines– absolute ref. frame– displacement field– very high precision
SCIGN & other PBO elements require sub-millimeter velocities on the plate boundary scale in order to answer the scientific questions
regional active faultsregional active faultsYou neverknow whatyou’re goingto get…
Fort Tejon1857
Long Beach1933
Arvin-Tehachapi1952
San Fernando1971
ECSZ sequence1992-1999
Northridge1994
CaliforniaCaliforniarelative plate relative plate motions formotions forthe past 20 the past 20
million yearsmillion years
Movie by T. Atwater, UCSB
Using GPS to study earthquakesUsing GPS to study earthquakes
Measure station positions every day to within a few millimeters
Track stations through time (time series) Velocities derived from stations’ time series Strain derived from the velocity field Strain is proportional to seismic hazard and to
seismic risk
Geodesy for risk estimationGeodesy for risk estimation
Getting from GPS data to Getting from GPS data to strain, and from strain to riskstrain, and from strain to risk
Daily positions showStation trajectory vs. time(a time series of position)
In North-South, East-Westand Up-Down components,we fit a line to the data
This gives a vector for eachstation’s velocity
time 1
time 2
t1t2
North by 40 mm in 2 yrs.
West by 70 mm in 2 yrs.
From station velocities to strainFrom station velocities to strain
Velocities of all stations then are used to calculate strain (between each set of three sites)
First case – all of the velocities are the same, so there is no strain occurring in the triangle between these three sites for this time interval
From station velocities to strainFrom station velocities to strain
Velocities of all stations then are used to calculate strain (between each set of three sites)
Second case – not all of the velocities are the same, so there is strain occurring in the triangle between these three sites for this time interval
t1
t2
SCEC Crustal Motion MapSCEC Crustal Motion Map
Combined EDM,
VLBI, survey-
mode and
continuous GPS
rigorously Released as a
SCEC product Set the bar very
high for the
SCIGN project
Types of Types of StrainStrain
Shear strain– San Andreas fault has right-
lateral slip (that results from shear strain between North American and Pacific Plates)
Plane strain– Compression across the
system of thrust faults beneath Los Angeles
Dilatational strain
Rotational strain
t1 t2
From station velocities From station velocities to strainto strain
Velocities of all GPS stations are then used to calculate strain (between each set of three sites)
Maps can then be made to show each component of strain, (e.g., the shear strain both parallel to and perpendicular to the San Andreas fault)
Strain values can then be contoured, or used in hazard calculations, to make maps that are more generally used, e.g., by engineers
Probabilistic Seismic Hazard AnalysisProbabilistic Seismic Hazard Analysis
SCECled usinggeodeticdata asinput totheirseismichazardAnalysis(BSSA, 1995)
How to improve hazard informationHow to improve hazard informationand reduce earthquake risk?and reduce earthquake risk?
[SoCal is home to 15 million people, and poses half of [SoCal is home to 15 million people, and poses half of the national earthquake risk (FEMA #366)]the national earthquake risk (FEMA #366)]
Better geodetic and other data as input to geodynamic earthquake source models
– SCIGN: state-of-the-art continuous GPS net– LARSE: deep crustal imaging of faults– TriNet: recording earthquake shaking
Better methods of simulating shaking and estimating damage likely to be caused in scenario events
– USGS & CDMG – official maps to have future revisions– SCEC ‘IT’, RELM – developing & testing new methods– HAZUS – convert hazard to risk – how to reduce it?
Operational Groups:
Major Funding:
Total $18 M
SOPAC
SCIGN is an integral part
of SCEC
The major objectives of theThe major objectives of theSCIGN array are:SCIGN array are:
* To provide regional coverage for estimating earthquake potential throughout Southern California
To identify active blind thrust faults and test models of compressional tectonics in the Los Angeles region
To measure local variations in strain rate that might reveal the mechanical properties of earthquake faults
In the event of an earthquake, to measure permanent crustal deformation not detectable by seismographs, as well as the response of major faults to the regional change in strain
Proper tools for the job…
- air rotary rock drill & auger rig- 185 installations by contractor
Had to build deeply anchored tripod
‘monuments’
MonumentationMonumentation
SCIGN – a SCIGN – a greatgreat GPS network: GPS network: carefully planned - well reasonedcarefully planned - well reasoned
Monuments– Each of 5 legs is drilled to 10
meters Lowermost 6 meters is anchored Upper 4 meters is isolated from soil
– Stainless for longevity Innovative geodetic tools
– SCIGN radomes & adaptors– Data acquisition software
Redundant precise processing– GIPSY and GAMIT– Rigorous comparisons ongoing
Accuracies are the highest ever achieved, exceeding even highly optimistic expectations for SCIGN
movie by John Galetzka, USGS
SCIGN project installation statusSCIGN project installation status
Los Angeles deformationLos Angeles deformation
Two main models advanced for the observed 5 mm/yr contractional strain across Los Angeles:
1. Strain on NW-SE and SW-NE strike-slip faults (e.g., Walls et al.)
2. Strain on thrust fault system [including blind thrust faults] (e.g., Argus et al.)
Deep crustal studies - LARSEDeep crustal studies - LARSE
Fuis et al. (SCEC & USGS) set off many explosions to reflect energy from deep faults
Tomographic imaging (like a CAT scan)
Shows deep geometry of thrust faults beneath Los Angeles
Images the faults, but doesn’t show if they are active or not
Geodesy can now be used to estimate rate of slip on these faults
Deformation across LADeformation across LA
SCEC and SCIGN GPS velocities - reduced so that San Gabriel Mtns. are held fixed (Argus et al., JPL)
Contraction may be in a narrow belt, concent-rated across the San Gabriel Valley
Simple elastic models don’t fit the GPS data?Simple elastic models don’t fit the GPS data?
So far, Argus et al. are finding that it’s going to take some fancy modeling to explain these GPS data
Bawden et al. find land subsidence effects on data, and after correcting for that, they can fit the data
Meanwhile, SCIGN is providing better data to help resolve these differences
and in the future…and in the future…
Soon we will build the SCIGN project’s 250th station
Early data have already been used for– Controversy over tectonic style in Los Angeles– SCEC v2 crustal motion map and Phase II hazards– Observation of static displacement field propagation and pushing the
real-time GPS envelope (broad-band seismology using GPS?)
Data so far are exceeding expectations, so we now expect to surpass original observational objectives (except where signals are masked by land subsidence)
Leading the world in continuous GPS network technology and software development
SCIGN project progress…SCIGN project progress…
GPS & telemetry/networkingGPS & telemetry/networking
Market for GPS boards is driven by Moore’s law (like PC’s) toward faster/better/cheaper, miniaturization, etc.
Spread spectrum radio and satellite telemetry leading to high bandwidth IP field networking (e.g., TDMA)
Allows higher sampling rates and more affordable real-time telemetry
Assess damage to infrastructureAssess damage to infrastructure
Were tilts or strains large enough to damage systems? (from regional measurements)
Did damage occur to critical structures or systems? (from site-specific monitoring)
Structure monitoringStructure monitoring Pacoima dam
GPS monitoring since Sept. 1995 with LA County
GPS data can indicate damage to engineered structures such as overpasses and tall buildings
New methods: high-resolution topographic New methods: high-resolution topographic mapping and digital photographymapping and digital photography
Laser scanning using an airborne platform requires high sampling-rate GPS data during flight to control aircraft position and attitude
SCIGN stations were operated at 1 and 2 sample per second rates via the radio network
The Plate The Plate Boundary Boundary
ObservatoryObservatory
SCIGN as a prototype Plans for 875 additional
continuous GPS stations, for earthquake research, throughout the Western U. S. A. (as well as in Canada and Mexico)
The San Andreas The San Andreas fault zone ‘focus fault zone ‘focus
array’ of PBOarray’ of PBO
Geodetic networks for earthquake research, especially for observing transitional behavior between creeping and locked fault surfaces, along-strike and at depth, and aseismic fault loading processes (that cannot be observed with seismological instruments)
ConclusionsConclusions SCIGN is the world-leading, state-of-the-art, continuously-
operating GPS network We have innovated & devised the best-ever geodetic methods
for earthquake research Having led in technology development, we now fully expect to
lead in research results (especially for the Los Angeles urban region) as the network is now nearly fully operational
– Original research on strain and fault motions– New and improved hazard and risk analyses
Data from SCIGN are freely available to all, and are used extensively in surveying, engineering, and for GIS mapping (spatial referencing users may become main sponsors of network operations, removing some burden from agencies like NASA, USGS & NSF)
For more information:For more information:
http://pasadena.wr.usgs.gov/scign/
http://www.scign.org/
Kenneth W. Hudnut, Ph.D.GeophysicistUSGS [email protected]
Arthur C. Clarke's 2nd Law:
"The only way of discovering the limits of the possible
is to venture a little way past them into the impossible."