Forest Service Management - FRAMES2 HP-7l Replaces TI-59for Fire Calculations in the Field Robert E. Burgan and Ronald A. Susott Behavior of the Life-ThreateningButte Fire Richard

I.i,c,,

United StatesDepartment ofAgriculture

Forest Service-

Volume 47, No, 21986

I''\

~,

"I

FireManagementNotes

......=-~\o.\.~

,:: .:po.:

" ./,

'0-..,~,; ,/ .... ' '-...../

~, ~ ;

FireManagementNotesAn international quarterly periodical devoted to

forest fire management

Contents3 Forest Fire Shelters Save Lives

Art Jukkala and Ted Putnam

6 Methods for Predicting Fire Behavior­You Do Have a Choice

Patricia L. Andrews

United StatesDepartment ofAgricultureForest Service

Volume 47, NO.21986

Fire Management Nores is published bythe Forest Service 01 the United StatesDepartment of Agriculture, Washington,D.C. The Secretary of Agriculture hasdetermined that the publication of thisperiodical is necessary in the transac­tion of the public business required bylaw of this Department. Use of funds forprinting this periodical has been eo­proved by the Director of the Office ofManagemer.\ and Budget 1t\ICUghSeptember 30,1984.

11

14

25

29

32

36

2

HP-7l Replaces TI-59 for Fire Calculationsin the Field

Robert E. Burgan and Ronald A. Susott

Behavior of the Life-Threatening Butte FireRichard C. Rothermel and Robert W. Mutch

An Application of NIIMS on the Uinta NationalForest

Helen Woods and Lyle Gomm

Current Status of BEHAVE SystemRoger L. Eubanks, Roger L. Bradshaw, andPatricia L. Andrews

Crew Mobilization: What's the Next Step?Stephen W. Creech

Late-Winter Prescribed Burns To PrepareSeedbeds for Natural Loblolly-Shortleaf PineRegeneration-Are They Prudent?

Michael D. Cain

Cover: Fire shelter with section cut away to show fully equipped firefighter prop­erly positioned. Fire shelters were heavily used during the 1985 fire season, espe­<:la!!y on the Butte Pire. See articles beginning on p. 3 ..",I 1'. 14.

Subscriptions may be obtained from theSuperintendent of Documents, U.S.Government Printing Office,Washington, D.C, 20402.

NOTE-The use of trade, firm, or cor­poration names in this publication is forthe information and convenience of thereader. Such use does not constitute anofficial endorsement of any product orservice by the U.S. Department ofAgriculture.

Disclaimer; Individual authors are responsi­ble 101 Ihe technical accuracy 01 the matelialpresented in Fire Management Notes

Send suggestions and articles to Chief,Forest Service (Attn: Fire ManagementNotes), P.O. Box 2417, U.S. Departmentof Agriculture, Washington, DC 20013.

Richard E. lyngU.S. Department of Agriculture

A. Max Peterson, ChiefForest Service

l.A. Amtcarena, DirectorFire and Aviation Management

Francis R. Russ,General Manager

Fire Management Notes

.,I

Forest Fire Shelters Save LivesArt Jukkala and Ted Putnam

Forester and equipment specialist, respectively.USDA Forest Service, Equipment Development Center,Missoula, MT

I

~

rI

The 1985 fire season proved espe­cially active for firefightersnationwide. Drought and adverseweather combined to set the stage forextreme fire behavior. Massivecrowning, spotting, and fast spreadcharacterized last year's fires, whichtrapped more than 200 firefighters.But thanks to good crew leadershipand the fire shelter, many deaths andserious injuries were prevented.

Fires trapped crews in the Eastearly in the year. Out West, the mostdangerous fires took place later in thesummer. Two serious incidentsoccurred in Idaho fires. Some 82firefighters deployed shelters whilebattling the Lake Mountain Fire onthe Payette National Forest. On theButte Fire, in the Salmon NationalForest, a fast-moving crown firechased four crews into safety zones.The firefighters set up shelters andsurvived despite intense heat andsmoke. According to one crew boss,the fire shelter made the difference:"The shelter saved our lives. We hadno escape altemative."

Reports from the Lake MountainFire tell us that most firefighters de­ployed their shelters under circum­stances that were not life threatening.Nevertheless, the shelter is creditedwith saving a few lives, and it pro­tected countless crew members fromserious burns and smoke inhalation.

Lives Saved on Butte Fire

The Butte Fire entrapments clearlywere life threatening. Eddie Abeyta,

Volume 47, Number 2

crew liaison officer, Santa Fe Na­tional Forest, who was one of thosetrapped, says, "We would've nevermade it without the shelter. There isno question about it. No shelter, nowalk out of there."

According to Nick Montoya of theCarson Hotshots, who was trapped inthe larger of the two safety zones,"Mortality might have been 75 per­cent without the shelter." Of the 73deployments in both safety zones, in­vestigators believe at least 60firefighters would have died withoutthe shelter's protection. Once againthe fire shelter had proved itself.

We estimate that the fire shelter hassaved more than 140 lives since its in­troduction in the early 1960's. Themain reason the fire shelter saveslives is because it gives firefighters away to protect face and airways.Breathing flames and hot gases is thegreatest hazard in fire entrapment;thus protecting face and airways is vi­tal. This fact cannot be stressedenough. A Federal Aviation Adminis­tration study of 1,140 bum cases in­volving 106 fatalities concluded thatif the lower respiratory tree (trachea,main bronchi, and secondary bronchi)is burned, death is almost inevitable.

We also believe the more 'you knowabout the fire shelter, the more confi­dence you'll have in it, and the betterprepared you'll be to stay put in yourshelter should you ever becometrapped. We have learned a lot fromour investigation of the Butte Fireentrapments and want firefighters toknow about the role the fire shelter

played and how they can increasetheir chance of survival. .

What the Shelter Can and Can't Do

The fire shelter protects primarilyby reflecting radiant heat. As demon­strated on the Butte Fire, even largecracks or tears do not reduce the shel­ter's protective capabilities when radi­ant heat is the principal hazard. Sev­eral people on that fire deployedshelters with 4- to 18-inch tears.These shelters let in some smoke butstill protected their occupants from ra­diant heat and heavier concentrationsof smoke.

The shelter's thin aluminum-glasscloth laminate can withstand only lim­ited contact with flames. Cracks,tears, or holes reduce its protection indirect flames. The shelter should bedeployed away from fuelconcentrations-both natural fuelsand flammable equipment.

Techniques for Survival

Avoid Entrapment-Experiencedfirefighters know the best way to en­sure their safety on a wildfire is toavoid entrapment. But when droughtconditions and severe fire weathercombine, as they did on the ButteFire, avoiding entrapment is not al­ways possible. Several people we in­terviewed said that before beingtrapped On the Butte Fire, they feltthey would never let themselves getinto such a situation and need a fireshelter.

3

We estimate that the fire shelter has savedmore than 140 lives since its introduction ....

Although entrapment can't alwaysbe avoided, you can make it unlikelyby:

• Following the 10 StandardFirefighting Orders.

• Knowing the 13 fire situationsthat shout "Watch Out!"

• Knowing the four major commondenominators of fire behavior thatlead to tragedy or near-miss fires.

Use the fire shelter as a last resort.Fallow proven escape proceduresfirst.

Select Safety Zones Carefully­The large safety zones on the ButteFire minimized direct flame contacton the shelters, allowing them to per­form effectively and reflect radiantheat as they were designed to do.

First, select safety areas carefully.Next, look for natural protectionwithin the safety zone and use it.Erect the shelter behind a large rock,dozer blade, or other heat shield.Take advantage of constructed or nat­ural depressions or small earth benns.On roads, deploy in the cut slopeditch. Such spots expose you andyour shelter to less heat, smoke, andwind.

Commit Yourself to theShelter-When you know you can'tescape entrapment, commit yourselfto the shelter. Deploy it quickly. Useany extra time to pick the best spotand prepare ihe site by scraping awayflammable fuels. Keep an arm or legthrough one of the shelter straps. Oth­erwise, you might lose the shelter inthe high winds the flame front gener­ates. A survivor of the Butte Fire-a

4

veteran of 23 fire seasons----estimatedwinds at 50 to 70 miles per hour.

The coolest, cleanest air is within afew inches of the ground. So stay lowwith your nose pressed to the earth.

Once inside the shelter, stick it outno matter how scared you are or howpainful it is. Remember, it is alwaysmuch better inside the shelter thanoutside. If you make a dash for it,your chances of survival are poor. Ifyou leave after the flame front passesbut while it's still smokey, you riskinjury-from smoke inhalation. It'sbest to stay put until a crew leadersays it's safe to leave.

Maintain Communications-Theentrapment reports we've studiedshow the psychological benefits ofreducing fear and panic by talking totrapped coworkers directly or by ra­dio. If you can't talk on the radio, lis­ten. Outside observers may be able toprovide valuable information aboutthe fire in your location. At its peak,the noise of a fire can be deafening,and you may not be able to hear any­one. Don't panic. As soon as thenoise subsides, resumecommunications.

Role of Other Equipment- Re­cent fire entrapments illustrate thevalue of flame-resistant clothing. Thisclothing protects you while you'reescaping entrapment or deploying andoccupying a shelter. Be sure to weargloves. Butte Fire experiences showthat holding the shelter down can be amajor problem in high winds. With­out gloves you may burn your handsand not be able to hold the shelter

down. We hope to improve shelterhold-down features in future designs.Wear your hardhat, equipment packs,and other gear inside the shelter tohelp keep hot surfaces away fromyour body. Be sure to leave tools,which can cut shelter cloth, outside.Any gasoline or fusees should be leftbehind or thrown far from any shelter.

In larger safety areas, you maywant to move the shelter to get awayfrom heat concentrations. This tacticwas effective on the Butte Fire. How­ever, firefighters reported that as theymoved the shelters were hard to holdonto, allowing smoke 'to get inside.Remember, if you move, there is thedanger of exposing your face to hotflames and gases.

Shelter Training-The more youknow about the fire shelter and whatto expect during entrapment, the bet­ter prepared you'll be should it hap­pen. A new film, "Your Fire Shelter;"is a good place to begin. Study thepamphlet of the same title carefully.Practice the imaging techniques, thenget hands-on training by deployingobsolete shelters. Refer to the pam­phlet for training recommendations.

Care aud lnspection- Experi­ences during the past several yearshave shown why proper shelter careand inspection are vital. Many shel­ters deployed on the Butte Fire hadcracks along the fold lines. Most shel­ters with this type of damage shouldbe screened out in a proper inspectionprogram. The pamphlet "Your FireShelter" contains details on care andinspection.

Fire Management Notes

The events of the 1985 fire seasonhave reaffirmed the worth of the fireshelter and the wisdom of having ev­ery firefighter canry one on the line.Entrapment experiences made believ­ers out of many firefighters duringthis past fire season. Learn from theirexperience-find out as much as youcan about how to use, care for, andinspect the fire shelter. This know­how could save your life someday.•

Shelter Use Observations

I. Fire shelters work, even whenthey are not in the best condition.

: Some shelters used in the Butte Fire.had 4- to 18-inch tears along folds.· 2. In indirect attack situations,· safety zones should be constructed to· provide effective backup if alternativeescape routes are cut off or earlyevacuation is not possible.

3. Safety zones on ridge topsshould be at least 300 feet in diameterin timber with fuel model G or 10.They may have to be larger in otherlocations.

4. The value of competent, well­trained, and experienced crew bosses,strike team leaders, and division su­pervisors cannot be overemphasized.In the Butte Fire incident, many liveswere saved through their actions.

Volume 47, Number 2

5. Overhead should recognize thatfirefighters using shelters may not beable to use their radios if turbulentconditions make it difficult to hold theshelter in place. On the other hand,one-way communications should con­tinue to give instructions and reassur­ance. It will be a challenge to over­head to effectively communicatereassurances to sheltered firefighterswhile still transmitting key informa­tion regarding the nature of a majorincident.

6. When in shelters, firefightersshould continue to talk to one anotherto maintain contact and reduce thechance of panic.

7. Once the fire has passed over,firefighters should stay in their shel­ters until the smoke has cleared.

8. Sheltered firefighters should notwet down their skin or clothing or wethandkerchiefs for breathing. Moistheat causes more damage to lung tis­sues than dry heat.

9. Sheltered firefighters should sipwater to prevent dehydration.

10. Incident commanders, opera­tion section chiefs, and emergencymedical technicians should follow upafter an incident to ensure that thoseinvolved in the shelter deployment arein the proper physical and mental con-

dition to continue in their fire assign­ments. A delayed stress response fol­lowing a traumatic incident couldseriously impair the safety and pro­ductivity of fireline personnel.

II. The life-saving value of shel­ters should be ensured through propercare and handling by firefighters.Throwing shelters around, sitting onthem, or other rough treatment willaccelerate the development of tearsand holes.

12. Missoula Equipment Develop­ment Center's field trial publication"Your Fire Shelter" (August 1984)contains the most up-to-date informa­tion on fire shelter use and inspection.All firefighting personnel should care­fully review this publication. Thepublication includes information onentrapment and on deployment, in­spection, and care and handling of the:shelter.

13. Measures need to be taken toensure that all firefighters know how ;to deploy and use the fire shelter.Contract sawyers, dozer operators,National Guard truck drivers, andother involved persons should be in­structed as well.

Richard C. Rothermel andRobert W. Mutch

5

Methods for Predicting FireBehavior-You Do Have a ChoicePatricia L. Andrews

Mathematician, Fire Behavior Project,USDA Forest Service, Intermountain Research Station,Intermountain Fire Sciences Laboratory, Missoula, MT

Figure I-Fire behavior predictions can be made using manual methods (such as nomograms or ta­bles), the TI-59 and HP-7JB calculators, or the BEHAVEfire behavior prediction andfuel modelingsystem.

Predictions of wildland fire behav­ior are used in various aspects of firemanagement: prescribed fireplanning, presuppression planning,real-time fire suppression activities.Methods for calculating fire behaviorcovered here represent continued im­provement of the packaging of mathe­matical prediction models for use byfire managers. Such improvement re­sulted from expanding user needs, ad­ditional research results, and newtechnology. Options available to man­agers range from manual methods(such as tables and nomograms), tohandheld calculators, to computers.These methods mainly differ in pre­diction capabilities and ease of use. r! .is important to understand that al·though the methods may differ, allproduce valid results.

In this article I will discuss themanual methods described byRothermel in "How To Predict theSpread and Intensity of Forest andRange Fires" (/5), the TI-59 calcula­tor with a CROM (Custom Read OnlyMemory) (8), the HP-71B calculatorwith a CROM (/8), and the BEHAVEfire behavior prediction and fuel mod­eling system (6, 7) (fig. I).

Computer programs not nationallyavailable to all agencies and privatefirms, such as FIREMOD (1) andFIRECAST (/2), will not be dis­cussed here. In addition, I have in­cluded only those methods that pre-diet site-specific fire behavior. Thisdiscussion, therefore, does not in­clude systems designed for other pur­poses, the National Fire Danger Rat-

6

ing System, for example. I will notcover the role and importance of ex­perience, except to emphasize that itis vital to any method of predictingfire behavior.

Manual Methods

Manual methods for calculating firebehavior include tables, graphs, andnomograms. Albini's nomograms forspread rate and intensity (2) were thefirst step in providing prediction mod­els to the field. As Dick Rothermelstated in the preface to "How To Pre­dict the Spread and Intensity of Forestand Range Fires," Frank Albini "letthe genie out of the bottle with publi-

cation of his book of nomograms in1976." Although that was 10 yearsago, the nomograms remain useful inthis age of computers. Nomogramsgraphically depict potential fire be­havior, showing relationships thatcannot as easily be seen in tables.Nomograms allow quick estimation ofspread rate, flame length, and inten­sity, based on a minimum ofinformation.

Rothermel (15) describes thenomograms and other manual meth­ods for calculating fire behavior thathave been developed through theS-590 Fire Behavior Analyst (FBA)course. Even with the availability ofcalculators and BEHAVE, S-590 con-

Fire Management Notes

tinues to include the manual methods.An FBA should be proficient in allmethods of calculating fire behavior(manual, calculator, BEHAVE, and,of course, experience), so as to beprepared to cope withcontingenciessuch as battery failure or lack of elec­tric power.

The S-390 Intermediate Fire Be­havior course (13) also covers manualmethods. So many students takeS-390 that it would be impractical torequire all of them to use a computeror calculator. In addition, the S-390Field Reference can be readily used inthe field. The reference has beennicknamed the "two-bit TI," meaningthat it can do what the TI does, at lesscost.

TJ-59 Calculator

Developing a spread and intensityCRaM for the TI-59 handheld calcu­lator gave users a quick, easy, andhandy means for calculating fire be­havior predictions in the field as wellas in the office. The automation was amajor step beyond manual methods.

Additional fire behavior predictionprograms for the TI-59 are availableon cards (4, 11). This calculator canprocess calculations too complex formanual methods. For example, thenomograms for maximum spottingdistance are limited to spotting from asingle torching tree on flat ground.The TI program allows for mountain­ous terrain and for spotting from agroup of torching trees, burning piles,and wind-driven surface fires.

Volume 47, Number 2

BEHAVE System

The next improvement after the de­velopment of the calculator was theBEHAVE fire behavior prediction andfuel modeling system (6, 9, 16). BE­HAVE is currently being expanded toallow additional predictioncapabilities (7). Anyone who hasprogressed from nomograms to the TIand then to BEHAVE can attest to theextent of the advancement. BEHAVEis not only the most comprehensive ofthe methods for calculating fire be­havior, but it is also the easiest to use.

Many of the prediction models inBEHAVE were already available inthe form of manual methods or TI-59programs. BEHAVE also includesmodels not previously available. Oneof the major features of BEHAVE isthe capability to design custom fuelmodels.

BEHAVE gathers the predictionmodels into one easy-to-use package.Tables of predictions can be generatedquickly. For example, in a few min­utes one can tabulate the effect of var­ious windspeeds on rate of spread,whereas it takes days to build tablesusing the TI-59. In the office, BE­HAVE is the logical choice for firebehavior calculations.

However, despite improved capa­bility to access computers from re­mote sites, handheld calculators arestill needed for predicting fire behav­ior in the field.

HP-71B Calculator

The HP-7IB calculator is replacing

the TI-59 calculator for fire behaviorcalculations (10, 18). The TI-59's are'breaking down and are no longermanufactured. Handheld calculatortechnology has advanced significantlysince the adoption of the TI-59 and itsCRaM. So the HP is much more thana replacement. Its capabilities go farbeyond those of the TI and are almostthe same as BEHAVE.

The HP fire behavior program ispatterned after the BURN subsystemof BEHAVE (the FIRE 1 and FIRE 2programs). The design, keywords,and worksheets are similar insofar asis practical. I anticipate that peoplewill frequently switch between BE­HAVE and the HP. For example, infire camp an FBA may have access toBEHAVE, but on the fireline will usethe HP.

Calculatiou Comparison

Table 1 shows aspects of fire be­havior that can be calculated, and al­ternative methods for doing so. Forexample, forward rate of spread, ifupslope with the wind, can be calcu­lated using tables, nomograms, BE­HAVE, and the TI-59 and HP-71Bcalculators. Only BEHAVE and theHP-71B, however, can calculate rateof spread for any specified direction.Containment by indirect attack can becalculated only by BEHAVE; this isthe only calculation that can be doneby BEHAVE and not by the HP-71B.The table also indicates that BE­HAVE provides the only means fordesigning custom fuel models; the

7

Table I-Major elements of fire behavior that can be predicted and various methods of calculation [Numbers in parentheses refer to publications in·theliterature cited section.]

Fire behavior element Manual methods TI-59 HP-?1B BEHAVE

Rate of spread; flame length; fireline intensity:Upslope with the wind Tables (13), nomograms (15) CRaM (8) CRaM (18) (6)

In the direction of maximum spread vectoring (15) CRaM (18) (6)

In any specified direction CRaM (18) (6)

Heat per unit area Nomograms (15) CRaM (8) CRaM (18) (6)

Reaction intensity Nomograms (2) CRaM (8) CRaM (18) (6)

Area; perimeter:With upslope wind Tables (13, 15) CRaM (8) CRaM (18) (6)With cross-slope wind CRaM (18) (6)

Length-to-width ratio Diagrams (13, 15) CRaM (18) (6)

Forward spread distance MUltiplication (15) CRaM (8) CRaM (18) (6)

Backing spread distance:maximum width of fire CRaM (18) (6)

Maximum spotting distance:From torching trees Nomograms (15, 3) Card (11) CRaM (18) (6)From burning piles Card (11) CRaM (18) (6)From wind-driven surface fires Card (11) CRaM (18) (7)

Containment (final fire size,line building rate, containment time):

Direct attack Card (4) CRaM (18) (6)Indirect attact (7) ,

Scorch height Graph (2) CRaM (18) (7)

Probability of ignition Tables (13, 15) CRaM (18) (7)

Ignition component CRaM (8)

Fine dead fuel moisture Tables (13, 15) CRaM (8) CRaM (18) (6, 7)

Custom fuel modelsDevelop (9)Use Card (9) CRaM (18) (6)

fuel models can then be used on the ior are based on Rothermel's spread nomogram, table, calculator, orTI ami HP. model (14). Therefore, given the computer.

Some factors related to individual same input, the predicted rate of Table I indicates that fine dead fuelpredictions deserve further discussion. spread will be the same whether the moisture can be calculated using ta-All methods for predicting fire behav- calculations are done using a bles, the TI-59 CROM, the HP-7IB

8 Fire Management Notes

CROM, and BEHAVE. Nevertheless,there are major differences in themethods. The TI-59 estimates finedead fuel moisture based on tempera­ture, relative humidity, and shade. Itshould be used only as a last resort.The S-590 tables allow adjustment forother factors: aspect, slope, positionon the slope, and time of day. TheS-390 tahles are a modification of,and produce results similar to, theS-590 tables. The prediction modelimplemented on the HP-71B and inBEHAVE is a highly sophisticatedsite-specific model (/7). The differentinput required for each of the modelsshould tell the user that the modelsare indeed different.

The models used to predict factorsother than fuel moisture are not dra­matically different. Answers may beslightly different, but not significantlyso, when one considers the applica­tion and the resultant decisions. Inmost cases input and output are thesame. The differences lie in the inter­nal workings of the mathematical

. model.The vectoring method for pre­

dicting spread under cross-slope windconditions includes some simplifyingassumptions that permit the use ofmanual methods. More sophisticatedcalculations are done in BEHAVEand on the HP.

The area and perimeter calculationsfor the tables and the TI are based ona double-ellipse formula (5), whereasBEHAVE and the HP use a simple

Volume 47, Number 2

elipse. The results are slightly differ­ent. This modification made it possi­ble to link size calculations to con­tainment calculations and to predictfire behavior in a cross-slope wind.

The containment calculations forthe TI, HP, and BEHAVE are allslightly different. The TI model hadlimitations and discontinuities thatwere overcome for the BEHAVE ver­sion. The HP requires a tabular ver­sion of the model in BEHAVE be­cause of the number of calculationsinvolved.

Probability of ignition is the samefor the tables, the HP, and BEHAVE.(The S-390 table is a condensation ofthe S-590 table.) Through oversight,ignition component rather than proba­bility of ignition was put on the firebehaviot part of the TI CROM. Igni­tion component was developed for usein the National Fire Danger RatingSystem; probability of ignition is usedfor fire behavior prediction.

Summary

Methods for estimating fire behav­ior vary from manual calculations tocomputer programs. Manually calcu­lated predictions are subject to manylimitations, and one must be highlytrained to use them. Nevertheless,manual calculations will always re­main useful, especially for a fire be­havior analyst on a wildfire suppres­sion overhead team. Those who needfire behavior predictions at a specified

time will not accept the excuse ofequipment failure. And when tailgatepredictions of fire behavior are calledfor, a quick look at a nomogramshould suffice.

Even with the availability of man­ual methods and BEHAVE, there hasbeen overwhelming demand to re­place the TI-59. Because of advancesin technology, the HP-7IB CROMhas capabilities far beyond those ofthe TI-59 CROM. The HP-71B isvery similar to BEHAVE, including auser-friendly interface. However, theavailability of the HP-7IB does notmean that each TI-59 should be re­placed with an HP. Although the TI iscapable of only about 10 percent ofwhat the HP can do, if that 10 percentmeets your needs and your TI is stillworking, there is no urgent need toimmediately switch to the HP. Thepredictions from the TI are as valid asever.

BEHAVE is at the automated endof the methods scale. It has the mostcapabilities and is the most user­friendly alternative. In most casesBEHAVE is the preferred choice;however, access to a computer is notalways possible. Although predictivecapabilities increase in the progres­sion from manual methods to calcula­tors to BEHAVE, each method forcalculating fire hehavior has its ownniche in fire management activities.You are now fortunate enough to havea wide choice in the method that youuse to calculate fire behavior. •

9

Literature Cited

I. Albini, F.A. Computer-based models ofwildland fire behavior: a users' manual.Ogden, UT: U.S. Department of Agricul­ture, Forest Service, Jnrermounrain Forestand Range Experiment Station; 1976. 68 p.

2. Albini. F.A. Estimating wildfire behaviorand effects. Gen. Tech. Rep. INT-30.Ogden. UT: U.S. Department of Agricul­ture, Forest Service. Intermountain Forestand Range Experiment Station; 1976. 92 p.

3. Albini, F. A. Spot fire distance from burn­ing trees-a predictive model. Gen. Tech.Rep. tNT-56. Ogden, UT: U.S. Departmentof Agriculture. Forest Service,Intermountain Forest and Range ExperimentStation; 1979. 73 p.

4. Albini, F.A.; Chase, C.H. Fire containmentequations for pocket calculators. Res. NoteINT-268. Ogden, UT: U.S. Department ofAgriculture, Forest Service, IntermountainForest and Range Experiment Station; 1980.17 p.

5. Anderson, H.E. Predicting wind-drivenwildland fire size and shape. Res. Pap.INT-305. Ogden, UT: U.S. Department ofAgriculture, Forest Service, IntermountainForest and Range Experiment Station; 1983.26 p. _

6. Andrews, P.L. BEHAVE: Fire behaviorprediction and fuel modelingsystem-BURN subsystem, Part I. Gen.Tech. Rep. IN1'-194. Ogden, UT: U.S. De­partment of Agriculture, Forest Service,Intermountain Research Station; 1986.130 p.

10

7. Andrews, P.L.; Chase, C.H. BEHAVE:Fire behavior prediction and fuel modelingsystem-BURN subsystem, Part 2. Gen.Tech. Rep. INT-OOO. Ogden, UT: U.S. De­partment of Agriculture, Forest Service,Intermountain Research Station; (in press].

8. Burgan, R.E. Fire danger/fire behaviorcomputations with the Texas InstrumentsTJ-59 calculator: user's manuaL Gen. Tech.Rep. INT-61. Ogden, UT: U.S. Departmentof Agriculture, Forest Service, .Intermountain Forest and Range ExperimentStation; 1979. 25 p.

9. Burgan, R.E.; Rothermel, R.C. BEHAVE:Fire behavior prediction and fuel modelingsystem-FUEL subsystem. Gen. Tech.Rep. INT-167. Ogden, UT: U.S. Depart­ment of Agriculture, Forest Service,Intermountain Forest and Range ExperimentStation; 1984. 126 p.

to. Burgan, R.E.; Susott, R.A. HP-71 replacesTI-59 for fire calculations in the field. FireManagement Notes. 47(2): 9-1 I.

11. Chase, C.H. Spotting distance from wind­driven surface fires-extensions of equa­tions for pocket calculators. Res. NoteINT-346. Ogden, UT: U.S. Department ofAgriculture, Forest Service, IntermountainForest and Range Experiment Station;1984. 21 p.

12. Cohen, J. FIRECAST user's manual. Riv­erside, CA: U.S. Department of Agricul­ture, Forest Service, Pacific Southwest For­est and Range Experiment Station, ForestFire Laboratory; [in pressj.

13. National Wildfire Coordinating Group. Firebehavior self-study course. 5-390. Boise,10: Boise Interagency Fire Center; 1981.

14. Rothermel, R.C. A mathematical model forfire spread predictions in wildland fuels.Res. Pap. INT-l 15. Ogden, UT: U.S. De­partment of Agriculture, Forest Service,Intermountain Forest and Range Experi­ment Station; 1972. 40 p.

15. Rothermel, R.C. How to predict the spreadand intensity of forest and range fires. Gen.Tech. Rep. INT-143. Ogden, UT: U.S. De­partment of Agriculture, Forest Service,Intermountain Forest and Range Experi­ment Station; 1983. 161 p.

16. Rothermel, R.C. BEHAVE and YOU canpredict fire behavior. Fire ManagementNotes. 44(4): 11-15; 1983.

17. Rothermel, R.C.; Wilson, R.A., Jr.;Morris, G.A.; Sackett, 5.5. Modelingmoisture content of fine dead wildlandfuels: input to the BEHAVE fire predictionsystem. Res. Pap. INT-359. Ogden, UT:U.S. Department of Agriculture, ForestService, Intermountain Research Station;1986. 63 p.

18. Susott, R.A.; Burgan, R.E. Fire behaviorcomputations with the Hewlett-PackardHP-7IB calculator. Ogden, UT: U.S. De­partment of Agriculture, Forest Service,Intermountain Research Station; [in press].

Fire Management Notes

HP-71 Replaces 11-59 for FireCalculations in the FieldRobert E. Burgan and Ronald A. Susott

Research forester and research chemist, respectively,USDA Forest Service, Intermountain Research Station.Fire Sciences Laboratory, Missoula, MT

The HP-7JB has several features that make it an effective field calculator.

If your Texas Instruments TI-59 isnearing its last gasp, you can replaceit with a newer calculator and enjoythe use of improved fire danger andfire behavior programs. The Hewlett­Packard HP-71B handheld calculatorhas been selected to replace the TI-59and is now available on a USDA For­est Service contract.

The HP-71 B has several featuresthat make it attractive as a fieldcalculator:

• An alphanumeric display thateliminates the need for keyboardoverlays.

• A liquid crystal display (LCD)that becomes easier to read in day­light, rather than more difficult.

• Use of complementary metal ox­ide semiconductor (CMOS) technol­ogy, which has a very low power re­quirement, thus pennitting 2 to 3months of normal use between batterychanges.

• Field-replaceable AAA batteries,rather than rechargeable nickel­cadmium batteries.

• A continuous memory that re­tains the information stored in the cal­culator even when it is turned off.

• A computational speed about sixtimes faster than the TI-59.

• A capability to be used with op­tional battery-operated printers, datacassettes, and disk drives.

• A powerful BASIC programminglanguage that is available for manyother applications.

• Output that can be routed to abattery-operated, field-portableHewlett-Packard inkjet printer.

Volume 47, Number 2

The National Fire Danger RatingSystem (NFDRS) program and thefire behavior program have been puton separate Custom Read Only Mem­ories (CROM's) for use in theHP-71B. Separate user's manualshave also been prepared for each pro­gram. The manual for calculatingNFDRS indexes and components is"Fire Danger Computations with theHewlett-Packard HP-71B Calcula­tor:' The fire behavior user's manualis "Fire Behavior Computations withthe Hewlett-Packard HP-7IB Calcula­tor." Both manuals are soon to bepublished by the Intermountain Re­search Station (2, 4). Separate self­study guides have been prepared forthe fire danger and fire behavior pro-

grams and are available through theagency coordinator.

NFDRS Program

The inputs required to perform1978 NFDRS (3) computations arethe same as for the TI~59 and othersystems. Weather inputs may be re­corded on the Weather Service's"1O-Day Fire Danger and FireWeather Record" fonn D-9b or onthe form provided in the user'smanual.

The major attributes of the NFDRSprogram are:

11

• Computes NFDRS indexes andcomponents from weather inputswhen the program's WEATHERmodule is selected.

• Automatically updates and storesthe values of those inputs that must becarried forward from day to day.When the WEATHER module isused, these values do not need to bemanually reentered each day.

• The NFDRS fuel models arestored in the calculator, not on mag­netic cards.

• Up to five supplemental "user de­fined" NFDRS fuel models may alsobe permanently stored in the calcula­tor memory, although no method cur­rently exists for building and testingsuch models.

Fire Behavior Program

The fire behavior program, whichis patterned after the BURNsubsystem of BEHAVE (1), imple­ments much more fire behavior tech­nology than was possible with theTI-59. Program capabilities are indi­cated by the following list of programmodules and their functions:

• FUEL MODEL-permitsinputting, loading, listing, saving anddeleting models, and listing names ofmodels stored in the calculator.

• DIRECT-calculates spread rate,heat per unit area, fireline intensity,flame length, reaction intensity, effec­tive windspeed, and direction of max­imum spread.

• SIZE-calculates area, perime­ter, length-to-width ratio, forward

12

spread distance, backing spread dis­tance, and maximum fire width.

• CONTAIN-calculates length offireline at containment time, time tocontainment, and final fire size or linebuilding rate required to stop the fireat a specified size.

• SPOT-calculates maximumspotting distance.

• SCORCH-calculates scorchheight.

• IGNITE-calculates probabilityof spot fire ignition.

• MOISTURE-calculates l-hourtimelag fuel moisture, fuel level tem­perature and relative humidity, per­centage of area shaded, and probabil­ity of ignition for a specific bum timeor on an hourly basis. .

• MAP-calculates fire dimensionsor spotting distance for plotting on amap.

• SLOPE-calculates slope steep­ness, elevation change, and horizontaldistance between two points.

• WIND--<:alculates midflamewindspeed from 20-foot windspeed.

• RH-calculates relative humidityand dew point from dry bulb and wetbulb temperatures and elevation.

• TWO-calculates weighted rateof spread for the two-fuel modelconcept.

Up to 19 user-defined fire behaviorfuel models may be stored in the cal­culator, in addition to the 13 standardmodels that are always available. Out­put may be produced as a list of oneto three values for each output item orup to a 3-by-3 matrix of output valuesfor a single output item. Program

modules can be linked as in the BE­HAVE system, to easily pass outputsfrom one module to another for addi­tional calculations. The program willaccomrnodate either English or metricinputs and outputs.

Purchasing

The HP-71 B calculator, fire dangerCROM, and fire behavior CROMhave been placed on contract for thefollowing agencies: USDA ForestService; USDI Bureau of Land Man­agement, Bureau of Indian Affairs,National Park Service, Fish andWildlife Service; and State Forestryagencies. Orders should be placedwith:

Government Marketing Services,Inc.701 E. Gude DriveRockville, MD 20850Attn: Art PhillipsThe original and one copy of the

order must be sent to GovernmentMarketing, and the order must statethe contract number: 54-3187-5-35.Contract prices are:

Calculator $349.12Fire Danger CROM 37.80Fire Behavior CROM 58.80Those not authorized to purchase

from this contract can order HP-71'sand CROM's from Government Mar­keting at commercial prices.

The appropriate battery -operatedprinter is the HP "Think-jet" printer,model 2225B. The printer is not re­quired for effective use of the calcula­tor in the field because the user's

Fire Management Notes

manuals include forms for recordinginputs and outputs. The printer is al­most a necessity. however, if the userplans to write other programs for theHP-71B. Ease of programming andfiling capabilities suggest that manyuseful programs can be added to theHP-71B calculator. •

Volume 47, Number 2

Literature Cited

1. Andrews. Patricia A. BEHAVE Fire behav­ior prediction and fuel modelingsystem--BVRN subsystem, Part 1. Gen.Tech. Rep. INT-194. Ogden, UT: U.S. De­partment of Agriculture, Forest ServiceIntermountain Research Station; 1986.130 p.

2. Burgan. Robert E.; Susott, Ronald A. Firedanger computations with the Hewlett­Packard HP-71B calculator. Gen. Tech.Rep. INT-QOO. Ogden, UT; U.S. Depart­ment of Agriculture, Forest Service,Intermountain Research Station; 1986. [inpress]

3. Deeming, John E.; Burgan, Robert E.;Cohen, Jack D. The national fire-danger rat­ing system-1978. Gen. Tech. Rep.INT~39. Ogden, UT; U.S. Department ofAgriculture, Forest Service, IntermountainResearch Station; 1977. 63 p.

4. Susott, Ronald A.~ Burgan, Robert E. Firebehavior computations with the Hewlett­Packard HP-7IB calculator. Gen. Tech.Rep. INT-QOO. Ogden, UT: U.S. Depart­ment of Agriculture, Forest Service,Intermountain Research Station; 1986. [inpress]

13

Behavior of the Life-ThreateningButte Fire: August 27-29, 1985Richard C. Rothermel and Robert W. Mutch

Project leader, Fire Behavior Research Work Unit, USDAForest Service, Intermountain Research Station, Missoula,MT; and fire use specialist, USDA Forest Service, North­ern Region, Missoula, MT

On August 29, 1985, 73firefighters were forced into safetyzones, where they took refuge in theirfire shelters for I to 2 hours while avery severe crown fire burned overthem. The incident took place on theButte Fire on the Salmon NationalForest in Idaho. Five firefighters werehospitalized overnight for heat ex­haustion, smoke inhalation, and dehy­dration; the others escaped uninjured.Investigators estimated that withoutthe protection of the,escape zones andthe fire shelters, at least 60 of the 73firefighters would have died. Thanksto preparation of safety zones, the ef­fectiveness of the fire shelters, andthe sensible behavior of thefirefighters themselves, disaster wasaverted.

Behavior of the Butte Fire, particu­larly its explosive movement on theafternoon of August 29, is of vital in­terest to fire behavior specialists, indi­vidual firefighters, and leaders whomake tactical decisions based on firebehavior projections. That an alreadylarge and intense fire could rapidlyescalate to even higher intensity­some have called it a firestorm-andmove fast enough to overrun 73firefighters warrants review by any­one concerned with fire management.

Immediately after the shelter inci­dent, a review team was dispatched tothe Butte Fire to document the mete­orological conditions and fire behav­ior that contributed to the life­threatening run up Wallace Creek.Results of the analysis were distrib­uted to all wildland fire management

14

agencies early the following week.The review team was composed ofDennis Martin and Hank Walters,Forest Service Intermountain Region;Clyde O'Dell, National WeatherService; Dick Rothermel, Inter­mountain Fire Sciences Laboratory;and Bob Mutch, Forest ServiceNorthern Region. The purpose of thisarticle is to augment and expand theresults of the initial review throughadditional interviews with those whohad been on the fireline and an analy­sis of photographs taken during andafter the fire run. Art Jukkala and TedPutnam of the Missoula EquipmentDevelopment Center have also pre­pared a report on the performance ofthe fire shelter based on many inter­views with those who used it on theButte Fire. The article "Forest FireShelters Save Lives" in this issue in­cludes information on their findings.

A separate review of the Butte Fireand adjacent fires in the Salmon River(termed the Long Tom Complex),conducted by the Forest ServiceIntermountain Region in October1985, examined such topics as strat­egy, tactics, and other issues. The re­sults of this review are on file in theForest Service regional office inOgden, UT.

Fire Environment

Severe drought characterizedweather in the Butte Fire areathroughout the summer of 1985,contributing to critically low fuelmoisture levels. The fire weather sta­tion at nearby Indianola along the

Salmon River measured only 0.31inch of precipitation in June and 0.23inch in July. Although more than halfan inch of precipitation fell on twodifferent days in early August, someof this as snow, only 0.12 inch fellbetween August 13 and August 31. Ata remote automatic weather stationnear the fire, 1,0OD-hour fuel mois­ture readings from the National FireDanger Rating System were rated at 8percent prior to the run up WallaceCreek.

The weather on the Butte Fire fromMonday, August 26, through Friday,August 30, was not unusual consider­ing the location. Elevation at BaseCamp was 7,400 feet; elevations onthe fire ranged from 6,400 feet nearthe confluence of Wallace and OwlCreeks to 8,200 feet near the twosafety zones. Typical late afternoonmaximum temperature reached 70 to78 of, with minimum relative humid­ity in the 12 to 21 percent range atSourdough Base Camp. The windiestperiod each day occurred between1400 and 1500 mountain daylighttime. The velocity was generally be­tween 10 and 12 miles per hour, withstronger gusts. Inversions occurredeach day, breaking between 1130 and1330.

Weather on the day of the blowup,August 29, was not unusual, either.In the afternoon the temperaturereached the mid-70's, and minimumrelative humidity was in the upperteens. At base camp, low-level windswere out of the south at 8 to 12 milesper hour in the afternoon, with occa-

Fire Management Notes

sional gusts to 17 to 20 miles perhour. District personnel reported thatfuel loadings ranged from 80 to 100tons per acre in spruce- fir stands indrainage bottoms, to 25 to 40 tons peracre in higher elevation lodgepolepine-fir stands. Fuel models 8 and 10characterized most of the WallaceCreek drainage.

One unusual feature of the areathreatened by fire was the topogra­phy. The upper slopes did not con­verge into sharp peaks as is com­monly the case in the RockyMountains, but tended to bedomelike, with continuous crowncover. Wallace Creek itself was awell-defined north-south drainage thatbecame progressively steeper at itsheadwaters near the two shelter sites.

General Fire BehaviorThe Butte Fire was started by light­

ning on July 20, 1985. This fire waspart of the Long Tom Fire Complexin the Salmon River drainage, whichincluded the Com Lake, Bear, Foun­tain, Goat Lake, and Ebenezer Fires.The Butte Fire was first contained onAugust 5 at just over 20,000 acres.Strong winds fanned smoldering fuelsand spread fire across control lines onAugust 24 and 25. Fire activitypeaked on August 27, 28 and 29, asthe fire made runs of 1,000, 2,000,and 3,500 acres, respectively. About3,000 of the 3,500-acre growth onAugust 29 reportedly occurred inabout 90 minutes. It was during thisrun up Wallace Creek that the 73firefighters deployed their fire shel­ters. Simultaneously, another run of

Volume 47, Number 2

lesser severity occurred in Owl Creek,the drainage east of Wallace Creek.Both columns were characterized bydense black smoke. By midafternoonthe Wallace Creek column hadreached 15,000 to 17,000 feet aboveterrain and had a firm cumulus cap.Another area of intense fire activitytook place on the western flank wherethe fire spread northward but was ap­parently pulled into the main fire inWallace Creek.

Events of August 29

On August 29 wind velocities werenot especially high. In the early after­noon, eye level winds were measuredat 7 to 8 miles per hour at the conflu­ence of Owl Creek and WallaceCreek. At the higher elevation nearthe head of Wallace Creek, the localwinds were stronger. Division Super­visor Jim Steele estimated winds to be10 to 15 miles per hour, with gusts to20 miles per hour across the ridges.Measurements nearby confirmed thisestimate, but with gusts of 25 to 30miles per hour.

Figure I shows the fire area at 0200in the morning on August 28, the daybefore the big run, and its extent by2200 in the evening. By 0200 in themorning of August 29, the fire hadspread considerably further, havingcrossed the lower end of WallaceCreek and moved up the ridge towardOwl Creek. The burned areas in lowerWallace Creek were patchy. Of spe­cial importance on the morning ofAugust 29 were the spot fires in themiddle portion of Wallace Creek and

along Owl Creek at the southeast cor­ner of the fire.

An understanding of the fire controloperations is essential to understand­ing many events during the 29th.Having had little success at close-indirect attack on the 26th and 27th, theoverhead team had decided to use anindirect attack strategy. On the 28thand 29th, a tractor line was builtalong the main ridge on the north endof the fire, approximately 1.5 milesnorth of the nearest spot fires inWallace Creek (fig. I). Fortunately,the line construction included severalsafety zones 300 to 400 feet in diame­ter at approximately 1/4-mile inter­vals. The plan for the 29th was toconduct a burnout operation in thelate afternoon when humidity was ex­pected to rise. An aerial drip torchwould be used for center firing in theupper end of Wallace Creek. Crewswere to be dispersed along the line toburnout from the line after a convec­tion column was developed.

During the morning of August 29,spot fires near the confluence ofWallace and Owl Creeks threatenedvaluable timber and seemed to havethe potential to outflank the controlline to the east. Thus, it was decidedto use the helitorch early in the day tobum out and stabilize the line in thisarea. Initial attempts began just to thenorth of Owl Creek (marked A on fig.I) about 1200. The area did not bumvery well, and ignition attempts wererepeated. Bill Williams, the opera­tions chief, reported that this fire wasineffective at developing a significant

15

fire column necessary for improvingthe fireline.

While attempts to burn out linenear Owl Creek were in progress, thefire was developing strength in lowerWallace Creek. Three reports substan­tiate the development of fire inWallace Creek. Bill Williams re­ported a large convection column eastof Dishpan Springs. Dave Broberg,division supervisor in Owl Creek, re­ported two strong columns devel­oping, one near drop point 30 at theupper end of Sourdough Creek andthe other east of Dishpan Springs.Gary Orr, the division supervisor onthe west side at drop point 30, sawthe fire east of him throwing fire­brands into Wallace Creek. Orr re­ported that the fire in this area wasbecoming active around 1100.

The spots along Owl Creek also be­came active and developed a strongconvection column by 1300 (fig. 2).Smoke from these spots and from thehelitorch fire was moving to thenorth. It appeared to some that thesecolumns were being pulled to thenorth by the larger column developingto the northwest. With the aid of in­drafts to these columns, the helitorchwas used to burn out hand line anddozer line in areas C and D near theconfluence of Sourdough and OwlCreek.

Meanwhile, Gary Orr at drop point30 reported lots of fire in lowerWallace Creek. Considerable red col­oration could be seen in the smokecolumns, and at 1300 or 1400 the fire

ur

28

TlNCUP HILL

o~- .....I\\\

%~ I

\ I,--

/

tN

o 1/4 1/2 3/4 1 MileI I

Figure l-Arrows depict major fire runs on the Butte Fire during the afternoon of August 29, 1985.The 73 firefighters deployed fire shelters at the lower shelter area and Tin Cup Hill shelter area.Areas A, B, C. and D indicate where the helitorch burnout operation was conducted that afternoon.

16 Fire Management Notes

)

) --- ~ l,

" "',,\

Figure 2-Convecrion column development near the confluence of Sourdough, Wallace, and Owl Creeks at about 1515 m.d.t, on August 29. These col­umns originated from spot fires and helitorch operations,

was intensifying and moving upWallace Creek.

The helitorch continued burning outthe line in area C. Later, at approxi­mately 1500, area D was burned ac­cording to Bill Williams and DaveBroberg. Photographs looking northtaken from a helicopter just to thesouth of the convergence of Sour-

Volume 47, Number 2

dough, Wallace, and Owl Creeks (fig.2) show the smoke columns buildingat about 1515. From this vantagepoint, the strongest column was fromthe burnout operation and spots inOwl Creek. All of the smoke wasmoving northward up Wallace Creek.The firing operation at the south endof the fire was completed successfully

about 1550, and the fire was con­tained along the southern line just asit was reaching full strength in upperWallace Creek.

Wallace Creek Run

About 1515, Jim Steele, at thenortheast end of the fire, who later

17

went into his shelter at Tin Cup Hill,reported that he was walking on thetrail above the large clearcut andcould see fire coming up over a ridgeto the south. He reported that at thattime he could not see fire in WallaceCreek because of intervening smokeand trees. The fire he saw to the southwas probably coming out of OwlCreek.

Bill Williams reported that aboutthis same time a large, strong convec­tion column was standing over thefire. This column was within the mainnorthern dozer line, and Bill stillhoped to use indrafts from the columnto complete the planned burnout inupper Wallace Creek. Because a verysevere crown fire started moving tothe north up Wallace Creek on a west­ern exposure (the east side of WallaceCreek) through extremely heavyfuels, the helitorch was never used inthis area as originally planned.

Gene Benedict, the incident com­mander, was returning to the fire byhelicopter between 1500 and 1515and reported that "while viewing thisfire I had three other convection col­umns in view: Goat Creek on theSalmon National Forest, Hand Mead­ows on the Payette National Forest (anew start), and a fire on the Nezpercenear Cotter Bar. All fires were ex­tremely active with apparent strongconvective activity and substantialrates of spread, except for GoatCreek, which was topographicallyconfined."

After landing, Gene received re­ports that the fire in Sourdough Creek

18

had moved into Wallace Creek andhad started a firestorm. I Initial reportssaid it covered about 2 miles in 15minutes. (This later proved to be anoverestimation.) Right after the majorrun, a second run started on the westside near drop point 30, apparentlyoutside the dozer line. Initially, itspread rapidly to the north, but thenveered to the east, probably due to in­drafts from the larger column inWallace Creek. This secondary runthreatened firefighters along the lineon the west side, who were evacuatedby pickup truck and helicopter. Al­though this rescue was overshadowedby the fire shelter deployment, it wasnevertheless an intensive effort ac­complished safely.

Neal Davis, air attack supervisor,flew by helicopter around the fire justafter 1400 and again at 1515. He pro­vided estimates of the fire location inWallace Creek before the fire devel­oped the extreme behavior reportedlater. On his next flight, at 1550,Neal saw the firefighters in the safetyzones preparing to go into theirshelters.

Firefighter Steve Karkanen, work­ing between drop point 28 and thelarge clearcut at the head of WallaceCreek, recorded the movement of thecrown fire as it progressed up WallaceCreek. Steve took color photographsof the fire, recording his location, thedirection he was shooting, and the es­timated time and location of the firefront. His notes were especially help­ful in reconstructing the fire move­ment. His notes at 1600 describe the

nature of the fire as it passed aroundthe large clearcut:

Experiencing intense heat and highwinds from all directions. At leastthree large whirlwinds passed overthat were strong enough to knockpeople off balance. The area be­came too smoky and dusty to takephotos. The smoke column com­pletely enveloped everyone, and itwas impossible to see the fire. Visi­bility was reduced to zero severalseconds at a time, the air was veryhot, and the area was showeredwith burning embers. Personnelwithin the clearcut did not take totheir shelters, a dozer was used tobuild fireline around the vehicles,and the pumper crew worked onsmall spot fires in flashy fuels.

Personnel at the lower shelter areareported that the fire reached them at16\O. Jim Steele reports that thefirefighters on Tin Cup Hill went intotheir shelters approximately IOta 12minutes before those in the lower areadid. This would have put them intheir shelters at just about 1600, or acouple of minutes before. Steele fur­ther reports that the fire approachedthem at about 1545 out of a draw tothe southeast. While Steele waspreparing to get into his shelter, hetalked by radio to Strike Team LeaderRon Yacomella at the lower shelter

I Although referred to as a firestorm. it shouldmore properly be called a conflagration, whichis a severe spreading fire. The term "firestorm"is normally used to describe a severe stationaryfire or burnout of an area within aconflagration.

Fire Management Notes

Table I-Behavior of Wallace Creek fire run on the afternoon of August 29, 1985

Time period Elapsed time Distance Rate of spread

min mi milhr ch/hr143(}-1530 60 0.32 0.32 26153(}-1550 20 0.48 1.45 116155(}-1555 5 0.29 3.48 2781555-1600 5 0.14 1.68 134160(}-1610 10 0.15 0.90 72

area approximately 1,000 feet away.Ron asked if he should start his back­fire at this time, which he did. Hiscrew burned out approximately 200feet in front of the lower shelter zonebefore the fire hit at 1610. Theirbackfire started easily. At first strongindrafts pulled the fire and smoke to­ward the fire front, but later thesmoke blew back over the crew.

The Nature of the Fire

From observations by Neal Davis,Steve Karkanen, Jim Steele, and RonYacomella, we have reconstructed theprobable location and time of the firefront as it moved up Wallace Creekand overran the crews (fig. I). Therate of spread during the run is de­rived by scaling the distances fromthe map at each time line.

It appears that up until about 1530,although crowning and developingstrong convection columns, the firebehavior was similar to the behaviorobserved on the two preceding days(table 1). The spread rate was low,about '13 mile per hour. After 1530the fire spread much faster, with anaverage rate of about 2 miles. per hourand a maximum of about 3V2 milesper hour. This period was describedas a firestorrn by observers. The firehad to travel slightly over I mile inhalf an hour to reach the safety zone.In order for the firefighters to reachthe large clearcut from the lowersafety zone, they would have had tobegin the evacuation by 1530.

As with any fire, this one musthave moved by surges, with some pe-

Volume 47, Number 2

riods of little or no spread. The recon­structed spread rates are too coarse toshow the surges and appear to beslower than the impression receivedby observers on the ground.

Jim Steele reported that on Tin CupHill, firefighters in their shelters werehit by three waves of fire, the firstone from the southeast. The secondone burned up the north side and thenburned back towards them at aboutthe same time as the people in thelower safety zone were going intotheir shelters. The third wave hit fromthe southwest. Each time they werehit by a new wave of fire, the fire­fighters moved, crawling along theground inside their shelters searchingfor cooler areas of the safety zone. Atone time they moved away from thedozer piles of slash that had beenmade during the clearing of the safetyzone. After 40 minutes in their shel­ters, they came out, but dense smokeforced them back in again for another30 minutes. The air entering the shel­ters around the lower edges was ap­parently remarkably free of smoke.

The fire that overran the crews wasvery large and very intense. Figure 3

shows the nature of the fire as itpassed over the shelter and indicatesthe size of the column in comparisonto the trees. In the original colorslide, the convection column, showsred coloration for hundreds of feetabove the trees. The fire at this timewas almost certainly an independentcrown fire (4).

Viewed from the front, the fire ap­peared as a wall of flame 200 to 300feet high. Viewed from the air, aheadof the fire, the flames were estimatedto be two to three times the treeheight. The fire front was advancingas a typical standing flame with thebase of the fire in the trees. Theflames in the front were not seen to be .rotating or turbulent. The smoke was'rising sufficiently so that the flamecould be seen clearly. The columnrose nearly vertically, then tilted to­ward the north. The rear of the col­umn was a turbulent, swirling mass,impressive in its extreme behavior.

After the run, aerial inspection ofupper Wallace Creek revealed a large,intensely burned area in which allcrown needles and smaller surfacefuels were essentially gone. There

19

Figure 3-A view of the fire as it reached upper Wallace Creek and overran the fire crews. The crews deployed their fire shelters in safety zones similar tothose seen in the foreground. This photo was taken from a helicopter looking toward the east.

was, however, no evidence from theair, or on the ground near the sheltersites, of firestorm activity such as thatseen on the Sundance Fire in theIdaho Panhandle in 1967. Trees werenot laid down in patterns that wouldindicate large firewhirl activity. Somefirewhirls had been observed duringthe fire, but trees were not knocked

20

down, uprooted, or hroken off as theywere in the Pack River Valley as a re­sult of the Sundance Fire.

Inside the Fire Shelters

That all the firefighters in the es­cape zones survived without seriousinjury borders on the miraculous.

Nevertheless, the approach and pas­sage of the fire was a terrifying or­deal. Many, in fact, doubted that theywould live through it. The trauma ofthe event was reflected in interviewswith the survivors.

Witnesses, all of them experiencedfirefighters, said that this was no ordi­nary crown fire. To some it was a

Fire Management Notes

Passage of the flame front was accompaniedby a roaring sound, like that of a jet airplane

I

~) standing wall of flame that reached200 feet above the treetops. Othersdescribed it as a huge, rolling ball offire with a bright orange glow. Somewitnesses reponed large balls ofexploding gasses in the flame front.

Passage of the flame front was ac­companied by a roaring sound, liketbat of a jet airplane or a train. Onefirefighter found this the most fright­ening pan of the ordeal: "The noisebuilds up until you can't hear yourselfthink and then the ground begins toshake." He estimated that the shakingand roaring lasted 10 minutes. Overthe roar of the fire he could hear theshouts of nearby firefighters scream­ing for reassurance, followed byshouts of encouragement from otherfirefighters.

Strong, fire-induced turbulencemade it difficult to deploy sheltersand keep them down. One witness re­poned a feeling of weightlessness, ofbeing lifted off the ground. Anotherreponed the shelter being slammeddown against his legs. Within thesafety zones, everyone moved as faras possible from the flame fronts bycrawling along under the shelter.

Within the shelters, firefighters ex­perienced extreme heat for as much as10 minutes. Shelters were so hot thatthey could only be handled withgloves. Light entering the shelterthrough pinholes changed from darkred at peak intensity, to orange, towhite, as the fire passed over. Onesurvivor said that at one point theground looked as though it had been

Volume 47, Number 2

painted a bright orange. Firefighterslearned to evaluate the color of thelight as an indication of the fire's in­tensity in order to judge when it wassafe to come out of their shelters.

After leaving the shelters, somefirefighters showed symptoms of car­bon monoxide poisoning: vomiting,disorientation, difficulty in breathing.Emergency medical technicians ad­ministered oxygen to several individu­als; five were evacuated to a hospitalfor treatment and observation. Allfully recovered.

Among those interviewed, the con­sensus was that without the sheltersnone would have survived. A fire­fighter with 20 years' experiencesummed it up as follows: "The mostfrightening, scariest experience I'veever had, The fire was over us,around us, everywhere. I was inVietnam for a year, but this beats itall. "

Factors Contributing toFire Behavior

Fire activity in the preceding dayscontributed to the ease with which thefire in Wallace Creek began. Fire be­havior on the afternoon of Thursday,August 29, was a repeat, albeit amuch more severe repeat, of the firebehavior of the preceding two days.Each day took out more acreage andconsequently left a larger holdoverfire for the following day. On themorning of the 29th, the north edge ofthe fire was uncontained. Fuels wereburned in patches, leaving large

amounts of scorched fuel and treeswithin the fire area. The continuousfuels and lack of topographic barriersallowed the fire to move up the slopesof Wallace Creek with only moderatewinds. The topography contributedsubstantially to the fire behavior anddifficulty of control. The slopes fromthe valley bottoms were steep,contributing to rapid upslope runs; theridge tops were rounded and coveredwith continuous fuels. Hence, therewere no definite fire barriers such assteep rocky slopes, sharp ridges, orscrubby subalpine fuels.

Examination of weather recordsfailed to reveal any factors that wouldhave contributed to the large-scaleconvective activity observed on Au­gust 29. The extremely dry spring andsummer probably contributed to therapid spread of the fire and difficultyin controlling it. As on other fires inthe northern Rocky Mountains at thattime, tree crowns were extremelyeasy to ignite. Certainly the dry fuelson the ground also contributed, al­though the major fire runs at this ele­vation (6,000 to 8,000 feet) carriedpredominantly through the crowns.

Fire Behavior Analysis

Postfire analysis of the potentialfire behavior in surface fuels wasmade with the BEHAVE fire predic­tion system (I) and displayed on thefire characteristics chan (fig. 4). Fuelmodel 10 was used. The values forfuel moistures ranged between 3 and7 percent. The light winds of the

21

,

Rate of spread (chains/h)100rr--,--,--.........----------------.

morning and early afternoon wouldhave produced fireline intensities of250 to 500 Btu/ft.sec, making the firedifficult to control. The strongermidafternoon winds would have pro­duced fireline intensities in the sur­face fuels of 600 to 1,500 Btu/ft. sec,virtually assuring an uncontrollablecrown fire. The range of the condi­tions is shown by the ellipses on thefire characteristics chart. The inputsto BEHAVE and the outputs producedare shown in table 2.

~

I

\I

)I

!•

Conclusions

The type of fire run observed in up­per Wallace Creek on August 29 wasnot unusual for fires in lodgepole pineduring the 1985 fire season through­out the northern Rocky Mountains.The high-intensity fire runs were theresult of drought-induced, extremelylow fuel moistures in all size classesand the speed of the transition fromsurface fires to torching, spotting, andcrowning fires. Because large areaswere burning unchecked by eitherfirel ine or natural barriers and asoutherly gradient wind had rein­forced upslope and upcanyon after­noon winds in Wallace Creek, the di­rection of fire spread and crown firedevelopment before 1530 were not asurprise. The distance the fire spreadfrom 1530 to 1600, and its severity,

going up steep slopes (2). If we com­pare the calculated rate of spread inthe surface fuels with the crown firevalues given in table 2, we find thatfor the period 1430 to 1530 the crownfire was 1.4 to 2.3 times faster thanthe surface fire. In late afternoon,from 1530 to 1610, the crown firewas 2.6 to 5.3 times faster. Thesevalues fall within the suggested rangementioned above.

There is a great deal of uncertaintyin this type of calculation, indicatinga strong need for research on crownfire behavior and better guidelines forpredicting the onset and spread ofcrown fires and potential blowupsituations.

4000

The calculated rate of spread in thesurface fuels was 11 to 19 chains perhour in the morning and early after­noon. The higher windspeeds inmidafternoon would have pushed therate up to 28 to 57 chains per hour.We do not have methods for calcula­ting crown fire rate of spread, but ithas been found that crown fire spreadcan be 2 to 4 times faster than the rateof spread calculated for fuel model 10in fuels exposed to the wind and asmuch as 8 times faster if the fire is

1000500

20

40

60

80

22 Fire Management Notes

Table la-Data used in BEHAVE to assess fire behavior in surface fuels on the Butte Fire

Rate of Heat per Fireline Flamespread unit area intensity length

chainsthr Btulfr Rtutft.sec tt11-19 1286-1487 251-523 s.r-s28-57 1286-1487 664-1563 8.9-13.3

Element

Fuel model1-hrfuer moisture10-hr fuel moisture100-hrfuel moistureLive woody moistureMidflame windspeed:

Early afternoon (sheltered)Midafternoon (exposed)

Percent slopeWind direction

Table 2b-BEHAVE outputs

Time

Early afternoonMidafternoon

were, however, unexpected. Thelarge area of holdover fire adjacent tocontinuous timber with heavy surfacefuels proved to be a juxtaposition ca­pable of generating an incredibleamount of energy in a short time.

Although crown fires are often as­sociated with strong winds, in thiscase winds of only 10 to 15 miles perhour, with some stronger gusts, weresufficiently strong to channel the flowup the canyon and produce the excep­tionally intense crown fire thatoverran the crews. The question aroseas to whether the burnout operationwith the helitorch on the south side ofthe fire directly accelerated the highintensity run up Wallace Creek. Inter­views combined with a careful inspec­tion of burning patterns on a 1/24,000

Volume 47, Number 2

Data

103 to 7'%6%9%75%

4 to 6 mi/h10 to 15 milh45%Directly uphill

aerial photo mosaic did not reveal anyfire behavior process whereby thehelitorch burnout could have acceler­ated the run up Wallace Creek. Thephoto mosaic showed a patchy patternof burned and unburned areas be­tween the helitorch burning at theconfluence of Wallace and OwlCreeks and upper Wallace Creek. Theburnout operation, however, probablycontributed to the shelter incident bypreoccupying the attention of somekey overhead personnel for so muchof the afternoon of August 29. The"eyes in the sky" reconnaissance thathad been routinely available on previ­ous days was not available during thecritical time on August 29.

Early reports on the Butte Fire esti­mated that the fire traveled 2 miles up

Wallace Creek in 15 minutes, or aspread rate of 8 miles per hour. Thisestimate now appears to be considera­bly higher than the actual rate ofspread. Reconstruction of the firefront location at various times indi­cated that the average spread rate wascloser to 2 miles per hour with a max­imum of about 311z miles per hour.

The safety zones that were bull­dozed into the tractor line at the headof Wallace Creek made it possible for73 firefighters to safely and effect­ively use their fire shelters and sur­vive one of the more violent fire runsobserved in the northern Rockies in1985. But, as one crew foreman ob­served after the incident, "the bestsafety zone is one where a fire shelteris not needed." This conclusion de­serves special emphasis whenever theButte Fire is discussed.

Preventing Future Incidents

What measures can be taken to pre­vent such a life-threatening eventfrom recurring in the future? If an in­direct attack strategy is selected, thena fail-safe warning system must be inplace to absolutely clear the line ofpersonnel well in advance of a high­intensity run. Another approach in co­nifer forests is to select a direct attackstrategy, build a line along the flanksof the fire from a well-secured anchorpoint, and attack the head of the fireonly when fuels, weather, and topo­graphic conditions allow firefightersto work safely.

23

Whatever the strategy selected, thefundamental principles of fire behav­ior and fire suppression should alwaysguide decisions that affect the healthand welfare of the firefighter. Despitethe remarkable progress made in firemanagement in the past quarter of acentury-better understanding of firebehavior, better trained and equippedfire crews, more flexibility in attackstrategy-conditions like those expe­rienced in the northern Rockies in thesummer of 1985 call for extreme vigi­lance in all aspects of fire suppres­sion. And the safety of the individualfirefighter is always the toppriority.•

Literature Cited

I. Andrews, Patricia L. BEHAVE: Fire be­havior prediction and fuel modelingsystem-BURN subsystem, Part 1. Gen.Tech. Rep. INT-194.. Ogden, UT: U.S.Department of Agriculture, Forest Service,

Intermountain Research Station; 1986.130 p.

24

2. Rothermel, Richard C. How to predict thespread and intensity of forest and rangefires. Gen. Tech. Rep. INT-143. Ogden,U'F: U.S. Department of Agriculture, For­est Service, Intermountain Forest andRange Experiment Station; 1985.

3. U.S. Department of Agriculture, ForestService. Your Fire Shelter. Missoula, MT:Missoula Equipment Development Center;1984.

4. Van Wagner, C.E. Conditions for the startand spread of crown fire. Canadian Journalof Forest Research 7:23-24. 1977.

Butte Fire Shelter VideotapeAvailable

A 33-minute videotape on the"Butte Fire Shelter.Deployment" isnow available' from the Boise Inter­agency Fire Center. Videotaped onsiteinterviews with shelter occupants justdays after the incident vividly demon­strate the importance of everyfirefighter having a shelter, knowingdeployment procedures, and caring

for the shelter properly. The i

videotape provides an excellent sup- "plement to Missoula Equipment De- i

Ivelopment Center's new training filmon fire shelters. Copies of thevideotape can be ordered from:

Bureau of Land ManagementBoise Interagency Fire Center3905 Vista Ave.Boise, ID 83705

The tape is offered in the followiugformats:

1. Order No. NFES 1523 3/4"U-matic $22.64

2. Order No. NFES 1524 112"VHS 11.243. Order No. NFES 1527 112"B~ II.~

Add 19% to all orders for shippingand handling.

Fire Management Notes

I

!I!I

i!

I

lI'

An Application of NIIMS on the UintaNational ForestHelen Woods and Lyle Gomm

Program assistant and chief, Branch of Recreation andLands, respectively, Uinta National Forest, Provo, UT

rojI,

~"

I

l

Since the implementation of theNational Interagency Incident Man­agement System (NIIMS), local of­fices of Federal agencies and Stateand local agencies, have banded to­gether to cooperatively manage com­munity emergencies such as fire,floods, toxic spills, aircraft disasters,rescues, earthquakes, and riots, Oneof the specific cooperative ventures, apump and hose-lay team, is describedin this article to show how agenciesare working together more effectivelyand efficiently under NIIMS,

Many fires along the Wasatch Frontof the Uinta National Forest occur inlocations where they can be effect­ively suppressed using water fromnearby lakes, streams, and irrigationcanals or ditches. Most of these firesstart in heavily used campgrounds andother recreation areas located in riverand creek bottoms. The burning ofditchbanks on farms, debris burning,and other misuses of fire also contrib­ute. In accordance with the NIIMSconcept, the Uinta National Forest,the Bureau of Land Management,Utah Division of State Lands and For­estry, and Utah County havecooperatively organized a pump andhose-lay team to fight fires withintheir jurisdiction.

The pump and hose-lay team is agroup of trained firefighters from thecooperating agencies who can effect­ively take advantage of availablewater sources lu fight fifes along thesteep Wasatch Front and in areas thatare often inaccessible to engines. Thisteam is trained to work together dur-

Volume 47, Number 2

INTERAGENY INCIDENTCOM AND

Figure l-Cooperating agencies worked together to set up a pump and hose-lay team to help fightfires in the Uinta National Forest.

.~, ...-.

Figure 2-Equipment used by pump and hose-lay team.

25

~llii~:=~-------1200 ft.

Figure 3-Diagram of use of equipment to fight a typical frontal fire.

equipment and use the trailer.The equipment carried on the trailer

is suitable for pumping operations,but it can also be used with city firehydrants. It can be used in remote ordesert areas with either large tankers

"

•

200 ft.

600 ft.

400 ft.

Lateral hoselaysfordirect attack

< ii·.···b.,----800 ft.

Locationof fire enginepump to first reservoir

Residential areaon valley floor

Burned area~

Direction of \fire spread

sary fittings have been contributed bythe Bureau of Land Management.Utah County took care of painting thetrailer. Utah County has also workedwith the Forest Service in trainingpart of its strike team to operate the

ing the initial stages of the incident toset up the system. After the hose sys­tem is in place, the crew can thensplit into two three-person shifts andutilize other, untrained personnel towork along with them. The crew isalso trained in the use of gravity-fedsystems that are used if the watersource is high or if a helicopter isused to fill a tank on a ridge and thefire lies below in a canyon.

The pump and hose-lay team is de­signed to provide initial and second­ary attack on frontal slope wildfireswithin and adjacent to Utah County.The high-risk and flashy fuels locatedalong the Wasatch Front require im­mediate and aggressive action once afire is identified. Any fire in this areahas potential to be threatening to lifeand property.

A trailer to house the unit was con­structed by the Forest Service and iscapable of carrying pumps and hoseequipment needed to run hose 3,000feet horizontally and 1,000 feet verti­cally (fig. 1-2). There are three largepumps, two Mark III pumps, and oneChrysler Flotopump. To augment thelarge pumps, the two smaller Mark IIIpumps (contributed by the Utah Divi­sion of State Lands and Forestry) areused to run smaller lines from relaytanks set up to help in moving thewater uphill. The exact pump config­uration, of course, depends on the sit­uation and application. Four light­weight relay tanks are included, alongwith a much larger SOD-gallon, open­top tank suitable for helicopter fill op­erations. Much of the hose and neces-

26 Fire Management Notes

Figure 4-Equipment used by the pump and hose-lay team is designed to make water available upI,OOO-foot vertical slopes.

ing capability. The computer will beused to store and retrieve the informa­tion necessary to manage a large fireorganization, information on overheadpersonnel, fire crews, earth-movingequipment, aircraft, fire engines, andso forth.

In the Uinta National Forest,NIIMS is working and working well.The cooperating agencies areexperiencing success in thejoint man­agement of fire and other communityemergencies that may arise. The ef­fective interagency cooperation shown

Initial attack at headoffire fromwithin burned area

Reservoir location

in developing and maintaining thecommand post trailer and the pumpand hose-lay team is evidence of thesuccess, of the NIlMS concept.

Utah County firefighters are excitedabout the training they are receivingin fighting wildfires. Forest Servicecrews are eager to use the new equip­ment. Each contributing agency cansee with pride the result ofcooperation. •

'"Drywest facing slopesdirectly aboveestablished communities

800 ft.

400 ft.

600 ft.

1000 ft.

Residentiaiarea

1200 ft.

or helicopter drops into artificialponds. These alternative watersources would be more expensivethan using a nearby lake or stream.

The equipment used by the team isdesigned to make water readily avail­able up I,OOO-foot vertical slopes.This is achieved through lightweightpumps, small reservoirs, and thou­sands of feet of linen hose. Theequipment is designed to bebackpacked up the slope as the firedevelops. It is also designed for fight­ing the fire from within the burnedarea. In this way, direct attack can bemade ahead of the fire without threatto firefighter safety (fig. 3-4).

Other pieces of equipment havebeen assembled, under the NIIMSconcept, to supplement the hose-layunit. The Interagency Incident Com­mand Post trailer contains all neces­sary equipment for an efficient firemanagement operation, A military­type command and staff organizationfunctions out of this trailer to handleincident objectives to which it is as­signed (fig. 5).

During the fire season, the trailer isoutfitted with a generator, lights,sleeping gear, and other personal gearfor six crew members. The unit alsocontains radio equipment for commu­nications to all cooperating agenciesand functions as a backup unit to thecentral dispatch office located in theUtah County sheriff's office.

A personal computer will behoused within this facility to providetelecommunication and word process-

'I,J

I

Ij'..

,.I

Volume 47, Number 2 27

------ ----~

1

•

Figur~ S-Command post trailer organized under NIfMS concept contains equipment for efficientfire management operation.

28 Fire Management Notes

•

\~

Current Status of BEHAVE SystemRoger L. Eubanks, Roger L.Bradshaw, and Patricia L. Andrews

Fuels management specialist, USDA Forest Service,Washington, DC; computer programmer/analyst, USDAForest Service, Boise, ID; and mathematician, USDA For­est Service, Missoula, MT

,

I

BEHAVE is a set of interactive,user-friendly computer programs thatare used for site-specific fire behaviorprediction. An overview of BEHAVEwas given in a previous Fire Manage­ment Notes article by Richard C.Rothermel, "BEHAVE and YOU CanPredict Fire Behavior" (5). That arti­cle included information on what BE­HAVE can do, where BEHAVEapplies, and how to make specific ap­plications. This article will give infor­mation on the current status of theBEHAVE system, documentation thatis available, how the programs can beaccessed, and additional developmentthat is underway.

Responsibility

BEHAVE was developed by USDAForest Service Research personnel atthe Intermountain Fire Sciences Labo­ratory (previously known as theNorthern Forest Fire Laboratory). TheAviation and Fire Management Staffin the Forest Service Washington,DC, office has accepted BEHAVE asa national system. Local specialiststrained in the BEHAVE system areresponsible for providing users withassistance in operating BEHAVE.BEHA VE instructors have beentrained in each Forest Service regionand in other Federal agencies thathave wildland fire management re­sponsibilities. Others who have re­ceived training as instructors includeState and university employees. Thiscorps of instructors has responsibilityfor training additional BEHAVEusers.

Volume 47, Number 2

Documentation

The primary documentation of theBEHAVE system consists of two For­est Service Intermountain StationGeneral Technical Reports (GTR).Burgan and Rothermel describe theFUEL subsystem (3); Andrews de­scribes the BURN subsystem (1).Rothermel's GTR "How to Predict theSpread and Intensity of Forest andRange Fires" (4) provides the basisfor his article "BEHAVE and YOUCan Predict Fire Behavior" (5) andAndrews and Burgan's" 'BEHAVE'in the Wilderness!" (2). The BE-HAVE Terminal User's Guide is aquick reference for worksheets,codes, program structure, etc. It isprinted in looseleaf format to allowfor future updates.

The two GTR's and the TerminalUser's Guide are available throughthe NWCG Publications ManagementSystem in Boise, !D.

Programs

The master copy of the BEHAVEprograms is maintained on the USDAcomputer in Ft. Collins, CO. Thesource code is available to users whowish to use BEHAVE on their owncomputers. The programs are writtenin ANSI Standard X3.9-1978 Fortran.The largest program, FIRE 1, requiresapproximately 260K bytes to load onData General MV/8000. BEHAVE isbeing successfully used on a varietyof computers.

Forest Service users with DataGeneral equipment will receive the

BEHAVE programs and all updatesthrough the regular Forest Servicesoftware update procedures. The localsystems manager is responsible fordeciding whether or not to place BE­HAVE on the system. Those notusing Data General may send a writ­ten request to Roger Bradshaw statingtheir name, address, and phone num­ber. Requests for tapes should includethe character set, tape density, block­ing, labeling, tracks, record length,and a blank tape.

Modification of the programs to runon a local computer is the responsibil­ity of the user. Users are also respon­sible for making sure that any localcopies of the program match the mas­ter. Version numbers are used to indi­cate changes made in one or more ofthe BEHAVE system programs. Achange in the tenth's digit indicatesminor changes that correct programerrors or enhance existing features;the one's digit indicates majorchanges such as the addition of newprediction models. At the time of thiswriting the version number of allthree programs is 3.3.

A BEHAVE conference is on theconference system at the FI. CollinsComputer Center. The conferencewill be used to announce new ver­sions of the programs, indicate neces­sary changes, and tell how to obtainupdates. Users are responsible formonitoring the conference or makingother arrangements for obtainingannouncements.

Requests from outside of theUnited States are handled by Patricia

29

System

Behavefire behaviorpredictionand fuel modeling system

Subsystems

Fuelfuel modelingsubsystem

Burnfire predictionsubsystem

Programs

New MOLinitial fuel modeldevelopment program

TSTMDLfuel model test andadjustmentprogram

Fire 1prediction program

Fire2prediction program

Modules

Sitespread rate and intensity calculation module(site - specific input)

Directspread rate and intensity calculationmodule(direct entry ofgeneral input)

Sizearea and perimeter calculation module

Containattack force requirement calculation module

Spotmaximum spotting distance calculation module

Dispstchautomatic linking of Direct, Size and Contain

{ Calculation ~odules to be added later

Subsystems, programs, and modules that make up the BEHAVE fire behavior prediction and fuel modeling system.

L. Andrews. She will see that foreignusers are kept informed of updates.An attempt will be made to arrangefor a single contact for each country.So far, BEHAVE has been sent toCanada, Australia, South Africa,Spain, Mexico, and Portugal.

Updates

A diagram of the current BEHAVEsystem design is shown in figure I.The FUEL subsystem consists of theprograms NEWMDL and TSTMDL.At this time the BURN subsystem hasonly the FIRE 1 program. However, aFIRE 2 program is being developed,and the FIRE 1 program is being ex-

30

panded. FIRE 2 will include the fol­lowing modules:

• MOISTURE-The fine dead fuelmoisture model that is now in SITE.This model will allow table output forranges of input parameters as well asgraphic output showing diurnalvariation.

• IGNITE-Probability of ignition.• RH-Relative humidity and dew

point from elevation; wet bulb anddry bulb temperatures.

The FIRE I program will be ex­panded as follows:

• SPOT-Addition of spotting dis­tance from wind-driven surface fires.

• CONTAIN-Addition of con­tainment by indirect allack.

• SCORCH-Scorch height.The keywords LOG and NOLOG

are in the current version of BE-HAYE, but they are not described inthe fonnal documentation. This de­scription was added in response touser request. Because the BEHAYEprograms are interactive, printing anentire session on paper is quite un­wieldy. The LOG option saves onlyinput listings and computed results in .a file. When the session is complete,the user is reminded to print the fileand then delete it. (The method ofprinting and deleting is a function ofthe user's computer and is not part ofthe BEHAYE system.) In this way auser can use a nonprinting terminal,

Fire Management Notes

.'

I1

'!

j,

yet have paper printout of the impor­tant parts of the run. The next versionof the BURN programs will also havethe keyword COMMENT to allowuser notes to be printed in the file.

Summary

BEHAVE has been an approvednational system since 1984 and is cur­rently being used for a variety of firemanagement applications. It is a dy­namic system that is expected tochange as new technology becomesavailable for field application. Watchfor future articles in Fire ManagementNotes describing the updates and ap­plications in more detail.

If you have any questions or sug-gestions, contact Roger Eubanks at:

USDA Forest ServiceP.O. Box 2417Washington, DC 20013(703) 235-8666FTS 235-8666

Volume 47, Number 2

Program maintenance is performedby Roger Bradshaw, who can bereached at:

USDA Forest Service3905 Vista Ave.Boise, ID 83705(208) 334-9458FTS 554-9458Requests from outside the United

States should be directed to PatriciaL. Andrews at:

Intermountain Fire SciencesLaboratoryP.O. Box 8089Missoula, MT 59807(406) 329-4827FTS 584-4827 •

Literature Cited

I. Andrews, Patricia L. BEHAVE: Fire be­havior prediction and fuel modelingsystem-BURN subsystem. Gen Tech.Rep. INT-194. Ogden, UT: U.S. Depart­ment of Agriculture, Forest Service,Intermountain Research Station; 1986.130 p.

2. Andrews, Patricia L.; Burgan, Robert E."BEHAVE" in the wilderness! In: Lotan.James E.; Kilgore, Bruce M.; Fischer,William E.; Mutch, Robert W., tech.coords. Proceedings, symposium andworkshop on wilderness fire; 1983 Novem­ber 15-18; Missoula, MT. Gen. Tech.Rep. INT-182. Ogden. VT: V.S. Depart­ment of Agriculture, Forest Service,Intermountain Forest and Range Experi­ment Station; 1985: 306-309.

3. Burgan, Robert E.; Rothermel, Richard C.BEHAVE: Fire behavior prediction andfuel modeling system-FUEL subsystem.Gen. Tech. Rep. INT-167. Ogden, trrU.S. Department of Agriculture, ForestService, Intermountain Forest and RangeExperiment Station; 1984. 126 p.

4. Rothermel, Richard C. How to predict thespread of forest and range fires. Gen.Tech. Rep. INT-143. Ogden, UT: U.S.Department of Agriculture, Forest Service,Intermountain Forest and Range Experi­ment Station; 1983. 161 p.

5. Rotherrnel, Richard C. BEHAVE andYOU can predict fire behavior. Fire Man­agement Notes. 44(4): 11-15; 1983.

31

Crew Mobilization:What's the Next Step?

Stephen W, Creech

State fire coordinator, Indiana Department of Natural Re­sources, Division of Forestry, Martinsville, IN

This is the second part of a two­part article dealing with interagencyfire crew mobilization. Part I dealtwith the preparation and planning thatare necessary to pull personnel to­gether in the most efficient manner tomeet mobilization deadlines. Part 2will deal with fire camp, fireline oper­ations, and demobilization. This in­formation is intended to address someof the problems that the crew bassistrike team leader, and squad boss, aswell aSJrookie and veteranfirefighters, may encounter and to uf­fer suggestions to make the assign­ment as positive an experience as pos­sible. We must all remember that ourrole in these assignments is to servethe host agency and not vice versa.

Part I left off as the crew boardedthe aircraft for the fire assignment.We will now pick up as the planelands, and we begin our assignment.However, you and your crew have notalready encountered lengthy delaysyou had better prepare yourselves forthe inevitable. These sometimeslengthy and always frustrating delaysare probably the number one enemyof good crew morale. Dealing withthe possibility of delays at the begin­ning can prevent more serious prob­lems later on.

What I would like to do now is de­viate from the story line a bit to lookat some of the National InteragencyFire Coordination Center (NIFCC)statistics for 1985. There were ap­proximately 83,000 reported firestarts that burned more than2,975,000 acres. To combat these

32

fires required in excess of 25 class Iteams, 2,510 miscellaneous overhead,276 category I crews, 1,010 categoryII crews, 749 smokejumpers, 473fixed and rotary wing aircraft (firesuppression aircraft), $9,721,000worth of catering services,$1,194,000 worth of shower facili­ties, and $8,500,000 worth of charterand contract aircraft. In addition,more than 50,000 firefighters weremoved throughout the country. Thesestatistics do not include ground trans­portation or the assistance providedfrom the military.

I think that now you can appreciatea little better why there may be someunexpected delays and perhaps beable to deal more effectively with thequestions fired at you from the crew.I will now jump ahead a bit and as­sume that we have safely reached firecamp.

Fire camp is always a very hecticplace. Don't expect to be greeted atthe gate and shown to your quarters.The first thing to do is to let someoneknow that your crew has arrived. Thefire camp may be operated under ei­ther the National Interagency IncidentManagement System (NIIMS) or theLarge Fire Organization (LFO). Thereporting procedure will vary. UnderNIIMS, the strike team leader/crewboss or the crew representative shouldfirst report to the planning section, re­source unit leader (RESTAT). UnderLFO the crew boss would report tothe maps and records section.

You should understand that yourcrew members may be utilized in one

of several ways. They may be consid­ered a single resource, they may bejoined with another crew and formedinto a strike team, or they may be­come part of a task force. How theyare utilized will depend upon theneeds of the incident and may changeaccording to changing needs.

Next, crew manifests and time­sheets must be taken to the financesection (this is the same under bothNIIMS and LFO). A close checkshould be made here to see that atimesheet is made out for each mem­ber of the crew. A little extra efforthere can prevent problems later on.The last stop will be to check in withthe staging area manager (NIIMS) orthe camp officer (LFO). Here youwill be assigned an area to assembleyour crew. Once this area is estab­lished, place a sign (one may be pro­vided) on the perimeter showing thecrew name and crew leader. This areawill become your home for the nextseveral days.

The strike team leader/crew bossmust decide if the crew is sufficientlyrested, considering travel time andrest opportunities, to accept a line as­signment. The incident may not re­quire that your crew be placed on thefireline immediately, but if you areasked to assume a fireline role, it isyour responsibility as leader to evalu­ate the crew's ability to safely carryout out the assignment. The crew'swell-being depends on you. Regard­less of whether you are immediatelyassigned to line duty, there are certainthings that must be routinely taken

Fire Management Notes

•I.

.'

I

l'

care of concerning your personnel.Your crew will do a better job if

they are happy and healthy. Preven­tion is the key to success. Let's lookat some simple yet effective preven­tive measures:

1. Hold briefings. Your crewmembers want to know what is goingon. They are not looking for detailedinformation-they just want to knowwhat is happening and what is theirrole. Where are they? Where will theybe going? What is the fire doing?What is the terrain like? How long isthe work shift? Tell them what youknow. Don't lie and don't speculate.Personnel who feel informed willhave fewer morale problems thanthose who are kept in the dark.

2. Information dissemination,Make sure good information gets backto the home unit and that the homeunit passes the information along tofamily and friends left behind. Thegood public relations that can be ob­tained from fire assignments goes along way to ensure that the crewmembers will be available the nexttime they are needed. As long as thefirefighting personnel feel that theirbest interest is being considered, theywill remain happy and productive. Itwill be necessary to plan ahead howand when family and friends will becontacted. The home unit has the re­sponsibility to disseminate timely in­formation. Planning is the key to suc­cess here.

3. Proper camp habits. Insist ongood eating, rest, and hygiene habits.Failure to adhere to anyone of these

Volume 47, Number 2

can spell disaster. The need to ob­serve this must be stressed at the earlybriefings. Individual counseling maybe necessary in some cases. Since thefire situation can change very rapidly,it may be necessary to alter plans. Forthat reason it is suggested that aschedule be established while thecrew is in camp, so that someone is atthe designated crew area at all times.This will greatly improve responsetime and will eliminate unnecesarydelays.

Note-As a strike team leader/crewboss you may be faced with a situa­tion where your crew is allowed to gointo a nearby town. This can happenduring staging or during 24-hour onl24-hour off work cycles. You are stillresponsible for your personnel, andyou have the ultimate authority to ap­prove or deny the privilege. It is notan easy decision and can prove mostunpopular.

4. First aid. First-aid problemsmust be attended to immediately.Problems can occur while on the lineor in camp. Don't assume the prob­lem will go away. Make sure promptmedical attention is provided. If first­aid is administered on the line, havethe individual report to the first-aidfacility upon return to camp. Docu­ment any problems to assist in futureclaims.

5. Personal problems, Action heremust be swift and firm. You mayneed help in dealing with certain mat­ters so don't be afraid to ask. Moraleor attitude problems, if leftunchecked, can rapidly spread to

other personnel and literally disable acrew. All members of the crew shouldfeel as though they can discuss theirproblems openly at any time.

6. Safety. Safety is one area thatcan never be stressed enough. Safetyin camp, on the way to the fireline,and on the line must be constantlymonitored. Unsafe conditions shouldbe corrected or brought to the atten­tion of others. Don't think that theSafety Officer can catch every safetyproblem. Safety is the responsibilityof all firefighters, but particularlythose in a supervisory position.

7. Per.formance evaluation. It isthe crew member's responsibility toobtain crew performance evaluationsfrom the unit supervisors they areworking for. Don't assume that theevaluations will automatically becompleted for you. The crew per­formance evaluation is proof of howwell your crew performed and shouldpoint out areas that need improve­ment. These evaluations can mean thedifference between being utilizedproperly and not being utilized at all.Incident commanders are looking forpersonnel who can do the job. Theevaluation is your proof that you andyour crew can perform satisfactorily.

The key to success on any fire as­signment is to maintain control andwork as a team (fig. I). Work to thebest of your ability, and don't beafraid to admit that you can't dosomething or that you are not properlytrained or qualified. Firefighting is adangerous profession and the more weall know and the more we work to-

33

Figure i-The fire crew's performance will be reflected in a crew performance evaluation.

gether the safer and more effective webecome.

We will now assume that our as­signment is completed and we havegone to demobilization status. Onceyou have fulfilled your time require­ment or the emergency has subsided,you must be released back to yourhome unit. The process of checkingout is much the same as checking in.You will be required to turn in anyaccountable, issued gear. You willhave to verify, sign, and take chargeof timesheets. You must obtain copiesof any request for treatment formsfrom the first-aid area. Your camparea must be cleaned up, and youmust sign out through RESTAT. Itmay seem harder to get out of camp

34

than it did to get in.Once you are sure of your demobi­

lization plans, particularly flightplans, it is a good idea to make onelast contact with your home unit to as­sure that someone will be there tomeet you. Remember that on a large­scale operation things can change andmessages and arrangements occasion­ally get messed up. It may not alwaysbe possible, but a short call just priorto demobilization may save a lot ofconfusion later on. You need to indi­cate mode of transportation, the carri­er's name, the destination (airport andgate location if known), and approxi­mate time of arrival. You might alsoplan to have a meal waiting just incase. This kind of planning isn't fool-

proof, but it sure helps smooth out theprocess.

Once you arrive at your officialduty station, you still have a fewloose ends to tie up before you canclose the books on this incident. Youneed to reclaim the fire gear fromyour personnel. Any materials thatcame from the fire cache should beput back, or plans should be made toreplenish depleted items. If it is feasi­ble, have the crew members acknowl­edge that they arrived home safely. Aquick phone call back tu the home of­fice or to the strike team leader/crewboss is all it takes.

You should plan to hold an assign­ment critique when time allows, pref­erably while the incident is still freshin everyone's mind. The critiqueshould look at the following areas:

I. Incident management. Youshould include an evaluation of eachof the phases of the assignment. Ex­amine mobilization, staging, firecamp, line assignment, safety, organi­zation, and demobilization. You arelooking for items, both good and bad,that could make future assignmentsoperate more efficiently.

2. Personnel evaluation. Youneed to look at all of your fire person­nel and evaluate their performance onthe assignment. Did they follow or­ders? Were they team players? Werethey good representatives of yourhome unit? Did they perform up tostandard? Could they be consideredfor transfer to a different position (forexample, firefighter to squad boss, or

Fire Management Notes

.1

'" squad boss to strike team leader/crewboss)? Were they physicaIly quali­fied? Poor performers should be re­moved from the roster and not al­lowed to participate in futureassignments until shortcomings arerectified. Remember the whole team'sevaluation may hinge on one person '8

actions.3. Training needs, You should

evaluate training needs very closely.Was there a need for training that youor yourcrew have not had? Do certainpersonnel need a refresher course? Isthere a need for more advanced train-

Volume 47, Number 2

ing for some of your personnel?Training is an ongoing consideration.It is needed both to maintain profi­ciency and to prepare personnel foradvancement.

Once you have completed the cri­tique you should make copies avail­able to your home office, the sendingagency, the receiving agency. andeach of your personnel. The critiqueshould be objective and honest. Youshould immediately develop a plan ofaction to correct those areas withinyour jurisdiction that need improve­ment. Your efforts here will payoff

on the next assignment for everyone.If you are a veteran leader you

should have picked up some helpfulreminders from this article. If you area veteran firefighter you should havegained some useful information incase you are ever given the responsi­bility to lead your own crew/striketeam. If you have only dreamed ofgoing on an out-of-state fire assign­ment you should have a better appre­ciation of what to expect and also abetter understanding of what goes intosuch a complex undertaking.•

35

Late-Winter Prescribed Burns toPrepare Seedbeds for NaturalLoblolly-Shortleaf PineRegeneration-Are They Prudent?Michael D. Cain

Research forester, USDA Forest Service, CrossettExperimental Forest, Southern Forest Experiment Station,Crossett, AR.

'f

I,

Density, percent stocking, and totalheights of first-year loblolly andshortleaf pine (Pinus taeda L. andP. echinata Mill.) seedlings weremeasured one growing season after alate-winter prescribed burn inuneven-aged pine stands in southernArkansas. These data were comparedwith data from the same year for anarea where there had been no pre­scribed burning. Although a partialseed catch on the late-winter burnarea produced significantly fewerpine seedlings, the burn coincidedwith a bumper seed year that resultedin a high-density, well-stocked crop ofpine seedlings after one growingseason.

Introduction

The use of prescribed fire forseedbed preparation in natural regen­erationof loblolly and shortleaf pine(Pinus taeda L. and P. echinataMill.) is widely practiced throughoutthe South. The recommended tech­nique is to use a winter burn to reducefuel accumulation and follow up withone or more hot summer burns to top­kill small hardwoods and expose min­eral soil (7). To maximize seed catch,prescribed burning should be donejust before seedfall (8. 16). Such tim­ing of prescribed bums will provide amineral soil seedbed yet will not de­stroy the current year's seedfall.

Natural seedfall for loblolly pinebegins in October and may persistinto spring. However, seedfalt rnoni-

36

toring in North Carolina has shownthat 71 percent of loblolly seeds fallby the end of November (1) and that84 percent of all loblolly seeds fallbefore January (13). Similarly, insouth Arkansas, Grano (12) foundloblolly-shortleaf pine seedfall to be77 percent complete by the end ofNovember and 92 percent completeby the end of December. Chaiken (6)monitored loblolly pine seedfall for 6years in South Carolina and noted that92 to 100 percent of all viable seedswere disseminated by February 1.

On the basis of these findings, aneffort should be made to coordinatethe timing of prescribed fires forseedbed preparation with natural pineseedfall. For example, Lotti andothers (14) surmised that if apreharvest prescribed winter fire takesplace after the main seedfall in lob­lolly pine stands, almost full depend­ence must be placed on the followingyears' seed crop for pine regenera­tion. The reasoning has been that pineseed on the ground will be destroyedby the fife.

Yet, fuel and weather conditionsfor prescribed burning in the Southare often most favorable during thewinter when cold weather fronts passthrough with steady north winds (8).It is therefore appropriate to investi­gate the effect of a late-winter pre­scribed bum on subsequent pine seed­ling establishment from a currentyear's seed crop.

Methods

Data for this investigation were ob-

tained from two active research stud­ies on the Crossett Experimental For­est in south Arkansas. Soil on thestudy areas is Bude silt loam(Glossaquic Fragiudalfs) with a siteindex of 85 to 90 feet for loblolly pineat age 50 years.

Study A-The purpose of thisstudy was to assess the effects ofprelogging hardwood control for es­tablishment of natural pine regenera­tion. When the study began, the standcontained about 100 square feet ofpine basal area per acre. On four0.25-acre plots all hardwoods havinga groundline diameter of I inch orlarger were stem-injected withpicloram (Tordon IOlR) in March1983. Herbicide spotguns were usedto treat four other 0.25-acre plots withhexazinone (Velpar-L) using 4 poundsa.i. per acre on a 4- by 4-foot grid inApril 1983. A basal area reductionharvest of overstory pines was done inJuly and August 1983, removing anaverage of 6,000 board feet per acre(International 'I.-inch rule). This re­duction left a 76-year-old even-agedstand of loblolly-shortleaf pinesaveraging 68 square feet of basal areaper acre in 28 trees per acre. Since theintent of the study was to facilitatepine regeneration from naturalseedfall, data from the study wereused as a standard for comparisonwith data taken from the area wherelate-winter prescrihed burning wasdone in study B.

Study B-The purpose of thisstudy was to determine the effect ofoverstory pine basal area and cyclical

Fire Management Notes

Table i-Averege cortditions during prescribed burning ott January 3/, 1984

'Oetermmed from fuel-moisture slicks at midd<lY.2Fireline intensity (I) (5)was calculated from over 100 ocular estimates of flame length (Lr). in feel,obServed dUringtne prescribed wms.: \ = 5.67 l{27

IIr

~',

i•

burning intervals on establishmentand growth of pine regeneration inuneven-aged stands of loblolly­shortleaf pine. Before the study, thesestands contained over 100 square feetof pine basal area per acre. Prescribedwinter burns were first conducted onJanuary 14 and 15, 1981. In Augustand September 1981, all hardwoods Iinch in diameter at breast height(d.b.h.) and larger were stem injectedwith Tordon IOIR. Merchantablepines were cut in June and July 1982,removing an average of 7,000 boardfeet per acre (International Y4-inchrule). This left a balanced uneven­aged structure with specified basalarea averaging 72 square feet per acrein 110 trees per acre. Plots remainedundisturbed after the 1982 harvest un­til January 31, 1984, when a secondprescribed burn was conducted (tableI). Pine seedling establishment fol­lowing the second late-winter bum isthe subject of this investigation. Seed­ling data were taken within 2.5-acregross plots where residual basal areaaveraged 78 square feet per acre inthe fall of 1984.

An estimate of pine seed produc­tion was obtained from 40 seed col­lection traps (llzo milacre each). Thesetraps were on all 16 plots in study Aand on 24 plots at 5 other locations inuneven-aged pine stands on theCrossett Experimental Forest. Seedcounts were taken once a week fromOctober 1983 through February 1984(22 weeks). All collected pine seedswere cut open, and those containingfully grown, firm, undamaged tissue

Volume 47. Number 2

Variable

Precipitation:Time since last accumulationAmount

Air temperature during burnRelative humidity during bumWind:

DirectionVelocity

Time 01 dayType of burnFine fuel moisture1

Mean fireline intensity"Range in lireline intensityEstimated ground coverage by fire

were judged as potentially viable (3).In October and November of 1984,

pine seedling counts were made onegrowing season after seedling estab­lishment in both study A and study B.Nine O.3-milacre circular quadratswere systematically established withinthe interior 0.1 acre on each of eight0.25-acre gross plots in study A andon each of six 2.5-acre gross plots instudy B. The number of pine seedlingwithin each quadrat was counted andtotal height of every fifth seedlingwas measured to the nearest O. I foot.Although older pine seedlings couldhave been found on plots in study B,only first-year pine seedlings were en­countered. The prescribed burn inJanuary 1984 had apparently elimi­nated all previously established pineseedlings.

A r-test for unpaired plots betweenstudy A (8 plots) and study B (6

Measurement

8 days1.22 inches34toSO°F70 to 24 percent

From the south3 miles per hour10 am to 4 prn CSTBack and flank tires12 percent92 Btu/ft-sec47 to 127 Btu/tt-sec83 percent

plots) was used for comparing meandensity, percent stocking, and totalheight of first-year pine seedlings, aswell as overstory pine basal area. Allstatistical tests were conducted at the0.05 level of significance. Arc sinetransformation was used in analysis ofpercent stocking.

Seed Crop

Total cumulative pine seedfall be­tween October I, 1983, and March I,1984, averaged over t, 170,000 seedsper acre with 81 percent consideredsound. According to Baker andBalmer (2), a good seed crop for lob­lolly pine is more than 80,000 soundseeds per acre, and an average crop isbetween 30,000 and 80,000 soundseeds per acre. Judging by thesestandards, the 1983-84 seedfall was abumper crop. Based on the 22-week

37

Table 2--Comparison of pine seedling establishment on seedbeds prepared before seedfall (A) andon seedbeds prepared after the bulk of seed had fallen (B)

lA = Hardwood control + logging; B = Late-winter burn.2Colurnnar means followed by the same letter are not significantlydifferent at the 0.05 level."eesmateo from 1982 postharvest inventory (9).

First-year pineseedlingsSeedbed 1984 merchantable pine

treatment' Density Stocking Total height basal area

Stems/acre Percent Feet Square feet/acre

A 62,979 a' 100 a 0,46 a 70 aB 20,607 b 89 b 0.36 b 78 b'

collection period, pine seedfall wasmore than 75 percent complete by De­cember I and 90 percent complete bythe time of the January 31 prescribedburn.

Even though seedfall was nearlycomplete before burning in study B,the bumper seed year provided suffi­cient seeds after burning for ensuringnatural pine regeneration. Ten percentof the sound seed crop, or nearly95,000 sound seeds per acre, was col­lected in traps after January 31. Thus,in this bumper seed year, 10 percentof the crop was equal to an entire seedcrop in an average seed year. Also,some additional seeding probablyoccurred after the March I termina­tion date for seed collection.

Seedling Establishment

As would be expected, study areaA, which had been site prepared fornatural pine seeding prior to seedfall,had substantially more pine seedlingsafter I year when compared withstudy area B, which had received onlypartial seeding following the late­winter prescribed bum. First-yearseedling density in study A, wherethere was a complete seed catch in thewinter of 1983-84, averaged nearly63,000 seedlings per acre. This figurewas significantly higher than thenearly 21,000 seedlings per acre instudy B with the partial seed catch(table 2). Stocking of first-year pineseedlings in study A averaged 100percent, again significantly higherthan the 89 percent stocking in study

38

B (table 2).Total height of first-year seedlings

in study A averaged 0.46 feet, signifi­cantly taller than the first-year seed­lings of study B, which averaged 0.36feet (table 2). The additional 8 squarefeet of basal area in the overstorypines and higher stem density of mer­chantable pines in study B resulted inmore shade and subsequently lessgrowth of pine seedlings, Brender andBarber (4) reported that the level ofpine overwood shade, measured bythe height to live crown, ranks withoverwood density as a major factoraffecting the growth rate of loblollypine seedlings and that survival andgrowth of understory pine seedlingsare poorest under low overwoodshade. Crown heights of the old­growth overwood in study A weretaller than those of the uneven-agedoverwood in study B.

Appreciably fewer pine seeds wererequired to produce a seedling instudy B (5 seeds per seedling) com­pared to study A (15 seeds per seed-

ling). Seeds that lay on the groundfrom early October until germinationthe following spring were available toany number of predators such as in­sects, birds, and woodlot mammals,Pine seeds that fell to the ground fol­lowing the late-winter burn wereavailable to predators for a muchshorter period of time before springgermination. There are two possiblesources of pine seedlings that arise onburned areas when burning is done af­ter the bulk of seed has fallen. Ac­cording to Chaiken (6), either newseeds are disseminated after the fire orseeds that had lodged in sheltered orprotected areas on the ground wherethey were not destroyed by fire werestill viable.

Density and StockingRecommendations

In uneven-aged management ofloblolly-shortleaf pine, the recom­mended density for the submerchan­table stand (<;;4-inch d.b.h. classes) is500 to 700 pines per acre (10). On

Fire Management Notes

,II

!' clear-cut areas in even-aged manage­ment, it is desirable to have between1,500 and 2,500 loblolly-shortleafseedlings per acre the first year afterestablishment from natural seedfall(11). Percent stocking is another crite­rion for judging the success or failureof natural pine regeneration. Accord­ing to Trousdell (15), areas that con­tain 90 percent stocking of loblollypine seedlings from natural seedfallare considered well stocked.

In the investigation, after the late­winter prescribed bums in uneven­aged stands of loblolly-shortleaf pinesduring a bumper seed year, sufficientseeds were available to result in ahigh-density, well-stocked crop ofpine seedlings. Although exact perio­dicity of bumper seed crops cannot bepredicted, better than average orabundant seed crops of loblolly pinecan occur at 2- to 4-year intervals (13,17).

Summary

When managing for naturalloblolly-shortleaf pine regenerationusing seed tree, shelterwood, or selec­tion (uneven-aged) cutting methods,timing of seedbed preparation can bea critical consideration. Late-winterprescribed burning for seedbed prepa­ration may result in an insufficientsupply of viable seeds for regenera­tion purposes if done in poor to aver­age seed years. For example, in this

Volume 47, Number 2

investigation, percent stocking anddensity of loblolly-shortleaf pineseedlings established following a late­winter prescribed bum were signifi­cantly less than stocking and densityof seedlings on areas undisturbed dur­ing seedfall. Even so, a bumper pineseed crop that coincided with the late­winter bum resulted in a pine seedlingstand that met or exceeded publishedrecommendations for both adequatestocking and density.

These data suggest that land man­agers who rely on prescribed burningto prepare seedbeds in advance of nat­ural pine seeding in the South canpossibly extend the burning seasonthrough January during bumper seedyears and still obtain adequate pineseed catch for regeneration purposes.

•Literature Cited

1. Allen, Peter H.; Trousdell, Kenneth B.Loblolly pine seed production in theVirginia-North Carolina coastal plain. Jour­nal of Forestry. 59: 187-190; 1961.

2. Baker. James B.; Balmer, William E. Lob­lolly pine. In: Silvicultural systems for themajor forest types of the United States.Agric. Handb. 445. Washington, DC: U.S.Department of Agriculture; 1983: 148-152.

3. Bonner, F.T. Seed testing. In: Seeds ofwoody plants in the United States. Agric.Handb. 450. Washington, DC: U.S. De­partment of Agriculture; 1974: 136--152.

4. Brender, E.V.; Barber, John C. Influenceof loblolly pine overwood on advance re­production. Sta. Pap. 62. Asheville, NC:U.S. Department of Agriculture, ForestService, Southeastern Forest ExperimentStation; 1956. 12 p.

5. Byram, George M. Combustion of forestfuels. In: Davis, Kenneth P., ed. Forest

fire--control and use. New York: McGraw­Hill Book co.. 1959: 61-89.

6. Chaiken, L.E. Extent of loss of loblollypine seed in winter fires. Res. Note 21.Asheville, NC: U.S. Department of Agri­culture, Forest Service, Southeastern ForestExperiment Station; 1952. 2 p.

7. Crow, A. Bigler; Shilling, Charles L. Useof prescribed burning to enhance southernpine timber production. Southern Journal ofApplied Forestry. 4(1): 15-18; 1980.

8. Davis, Kenneth P., ed. Forest fire-c-coruroland use. New York: McGraw-Hill BookCo.; t959. 584 p.

9. Farrar, Robert M., Jr. Densitycontrol-natural stands. In: Karr, Bob L.;Baker, James B.; Monaghan, Tom, eds.Proceedings of the symposium on the lob­lolly pine ecosystem (west region); 1984March 20-22; Jackson, MS. MississippiState, MS: Mississippi Cooperative Exten­sion Service; 1984: 129-154.

10. Farrar, Robert M., Jr.; Murphy, Paul A.;Willett, R. Larry. Tables for estimatinggrowth and yield of uneven-aged stands ofloblolly-shortleaf pine on average sites inthe west gulf area. Bull. 874. Fayetteville,AR: Division of Agriculture, University ofArkansas, Arkansas Agricultural Experi­ment Station; 1984. 21 p.

11. Grano, Charles X. Growing loblolly andshortleaf pine in the midsouth. Farm. Bull.2102. Washington, DC: U.S. Departmentof Agriculture; 1967. 27 p.

12. Grano, Charles X. Conditioning Ioessialsoils for natural loblolly and shortleaf pineseeding. Res. Note SO-116. New Orleans,LA: U.S. Department of Agriculture, For­est Service, Southern Forest ExperimentStation; 1971. 4 p.

13. Jemison, George M.; Korstian, c.F. Lob­lolly pine seed production and dispersal.Journal of Forestry. 42: 734-741; 1944.

14. Lotti, Thomas; Klawitter, Ralph A.;LeGrande, w.e. Prescribed burning forunderstory control in loblolly pine stands ofthe Coastal Plain. Sta. Pap. 116. Asheville,NC: U.S. Department of Agriculture, For­est Service, Southeastern Forest Experi­ment Station; 1960. 19 p.

15. Trousdell, Kenneth B. Favorable seedbedconditions for loblolly pine disappear 3years after logging. Journal of Forestry. 52:174-176; 1954.

16. Van Lear, David H. Natural regeneration ofsouthern pines. In: Mann, John W., ed.Proceedings of a seminar on site prepara­tion and regeneration management; 1980November 18-20; Long Beach, MS. LongBeach, MS: Forestry and Harvesting Train­ing Center and Clemson University; 1980:64-72.

17. Wahlenberg, W.G. Loblolly pine: its use,ecology, regeneration, protection, growth,and management. Durham, NC: Duke Uni­versity School of Forestry; 1960. 603 p.

39

United StatesDepartment of Agriculture

Washington, D.C.20250

OFFICIAL BUSINESSPenalty for Private Use, $300

III II IU.S.MAlL

(0)

Postage and Fees Paid

United StatesDepartment of Agriculture

AGR·l01

. '. i

ORDER FORM To: Superintendent 01 Documents, U.S. Government Printing Office, Washington, D.C. 20402

ECredit Card Ord,ra Only

Enclosed is S 0 check, VISA" Total Chari/esS Fill in the boxes below.o money order, or charge to my ~

Deposit Account No. Credit') ,-,-,-,-.,.,.,.,.,.,.,.,.,.,,---,---,--,

ITITITIJ-D [EEl Card No.Expiration Date ITIIJ

Order No. Month/Year

Please enter my subscription to Fire Management Notes at $13.00 domestic, $16.25 foreign.(Prices are eubtect to change without nolice.)Make checks payable to Superintendent of Documents.

Company Of personal name

For Office Use Only.

Quantity Charges

EnclosedTo be mailed

Subscriptions

Postage

Foreign handling

MMOBOPNR

UPNS

Discount

Refund