This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. HORNBY'S PRINCIPLES OF FIRE CONTROL PLANNING BY H. T. GISBORNE • NorthernRockyMountainForestand RangeExperimentStation On August 27, 1937, Lloyd G. Hornby died o.f heart failure on the Toboggan Creek forest fire in the Clearwater National Forest. Few if any men in or out of the U.S. Forest Servicehave made a greater contribution t• fire .control plannin,g than did he. In the following article, H. T. Gisborneoutlines the principles of fire control planning developedby Mr. Hornby, emphasizingthe major features which have too often been obscuredby the detailed technical proce- dures of the componentprocesses of fuel type classification, seen area mapping,smokechaser coverage,transportationplanning, etc. fire makesperimeter. chargeof fire controlplanningresearch 2. Fuel type classificationis necessaryto at the Northern Rocky Mountain For- showthe two basicfactors:rate of spreadand est and RangeExperimentStation,he brought resistance to control. to the taska background of training,experience, 3. Plans must be made for first-attack conto spot or crown and inherentability whichwasuniqueas it was trol before the fire commences well suitedto the opportunities of this work. exceptunderclass6 or class7 dangerin the His training in both engineeringand forestry, extremefuel types. of fires, and values his fifteenyearsof field experience from smoke- 4. Fuel type, occurrence chaserand ranger to supervisorof three na- at stake must be coordinated for most economitional forests,and his exceptionalingenuity, cal yet adequatefire control. 5. Lookouts, firemen, and crews have the practicallyassureda researchthat would make a major contribution to forest fire control. dual responsibilityof detectionand smokeIt is universallyconcededthat his Fire Con.trol chasing. Planr•ingin the Northern Rocky Mountain Re-' 6. Transportationand communicationplangion has revitalizedand very nearly revolution- ning shouldfollow and be basedupon fire conized fire control planning processes through- trol planningand other forest-use requirements out the United States. for multiple use. It was Hornby'splan, beforehe died on the 7. Fire controlplanningfor the "worstfirst" fire line in August 1937, to summarize automaticallysimplifiesthe process. 8. The conditionscreating a fire problem the principlesbasicto his work, but exceptfor a three-pagelong-handdraft of a memorandum are not static. Fire control planningis therefound in his files after his death, this sum- fore a continualprocessof revisionand refine- 'HEN Lloyd G. Hornby was placed in mary was never written. On the basis of this men t. memorandum,his various reports, and six years of almost daily contactand work with him, the writer here attemptsto outline those principlesas accuratelyas one personcan attempt to presentmaterial for another person. The analogiesused in several casesto emphasizethe similaritybetween fire controlplanning and engineeringpracticesare Hornby's own,whichhe employedfrequently. The eight major principleswhich,with one exception,are apparent in his publishedreport can be briefly expressedas follows: The first, and perhapsmost basic, principle developedby this researchwas that successful fire control requires"held line to be built faster than the fire makes perimeter." This principle had been subconsciously and dimly recognizedin the past, but it has been within the past few years only that the old method of estimatingthe job of line building on the basisof acresof fire was replacedby estimates basedon perimeter. This first principle recognizes the superiority of the perimeter basis, 1. Held line must be built faster than the and in addition considers the fact that fuel type is the major determinantof the rate of spread or rate of perimeter increase, which •Acknowledgmentis ma.de to Clayton S. Crocker mustbe assumed for eachareain planningnorand ClarenceB. Sutliff, who also ha,d worked closely mal controlmeasures for it. Omitting proper with Mr. Hornby, for reviewing this article and for of fuel typesin fire control plantheir suggestions for improvingit.saccuracyand scope. consideration 292 HORNBY'S PRINCIPLES OF FIRE CONTROL PLANNING 293 beingequal." ning is analogousto an engineer planning to requiredit, otherconsiderations construct a massivebuilding without examinOne impedimentto the understandingand ing the rock or quicksandon whichit is to rest. properuseof fuel type mapsis the fact that "Proper consideration"may vary, of course, from broad generalizationsin regions of uniform fuel types to detailed measurements, surveys, and maps in regions of great variation in rate of spreadof fire. The secondprinciple, which logically follows, is to measure and rate this basic factor, individualfires often spreadat ratesdifferent from that indicatedby the map. It is obvious, however, that even the most carefully identified and mapped rate of spread,at a certain, averagebad level of burningseverity,will not be reached under favorable weather conditions and will be exceeded when the weather is worse fuel type. To do this Hornby originatedthe than averagebad. No single map could be dual basis, physical classificationof fuel •nade,and kept simple, which would showall types including rate of spread and resistance rates of spread for all weather conditions. to control. His methodsof evaluating these The major purposeof this work is not to proare describedin "Fire Control Planningin the ducea fuel type map whichcanbe usedfor deNorthern Rocky Mountain Region," and "Fuel terminingthe manpowerto sendto individual Type Mapping in Region One."•He assumed fires, under any combinationof weathercon"burningseverity,"classof fire danger,or fuel ditions. The chief purpose is to determine condition,that shouldbe usedin fuel type map- the normal high-rate of perimeter increase ping to be "averagebad," leavingits precisedefi- which can be expectedso often or so consistnition to be determinedlater by actualmeasure- ently that it should be planned for. This ment. Here he followedthe practiceof an en- normal, high-ratevariessignificantlyaccording gineer who would estimatehis costsof making to fuel type. This principle thereforeholds excavations in clay on the basisof the clay that it is better engineering,better forestry, beingwet, and thereforemoredifficultto handle better science,to identify such a factor conthan if it were dry. The precisedefinition of sciously and accordingto certain standards and without uni"wet" is of lessimportancethan would be the rather than subconsciously complete failure to consider this condition. versal standards. The measurement rather than the approxima- This illustrates the researchquality of this tion of eachfuel type is one phaseof this work work. The objective is the evolution of a ß requiring much additionalresearch. Hornby methodwhichcan be usedsimilarlyby all men, was engagedin suchwork the day he died. in preferenceto "the judgmentof someperson A brief statementof Hornby'sconceptionof or persons with or without going to the fuel types and their basicuse was includedin trouble" of identifyingand evaluatingall the his final memorandum:"When a particular factors. It has been consistentlyrealized that level of burningseverityprevailsover a large the presentevaluationsof fuel type can be iraarea, the same speed and strengthof attack proved,with time and additionalresearch,but is not requiredfor everykind of fuel" . .... fire controlplanningwithoutfuel type classifi"One way to station manpowerover an area cation is, to use another analogy, like estiis to acceptthe judgmentof somepersonor mating the cost of road or railroad construcpersons,with or without going to the trouble tion from a topographicmap alone with no of comparing fuels and mapping their loca- k•owledge of the soils, clay banks, or rock tions. Anotherway is to map the fuels,set up ledgesthroughwhichthe cutsmustbe made. •ninimumstandardsof speedand strengthof The third principle is that control of a fire attack accordingto severityof burning condi- (in Region One) by the first attackingforces tions,and stationmanpowerto satisfythesere- shouldbe plannedfor "beforethe fire beginsto quirements. Under the latter procedurethe spot or crown." Other criteria suchas "consame class of fuels would receive the same trol before 10 a.m. of the next burning petreatment (degree of fire control) wherever riod," and the "percentageof area burned," it existedand wheneverthe burningseverity by timber types as recommended by the regional foresters, Washington conferenceof altornby, L. G. Fuel type mapping in Region One. Jour. Forestry 33:67-72. 1935. 1930, have been used, but for the reduction and gradual elimination of those large fires 294 JOURNAL OF FORESTRY which are chieflyresponsiblefor overburn,a the need of attemptingto apply this fourth more basically natural criterion is needed. principle, including values as well as fuel Hornby insisted that "after the crown fire typesand the occurrence rate. The fact that valid inferences cannot be stage is reached, no assuranceof control at any particular size can be given," and under drawn unlessthe measurements or ratings of certain conditionsthis stage may be reached all the basicfactorsare integratedled Hornby before 10 a.m. of the next burning period. to his searchfor a "total dangerrating" which He also pointed out that the analysisof the would integrate fuel types, occurrence,and recordsof 12,056 fires in Region One shows values. He was nevercompletelysatisfiedwith that over a periodof yearssilviculturallyade- the formula that he originated. By this forquate protection,the basisof the Washington mula there is obtained from the fuel type conference standards,can be providedwithout maps an estimateof the acres per hour that 99.99 percent control of all fires in the first will probably burn during the first few hours or even the secondwork period. after origin on an average bad day. This, The likelihood of spotting or crowning multipliedby the damage,in dollars,that would varies,of course,both with fuel type and with be done, is called "damageper hour." Then burningseverity. The presentprinciples,how- damageper hour plus expenseper hour (based ever, do not include plans to meet the worst on the fuel type classificationof resistanceto possibleconditions of eitherfuel type or class control) gave total probable"cost per hour" of fire dangerin the first attackany morethan for any unit of area. And costper hour mulan engineerbuildsto withstandthe worstpos- tiplied by the occurrenceindex for that area sible earthquakeor the worst possibleflood. gives a figure approximating"total danger." Engineers knowthat the damagein suchcases Such numericalratingsfor differentareascan is often less than the cost of completepro- be compared and the relative needs for fire tection. This third principle of "control be- control facilities and manpowerthereby deterfore crowning"therefore excludesthose rare mined, with less than previous dependence firesoriginating in extremefuel typeson class upon the personaljudgment,knowledge,lack 6 or 7 days, when fuel moistures are under of knowledge,or bias of the individual. 5 percent,relativehumidityunder 15 percent, This attempt to obtain numerical ratings is and winds19 m.p.h. or more. The damagesadditional evidenceof the scientificquality of fromsuchexplosive firescanmosteconomicallythis research. It is emphasizedby Lord Kelbe minimizedby large crew suppression ac- vin's statement,"that when you can measure tion, the same as for other fires which for other what you are speakingabout,and expressit in reasonsescapecontrolby a reasonably econom- numbers,you know somethingabout it. But ical first attack. when you cannotexpressit in numbersyour The fourth principleis that fuel types,oc- knowledgeis of a meager and unsatisfactory currence of fires, and values at stake must kind. It may be the beginningof knowledge, be coordinatedby the processes of planning but you have scarcelyin your thoughtsadif the resultantpermanent facilities(buildings, vancedto the stageof a science." roads,and telephonelines) and the temporary The impossibilityof immediatelyevaluating manpower(contactmen, lookouts,smokechasers,all of the factorsof "total danger"with conand small crews), are to be held to an econom- sistent dependability is obvious. But here ical minimum,yet adequate protection is to be again, as in fuel type mapping,it is evident providedduringseasons of criticaldanger.The that carefully prepared estimatescan now be use of this principle is not as apparent in made, with a sufficientlylow error to give a Hornby'smajor reportsas are the otherprin- better basisfor fire control planning,more ciples. This is beca/lseit has not beenex- "independentof the idiosyncrasies of the inpedientto date to attemptto place dollar dividual." The fifth principle which is contrary to values'on eachof the diversifiedusesor purposesof forestareasincludingstreamflowregu- l•racticesin some sectionsof the country is lation, soil conservation, wildlife production,the "dual responsibility"of lookouts,firemen, recreation,live-stock,and timber products.Fire and crews. This had beenappliedwith graducontrolplannerscannotescape,however,from ally increasingintensityin Region One begin- ttORNBY'S PRINCIPLES OF FIRE CONTROL PLANNING 295 ning about 1920. As national inspector of large crewcoverage(providedmerelyby roads fire control,Major E. W. Kelley gave marked from outsidesourcesof labor) is planned for impetusto the practicein 1927 to 1929. By all areas of fuels in which fire could be exthis principle lookout men are not only to pected to escape any reasonablyeconomical chase smoke when that can be done without first attack, as mentioned in discussingthe impairing the detectioncoverage,but look- third principle. Naturally, thesemost difficult out station locations are to be selected, other fires are most likely and should be expected things being equal, so that thesemen can serve most frequentlyin the worst fuel types. The this dual purpose. It would appear to be de- actiondemanded,undersuchconditions,is supcidedly uneconomicalto build a lookout sys- pressionby large crews, and roads must be tem capableof first discovering99.99 percent available reasonablynear such areas of worst of all fires when firemen, trail crews, and fuels to expeditearrival of these large supseveralother units of the fire control organiza- pressioncrews. If these roads are planned tion can be dependedupon for considerable first, later planningof roadsfor use by small crews, smokechasers, and firemen-lookouts,will detectioncoverage. Closely related to this fifth principle of automaticallybenefit. By the reverseprocess "dual responsibility" is the sixth, or "multiple roadsare first plannedfor the lookouts,smokeuse of roads." Althoughnot generallyrecog- chasers,and patrolmen,and are often found to nized at the time Hornby was developing be unsatisfactoryfor expeditingthe arrival of his principles, it is now standard practice large crews at the areas of worst fuels where to plan for the lower cost, to each line of the catastrophicfires are most likely to origwork, and the greater efficiencyof a transpor- inate. As the reductionof the sizeof large fires tation systemdesignedby correlatingthe re- is the presentmajor requisiteto satisfactory quirements for (1) fire control, (2) utiliza- fire control, logical planning must be for the tion or removal of forestproducts,(3) recrea- "worst first." tion and (4) general administration. By this The eighth, and final, principle applied principle, transportationplanning is a distinct- throughoutall phasesof fire control planning ly accessoryprocesswhich should follow and in Region One is that none of the basic facbe basedupon the previousdeterminationof tors of the fire problem--fueltype, occurrence, manpowerplacementadequatefor fuel types or values at stake--is constant or static. Each as well as for other forestrequirements. Com- of theseis known to change,perhapsslowly, municationplanning similarly becomesa sub. with time. The rating of the relative weight sequentprocess;it is evendependentupon, or of each factor therefore undoubtedly will at least affectedby, the resultsof transporta- change, slowly but steadily with increased tion planning. Transportationor communi- knowledge,and perhaps radically for short cationplanningwithoutcompleteconsiderationperiods accordingto the dominanceof some of fuel typesand their manpowerrequirements, individual's philosophy. Preliminary plans or without correlatingutilization, recreation, thereforeshould deal with averages,with ingrazing, and other needsis analogousto dam dicative trends, with combinations of factors, construction without streamflow measurements. always endeavoringto prevent single and exSuch dams have been built--and rebuilt, after ceptionalcasesfrom resultingin over-'building the high-waterthat shouldhave been foreseen. or completelyneglecting any essentialphaseof The seventh of these principles for fire the fire control organization. Likewise, this controlplanningis to plan first for the worst, principle shouldtemperall finality of concluor for the "worst first." By "worst" is meant sionand shouldpredicatecontinualrevisionand the mostdangerous fuel types. The plansfor refinementof the plans. 10 national forests comprising 17 million Hornby's contention,supported by ample acresof land in RegionOne have repeatedly evidencein RegionOne, that a strongand apdemonstrated that if provisionsare made first parentlycostly first attack is usually the most for the worstfuel typesa large proportionof economicalin the most dangerousfuel types, the easier types is automaticallygiven satis- can hardly be called a principle. It is, howfactory, or even superlativedetection,smoke- ever, a generalitysupportinghis worst first chaser,and crewcoverage.By this principle principle,and onewhichis basicto the proper 296 JOURNAL OF FORESTRY financingof adequatefire control in this re- forest fire control forces, and had seen the deleterious gion. effects of the short-term limitation The personnelfactor was recognizedby on so many other forests,that he was occaHornby as vital. He stressedthe requirement sionally pessimisticconcerningthe successful of both adequatenumbersof men and ade- applicationof his or anyone else's fire conquate training. His last memorandumstates: trol plans. He knew that the best of plans, "Errors (on the fire line) are so numerous basedon specificidentification of all significant that their elimination constitutes an outstand- factors, each well evaluated, and all consist- ing problem,"and, "Ratesof fire suppressionently integrated,could not function most efta. work, per man per hour, are so slow that ciently unless applied by a competentperunwieldynumbersof men are required even sonnel. He knew that any lack of policy or on firesof moderate size." Thesephases of fire financial limitation, which interfered with the controlplanning,and the preventionof fires, retention,year after year, of a trained and were recognizedby Hornby as calling for competentpersonnelcould prevent the full such distinctly different action as to be out- improvementof fire control efficiencythat sidethe scopeof his particularresearch. might be expectablefrom planningmethods The little pessimismthat Hornby felt con- that he or anyoneelse might devise. He concerningthe possibleprogressof fire control finedhis work,however,to developingmethods planningarosefrom the indeterminacy of the and a planning procedurewhich would propersonnel factor. He had had so muchper- vide any personnelwith facilitiesimpersonally sonal experiencein the loss of trained and suitedto local conditionsof fuel type, occurcompetentshort-termmembersof his national rence of fires, and values at stake. A DESIGN FOR EXPERIMENTS IN THINNING FOREST STANDS BY J. G. OSBORNE SouthernForest ExperimentStation• Thinning experime•nts have taken up _aconsistentlylarge shareof forest-research programs,and can be expectedto continueto do so. This paper presentsan experimentaldesign which is intendedto aid iu increasingthe amountof informationobtainedfrom such experiments. of thinning employed,and the kind of final ANY ofthe early e:•periments infor- estry were madeto determinethe effects of various methodsof thinning on tree growthand on the yield per acre in variousforestproducts. Sincethesestudies have continuedto occupya prominentposition in forest-research programs,and since, as stated by Yates (3), it is profita'bleto expendconsiderablework in developinga techniquein a field in whichmuchexperimental work is likely stahdsought varywidely,but certainprinciples of experimentation are common to all thinning research.Effortstowardsdeveloping an ade- quate experimentaldesign,therefore,must be madein accordwith at leastfour basicrequirements,whichbriefly statedare: (1) That adequateprovision be made to assurethe reliability of the resultsobtained; (2) That the experiment be efficient,which to be undertaken,it has been consideredworth meansthat early and accurateinformationbe while to devote at least some research to imobtained, andthatthenumberof plotsandtotal provingthe techniqueof thinningexperiments.area necessary to providea givenbody of inOBJECTIVESOF A SATISFACTORY TECHNIQUE formation be at a minimum; (3) That the experiment shouldpermitthe The typeof standto be thinned,the method inductionof generalprinciplesof response, •The author gratefully acknowledgesthe assistance from which the probableresultsof thinning and ,helpful suggestionsgiven by R. A. Chapman, It. underotherconditions may be deduced; Bull, and A. L. MacKinney of the Southern Forest Experiment Station, and by F. X. Schumacherformerly of the U.S. Forest Service, Washington, D.C. (4) That, in order to obtain assuredreliabil- ity of the results,the limitationsas to original