Fuel Moisture, Seasonal Severity and Fire Growth Analysis ...

1

FuelMoisture,SeasonalSeverityandFireGrowthAnalysisintheUSFireBehaviorAnalysisTools:UsingFireWeatherIndex(FWI)CodesandIndicesasGuidesinAlaska

1 IntroductionEffortstoconduct,interpret,andapplyfindingsfromfiregrowthanalysisusingtheWildlandFireDecisionSupportSystem(WFDSS)andInteragencyFuelTreatmentDecisionSupportSystem(IFTDSS)toolsareheavilydependentonweatherobservationsandforecastsfromlocalweatherstationsandlandscapefuelclassificationsfromLANDFIRE.

Additionally,analystsapplyaconsiderablenumberofsubjectiveinputstotheiranalyses,suchasInitialFuelMoisturevaluesforliveanddeadfuels,bestweatherstationtouseforwindandfuelmoistureassessments,crownfirepotentialandmanifestation,andspottingfrequency.

Thetypicalapproachutilizedbyanalystswheninitializingtheirfirstanalysesistousedefaultinputsasmuchaspossibleand“calibrate”themodeltoknowfiregrowthevents.Thismethodcanbetimeconsuming,assumesthatthefirehasalreadyexperiencedoneormoresignificantgrowthevents,andsometimesleadsanalyststoadjustfactorsthatmaynotberesponsibleforchangesobservedontheground.

ThisguideoffersrecommendationsforusingCanadianForestFireDangerRatingSystem(CFFDRS)fuelmoisturecodesandfirebehaviorindicesfromtheFireWeatherIndex(FWI)systemtoprovideobjectiveguidanceforinitialsettingsformanyoftheseanalysisinputs.TheFWIsystemhasbeenformallycalibratedfornorthernborealecosystemsandeffectivelyidentifiessignificantthresholdsfortheAlaskalandscapesaswellasimportanttrendsinchangingfiregrowthpotential.

TheprimarytoolsconsideredhereincludeWFDSSandIFTDSSanalyses.IncludedareShort-termFireBehavior(STFB)thatisbasedontheFLAMMAPfiregrowthmodelingsystem,Near-TermFireBehavior(NTFB)basedontheFARSITEfiregrowthmodelingsystem,andFireSpreadProbability(FSPro)basedonFLAMMAPandNFDRSinputsusingFireFamilyPluswithinWFDSS.IFTDSSusesprimarilyFLAMMAPtoolsforitsfiregrowthanalyses.

Allanalysesusefuelmoisturescenariosincluding1hr,10hr,100hr,Woody,andHerbaceousfuelmoistures.Analystsareencouragedtoeditthesesettingsingeneral,orforspecificfuelclasses.FSProutilizeswindclimatologyfromaselectedweatherobservinglocationandallowstheusertomakebothcoarseandfineadjustmentstothatdistribution.FSProisheavilydependentontheEnergyReleaseComponentforfuelmodelG(ERCg)toidentifydailyfuelmoistureandspottingscenariosforbothdeterministic(forecast)andprobabilistic(climatology)portionsoftheanalysis.AnalystsarefindingthattheyneedtoedittheERCclassesandstreamsheavilytoreflectexpectedconditions.Attheveryleast,thesedailyFWIfuelmoisturecodesandfirebehaviorindicesareausefulcross-referenceswhenconsideringanalysisinputsandoutputs.

Therearetwosectionsthatfollow.

• ThefirstisadiscussionoftheFWIfuelmoisturecodes,theirfuelmoistureequivalents,andhowtheycanbeusedtofacilitateeditstofuelmoisturescenariossothattheyreflectcurrentobservedconditions.

• ThesecondshowshowBuildupIndex(BUI)andFineFuelMoistureCode(FFMC)canbeusedtoinformERCClassTablesandStreamstoreflectcurrentseasonseverityandfacilitatelocal“burndays”climatologytotheanalysis.

2

2 FuelMoistureInputsandFWIFuelMoistureCodes2.1 FineDead(1-hour)FuelMoistures

Whilelivefuelmoistures(WoodyandHerbaceous)havelargeimpactsonthefirespreadmodels,theyarefixedoverthedurationofbothWFDSSandIFTDSSanalyses.Themostvariablefuelmoistureinputisthe1-hourfuelmoisture.WFDSSSTFBandNTFBusehourlyweatherdataaswellasslope,aspect,andshadingfactorsto“condition”1-hour(and10-hour)fuelmoisturevaluesfrominitialsettings.

Thisdiurnalplotoffinedeadfuelmoistureillustratestheeffectofhourlyweather.IncludedaretheoriginalFosberg(1971)model(1h)inblue,the-Nelson(2000)model(1h)ingreen,andtheWotton(2009)GrassFuelMoisture(GFM%)inorange.Noticethatboth1h-NelsonandGFM%showgreaterresponsivenesstoovernightrecoveryandprecipitationevents.However,the1h-Nelsonestimatereflectsa2-4%increaseintheestimateduringthedryburnperiods.

AssumingtheGFM%estimateismorecompatiblewithexistingfirespreadmodelsandmoreresponsivetoday-to-dayvariationresponsibleforchangingfirespread,theanalystcouldconsideradjustingthe1-hourfuelmoistureestimatebasedononlineevaluationsorthistable.

KeepinmindthatNTFBusesconditioned1-&10-hourfuelmoisturesthroughouttheanalysis.ConsiderusingSTFBwherepossibleandsetting“conditioning”daysto0.

3

2.2 10-hourFuelMoisture

Whilethe10-hourFuelMoistureexertslessinfluenceontheRothermelFireSpreadmodeloutputs,itisestimatedforeachanalysisandisalsosubjecttotheinfluenceoftheNelsonDeadFuelMoisturemodel’stendencytoraisemoistureestimates.

Inaddition,theFWIFineFuelMoistureCode(FFMC)isverylikea10-hourtimelagfuelmoisture,estimatedbyFWIdevelopersassomewherebetweena5-and16-hourtimelag.Thoughproducedasa“unitless”code,itiseasilyconvertedtoafuelmoisture,representinganestimateofshadedlitterfuelsunderforestcanopy.Assuch,itassumesthatslope,aspect,andvariationinshadingislesssignificantthanthedryingeffectsoftemperatureandhumidity.

Infact,while1-hourestimateswerediscussedabove,finedeadlitterfuelsandfeathermossfuelbedsundertheborealforestcanopymayrespondtoweatherconditionsmuchmorelikeFFMCandmaybeappropriatelysetequaltothe10-hourestimatedescribedhere.

ThisgraphandtabledepicttherelationshipbetweentheestimatedFFMCandprospective10-hourfuelmoistureequivalents.Onthegraph,thebluepointsreflecttheformulausedintheFWIsystemtoconvertbetweencodeandfuelmoisturecontent(%).10-hourfuelmoisturesderivedinthiswayrepresentshadedforestlitterdeadfuelmoisture.Inthetableandonthegraph,inorangeistheconversionbetweenmeasurementsoffuelmoistureunshadedNFDRS“sticks”.

4

2.3 SlowlyRespondingFuelMoisturesWFDSSandIFTDSSanalysesbothincludeawiderangeoffuelmoistureestimatesthatrespondmoreslowlyoverweeksandmonthsduringthefireseason.100-hourand1000-hourfuelmoisturesrangefroma4-to40-daytimelaginheavierdeadfuels.Livefuelmoistures,herbaceousandwoody,areusedtoaccountfortheinfluenceoflushgreenvegetationasaheatsinkinthefireenvironment.

Currently,thereislittleobserveddatatoinformtheinputsforfuelmoistureandflammabilityconditionsfortheselivefuelsfoundinAlaska.Despitethis,manyanalystsusetheseinputsastheirprimarytoolincalibratingfiregrowthmodelsagainstobservedfirespread.

TheFWIDuffMoistureCode(DMC),a“unitless”indexofanassumedintermediate“timelag”fuelmoisture,takesadifferentapproach.Itintegratesfuelmoistureconditionsacrossthisbroadrangeofavailablefuelcharacteristics(otherthanfinedeadfuels)andrepresentsavailabilityandflammabilityinthoseclassesmoregenerally.Ithasbeencalibratedtodryinginthedufflayerbelowlitterontheforestfloor.

ThisgraphdepictstherelationshipbetweenthedailyestimateofDMCanditsequivalentdufffuelmoisture,inpercent.Further,itdepictsaconversiontofuelmoistureestimatesofanabove-grounddeadfuelofapproximately5”diameter.DMCestimatesareavailablefornearly200weatherobservinglocationsacrossAlaska.Theserepresentobjectivecharacterizationsthatcanbeusedtoadjustandapplyfuelmoistureinputsforanalysispurposes.

OnlytherelationshipbetweenDMCanditsequivalentdufffuelmoisture%hasbeenrigorouslyevaluated.Recommendationsforestimating100-hrandHerbaceousfuelmoisturecanbeappliedforanalysispurposes,butshouldbeevaluatedcritically.FeedbackconcerningthesemethodsshouldbedirectedtotheAlaskaWildfireCoordinatingGroup’s(AWFCG)FireModelingandAnalysisCommittee(FMAC)ortheAlaskaFireScienceConsortium(AFSC).

5

2.3.1 100-hourFuelMoistureInthegraphabove,DMCcanbeconvertedtoadufffuelmoistureequivalentthatrepresentsa360hourtimelagtrend.Thisshouldbeintermediatebetween100-hourand1000-hrtimelags.

Thisexampleseasonplotfrom2015attheHogatzaRAWSdemonstratestrendsforthesethreefuelmoistures.TheDMCEquivalent“duff”fuelmoisturewasrescaledtooverlaythe100-hourand1000-hourtrends.TheDMCmoisturetrendis,infact,intermediatebetweenthe100-and1000-hrtrends,representinga360hrtimelagfuelmoisturetrend.

Usingthe360-hourfuelmoistureestimatedfromtheDMCconversiongraphabove,the100-hourand1000-hrfuelmoisturescanbeestimatedasslightlylowerandhigher.Inthisexample,onJuly6th,the100-hrcouldbeadjustedtobemoreliketheDMC’s360-hrestimate,between4and5%.

2.3.2 HerbaceousFuelMoistureHerbaceousfuelmoisturehasbecomeacriticalinputforfiregrowthanalysisinWFDSSandIFTDSS.Butinsteadoffaithfullyobtainingandusingestimatesofmoisturecontent,thisinputisusedasacalibrationtoolforadjustmentoffirespreadestimatesinthoseanalyses.Itworksprincipallybytriggeringafuelloadtransferbetweenherbaceousloadsandfinedeadloadsformanyofthefuelmodelscurrentlyused.Transferredloadswouldthentakeonthe1-hrfuelmoistureestimate.However,alongwithwoodyfuelmoisture,theseloadsandtheirelevatedfuelmoisturesalsoimposeimportantheatsinksduringthegrowingseason,mutingsimulatedfirespreadwithinthemodels.

Despitelittleobservationdatatosupportinputvaluesinmanycases,herbaceousfuelmoistureestimatesusedinanalysescanhavealargeinfluenceonresults.Andoncethevalueisset,itsinfluenceisfixedforthedurationofthatanalysis.Assumingthattheherbaceousfuelmoisturewillremainfixedovera1-14-dayanalysisdurationmaybeaccurateorproblematic;wecannotbesure.Evenso,makinglargeadjustmentsinthisvaluetocalibratetoaknowngrowtheventmaynoteffectivelyrepresentthefactorsresponsibleforday-to-dayvariationinfuelavailabilityandflammabilitywithinthemodels.Modeledspreadcalibratedtoobservedfirespreadbasedprimarilyonsensitivitytolivefuelmoistureestimateswillproduceinconsistentresultswhentheassumptionsareappliedtoforecastconditions.

ThemethodsdescribedhereassumethattheDMCequivalentduffmoisture%isagoodproxyforgrowingseasonherbaceousfuelmoistureinputsresponsibleforguidingfuelloadtransfersandestimatingheatsinkfactorsforanalyses.ThiswouldallowanalyststoevaluatecurrentDMCvaluesinthefirearea,viewDMCforecasttrends,andobjectivelyapplyherbaceousfuelmoistureinputs.

0

40

80

120

160

200

240

280

320

360

0

5

10

15

20

25

30

35

40

DMCEq

uivalentFu

elM

oisture,%

Dead

FuelM

oisture,%

DMCEquivalent&DeadFuelMoisture2015Hogatza(AK)RAWS

100h 1000h DMC%

6

UsingtheDMCtofuelmoistureequivalentconversiononpage4,currentand/orforecastDMCsinthefireareacanbeconvertedtoDMCequivalentfuelmoistureforusedirectly(orasaguide)fortheherbaceousfuelmoistureinputtothefiresimulationanalysis.

Intheexampleplotshownhere,DMCequivalentfuelmoisturesbasedonweatherinputsfromtheHogatzaRAWSin2015arecomparedtoLFI-basedand1000-hrbasedherbaceousfuelmoistureestimates.

First,DMCequivalentfuelmoisturescannotbeusedtoestimatepre-greenandcuring/curedstatesinthefall.Theseareasareshadedoutonthegraph.Inthosecases,estimatesofherbaceousfuelmoistureshouldreflectcuring/curedconditions.

FromJune8ththroughJuly6th,DMCequivalentfuelmoistureestimatesfellfromahighof200%to45%.Estimates,suggestedonthegraphrangingfrom45%to75%duringthe15daysbeginningJune22nd,wouldimposesignificantfuelloadtransfersandenhancefirebehaviorpredictionspreciselywhendryfuelconditionswassupportingextremefiregrowthevents.ThroughthemiddleofJuly,therewasalullinsignificantfireeventsinthisareaandDMCfuelmoistureestimateswerebetween90%and130%,reducingandeliminatingfuelloadtransfersandincreasingtheheatsinklivefuelsprovide.ForseveraldaysinearlyAugust,firesintheHughesareabecameactiveandmadeseveralsignificantruns.DMCfuelmoistureestimatesduringthisperiodwouldhavebeenbetween50%and60%.

NoneofthissuggestthatthisisthephenologicaltrendofmoisturecontentinherbaceousfuelsduringthegrowingseasoninAlaska.Butitwouldbedifficulttoobtainsatisfactorysimulationsusingherbaceousfuelmoisturesbetween135%and240%asestimatedbytheLFIbasedmoisturemodel.Infact,mostanalystsheavilyeditthelivefuelmoisturesformostoftheiranalysesduringthegrowingseason.

2.4 FuelMoistureClimatologyforFSPro1-hr 10-hr 100-hr Herbaceous Woody

ERCgclimatologytendstomutetheobservedvariationinfinefuelmoisture.considerlowering1-hrintoptwobins,possiblyto3%or

4%.

Notalargefactorin

spreadmodel.ConsiderFFMCclimatologyasadefault(6-7%,8%,9%,12%,15%.

Again,generallysmallinfluence.

DMCclimatologysuggests

defaultsof6%,7%,8%,12%,and17%.

UsecurrentDMCestimateandforecast/outlooktosuggestrangeofDMCvaluesexpectedover

analysisperiod.UsefuelmoistureconversionandspreadrangeoverERC

classes.

ReviewNFMDrecordsforBlackSpruceneedle

moisture,generally<100%.Othersshrubsgenerallyhigherduring

growingseason.

Pre-Green Curing/CuredGrowingSeason

7

3 FireSeasonSeverityandFSProERCClassesandStreams3.1 EnergyReleaseComponent(ERC)andBuildupIndex(BUI)asSeasonIndicators

Totheleft,thesethreefiredangerratingsgraphsfromWFDSSdepictannualERCgclimatologyforobservinglocationsinthewestern,centralandeasterninterior.Alongwiththese,thesingleBUIseasongraphbelowshowsclimatologyforalltheinteriorwithmedianweeklyMODISdetectionsrepresentingareaburnedduringtheseason.Graydashedlineshighlightthedivisionsbetweenthefire“seasons”(Wind-Driven,DuffDriven,DroughtDriven,DiurnalStage)onallfourgraphs.TheprecisedatesofthedivisionsvaryfromseasontoseasonandarelessimportantthantheERCandBUItrendsthrougheachoftheseseasons.

TheMODISdetectionsconfirmthattheBUItrendscorrectlyrepresentthe“DuffDriven”and“DroughtDriven”seasonsaspeakseasons.Thereislesseroverallareaburnedinthe“WindDriven”and“DiurnalStage”shoulderseasons.

ThecorrespondingERCgseasonaltrendsunder-representseasonalpotentialforthe“DuffDriven”and“DroughtDriven”seasons,withtheaverage(gray)trendpeakingveryearlyinthe“WindDriven”shoulderseasonandshowingsteadydeclinethroughoutthepeakseasons.Thisskewedrepresentationofseasonaltrendistheresultofthefuelloadtransfersfromtheherbaceouscategorytofinedeadfuelsduringtheearly,pre-greenperiodandthelargeheatsinkprovidedbyelevatedherbaceousandwoodyfuelmoisturesduringthegrowingseason.Whilethismodeledheatsinkcharacterizationworkswellformanylandscapes,itinaccuratelydiminishespotentialduringthesepeakseasonsinnorthernconiferforests.

WindDriven

DuffDriven

DroughtDriven

DiurnalStage

8

ERCgperformsseveralcriticalfunctionsintheFSProanalysis.First,asadefault,itprovidesafrequencydistributionof5fuelmoistureandfirebehaviorscenariosbasedonitswholeseasonclimatology.Second,that climatologyprovidesday-to-daystreamsofthosefuelmoistureandfirebehaviorscenariostomodelfirespreadprobabilitiesweeksintothefuture.TheprocessexplainedbelowwilldemonstratehowknowledgeofobservedFWIelementscaninformadjustmentstoboththefrequencydistributionandtheERCgstreamsusedinthoseanalyses.

3.2 EditingtheERCStreamtoReflectFFMCandBUITrendsInFSProanalysis,theERCStreamisdisplayedasasequenceofdaysintherecentpastandtheestimatedERCgvaluesforthosedays.Aforecaststream,basedontheNationalDigitalForecastDatabase(NDFD)weatherforecast,canbeincluded.Andafterthosedays,climatologyapproachingtheaverageERCgtrendprovidesarangeofERCsequencesfurtherintothefuturefortheanalysisperiod.

Inthisexample,withtheminimumburnableERCgvalueat38,alloftheobservedandforecastERCstreamfallsbelowthatthreshold.Giventhat,themapshowstheresult,withaverylowprobabilityofanysignificantfirespread.Thatmaybecorrectinthiscase,butwithERCgexaggeratingtheinfluenceoflivefuels,itmaybeaseriousunderestimate.

AccuratelyportrayingtheobservedandforecastERCstreamarecriticaltotheaccuracyofFSProoutput.ItispossibletouseFFMCandBUIfromtheFWIsystemtoadjusttheERCstreamwhenpreparinginitialanalyses.ThetabletotheleftshowsFFMCandBUIclassesandsuggestshowtheyarecombinedtoidentifywhereintheERCfrequencydistributioneachdayfalls.

AnalystsshouldevaluateERCvaluesusingFFMCandBUIvaluesobservedfromrepresentativelocalweatherstationsandfindthecellthatrepresentsthatcombinationofvalues.ERCglevelscanbederivedfromtheclasslevelthetablesuggests.

Forexample,iftheFFMCis91andtheBUIis80,thecombinationsuggeststhattheERCvalueshouldbeinthethirdERCClass,withavaluebetween49and53.Because91and80arebothintermediatewithintheirclasses,theERCmightbebestrepresentedas51or52.ConsiderestimatingERCvaluesforupto3daysintheobservedERCstreamandalltheforecastedERCvaluesbeforeconductingtheinitialFSProanalysis.

9

3.3 EvaluatingandEditingERCClassTableusing“BurnDays”ClimatologyTocontinuewiththeexampleabove,thisERCClassBurnDaysSummaryshowstheresult.ERCClimatologywiththedefaultERCstreamproducedonlyabout300burnabledaysamong7000totaldays,lessthan5%.Thatamountstooneburndayin3weeks.Thesedaysallcameinthelowesttwoclassesrepresentingmoderatedburningconditions.Infact,therearefrequentinstancesinthehistoricrecordwheredryingconditionsaroundactivefiresincreasedtheriskofspreadinmuchlessthantwoweeks.

ThoughfirespreadpotentialinboreallandscapesmaynotrespondastheERCgsuggestsduringthepeakseasons,thereisanobservedepisodiccharactertofirespreadwithfireslyingdormantfordaysandthengrowingaggressivelyafterashorttransition.Thissuggestsaninfluenceoftheheatsinkinlivefuels.

Thisgraphic,basedonanalysisofFFMCandBUItrendsincombinationwithconcurrentobservedMODISdetectionshighlightsanaveragefrequencyofburndaysunderarangeoffireseasons.Overall,itsuggests1-2daysofactivespreadpotentialperweek,or15-30%ofalldaysinananalysisperiodforpeakseason.Formoreactiveseasons,thatpercentagemayriseto40%(3days)ormoreoverall.

WithverifiedERCStreams,initialFSProanalysesforagivenstartdateanddurationwillsuggestadistributionofburndaysproducedbytheclimatology.Theanalystshouldreviewthatfrequencydistributionandmakeeditstoreflecttheclimatologydemonstratedhereandtheforecastandoutlookguidanceavailable.Methodologiesaresuggestedbelow.

10

3.3.1 DuffDrivenSeason:TooManyBurnDaysInthiscommonexampleduringthepeakseasonsurroundingthesummersolstice,ERCglevelisonlybeginningtofallfromitspre-greenpeaklevels.Overall,theanalysisassumedthat82%ofalldayswereburndays,nearly6daysaweekoverall.Thereislittleevidencetosupportthisfrequencyofsignificantgrowtheveninextremeseasons.Theremaybeindividualperiodswith6-7dayswithdailysignificantspread,butnothingthatsuggestthatforanoverallaverage.

AdjustmentsintheERCstreammayalterthisdistributionofburndayssignificantly,butassumethatthestreamhasalreadybeeneditedasrecommendedabove.ReducingthefrequencyofburndayscanbeaccomplishedeasilybyreducingthenumberofERCclasses.Inthiscase,eliminatingthelowertwoclassesreducedthefrequencyfrom77%to47%.This,ineffect,ismodelingtheresistancefromtheheatsinkinlivefuels.

3.3.2 DroughtDrivenSeason:NotEnoughBurnDays

ThiscorrespondingexamplefromlaterintheseasonhighlightsthedifficultyERCghasinrepresentingfuelavailabilityandflammabilityatthattime.CurrentERCglevelsarewellbelowburndaythresholds,andtheanalysiswillproduceveryfewactiveburndaysasaresult.Giventheguidanceforburndayclimatologyabove,itwouldbeprudenttosuggestatleast15%burndaysover2weeks.Infact,ifasignificantdryingtrendisforecast,frequencyof30-40%maybeanappropriatefrequency.

Addinga6thERCClasswillproduceadditionalburndays.Butifthatisinsufficient,editingtheERCStreamevenfurthermaybenecessary.

11

4 ConclusionTheserecommendationsarepreparedspecificallyforspatialanalysisinAlaska,withemphasisonitsboreallandscapes.TheremaybesufficientapplicabilityintheWesternGreatLakesForestofMichigan,Minnesota,andWisconsintoconsidersimilarapproaches.Theseguidelinesarebasedonfirepotentialreflectedduringthegrowingseasoninnorthernforestswithsignificantlivefueladmixturesbothonthesurfaceandinthecanopy.TheydonothaveapplicabilityforcuringandcuredfuelbedsthatrepresentpeakseasonconditionsthroughoutmuchofthewesternUS.

Thegoaloftheserecommendationsistoemphasizerealobservedconditionsasinputstothemodel,toidentifywhereinthemodelsrealvariationsoffirebehaviorandfirespreadphenomenonarebestreflected,andtominimizetheneedforusingcalibratingfactorsthatmaynotreflectthemostfrequentlychangingfactorsthatdriveday-to-dayvariationinfirebehavior.ManyoftherecommendationsincludeuseofCFFDRSFireWeatherIndexsystemcodesandindices.Recent,current,andforecastvaluesandtrendscanbeexploredathttp://akff.mesowest.org.Upto3forecastdaysarenowavailableforuse.

Further,thisapproachassumesthatnearlyallsignificantgrowthoccursonfewerdayswithmoreflammableconditionsthatencouragefirestoovercometheheatsinkofthelivefuels.UsingthisapproachtoreducethefrequencyofburndaysinFSProandreduceoreliminateindividualburndaysinERCstreamsorinNTFBsequencesrequiresaconcurrentcommitmenttomodelcrownfirepotentialintheconiferfuels,especiallyinBlackSpruce.Analystsusetwoapproachestoaccomplishthis:

• Earlierinthegrowingseasonwhenhardwoodandmixedwoodforestshavegreaterlivefuelheatsinkstodiscouragespread,crownfireinBlackSprucecanbeencouragedbyconvertingthestandardfuelmodels,tu3/163(timber/grass/shrub)and/ortu4/164(dwarfconifer)tosh5/145(chapparal).Fuelloadingsarecomparable,anditeffectivelymodelsindividualgrowtheventswithobservedenvironmentalinputs.

• Laterinthepeakseason(drought-driven),whenlivefuelsmaybemorestressedacrossthelandscape,hardwoodsandmixedwoodforestsmaybemoreavailableandflammablefuelbeds.Inthiscase,selectingtheScott&ReinhardtCrownFiremethodproducescrownfireacrossthewiderspectrumoffuelmodelsdistributedacrossthelandscape.Inthiscase,itmaybeunnecessarytoconverttu3/163and/ortu4/164fueldesignations.

• Asacaution,whenusingtu3/163andtu4/164torepresentblacksprucecommunitiesoravarietyofgrassandgrass/shrubmodelsfortundralandscapes,keepinmindthatmoistureofextinctionisaslowas12%.Undertheinfluenceoffuelmoistureconditioning,therewillbenumerousinstancesthatanalyseswillproduceelevated1hr-and10h-fuelmoisturesthatcomeunderthedampeninginfluenceofthatlowmoistureofextinction.ThisisespeciallyproblematicinNTFBwheretherearenosettingstomitigateitseffect.ThereissomefacilitytodothatinSTFBwheretheanalystcanselect0(zero)conditioningdaysandsimplyuseinitialfuelmoistureinputs.Thiscanproduceacceptableresultsforsurfacefuelsunderforestcanopyandinopenflattundra,whereconditioningfactorsareminimizedontheground.

Theserecommendationsshouldhelpproduceeffectiveanalysesearlyinanincidentwithoutanysignificantcalibration.However,asthefiredevelopsahistoryofgrowthevents,perceivablevariabilityinweatherinfluences,andanaccumulationoffirelineobservationsitisappropriatetocriticallyevaluatetheseguidelines.Yourexperienceusingthemandrecommendationsforchangesareimportant.ContacttheAlaskaWildfireCoordinatingGroup’s(AWFCG)FireModelingandAnalysisCommittee(FMAC)ortheAlaskaFireScienceConsortium(AFSC)iftherearecontributionstooffer.

12

5 ReferencesWildfireAssessmentWebsites:Maps/Imagery/GeospatialServices:https://fire.ak.blm.gov/predsvcs/maps.phpWeather:https://fire.ak.blm.gov/predsvcs/weather.phpFuels/FireDanger:https://fire.ak.blm.gov/predsvcs/fuelfire.phpAirQuality:https://fire.ak.blm.gov/predsvcs/airquality.phpOutlooks:https://fire.ak.blm.gov/predsvcs/outlooks.phpFireWeatherIndex(FWI):http://akff.mesowest.orgNWSAlaskaFireWeather:http://w2.weather.gov/arh/fireNWSNationalFireWeather:http://www.srh.noaa.gov/ridge2/fire/AlaskaClimate:http://climate.gi.alaska.edu/NIFCFireEnterpriseGeospatialPortal(EGP):https://egp.nwcg.gov/egp/default.aspxWildlandFireDecisionSupportSystem:http://wfdss.usgs.govWildlandFireLibrary:https://firelibrary.org/

GuidesAlaskaFuelModelGuide(draftupdate,2016andoriginal,2008):https://www.frames.gov/files/9614/6482/3097/Revised_FuelModelGuide_Draft_May2016_Posted.pdfhttps://www.frames.gov/files/2814/3352/8200/Alaska_Fuel_Model_Guidebook_0620081.pdf

FSProAnalysisinAlaskahttps://www.frames.gov/documents/alaska/docs/FSProAnalysisAK_V1.1_Mar_2012.pdf

FieldGuidestoCFFDRSFireWeatherIndex(FWI)andFireBehaviorPrediction(FBP)Systems:https://www.frames.gov/files/3014/2309/6588/AK_FireWeatherIndex_FieldGuide_2015.pdfhttps://www.frames.gov/files/6914/2309/6585/AK_FireBehaviorPrediction_FieldGuide_2015.pdf

FireEndingEventWorkshop:https://www.frames.gov/files/1513/8749/6485/AWFCG_2008_Fire_Ending_Event_Workshop.pdf

HowtodownloadAKfireperimetersfromAICChttps://drive.google.com/file/d/0Byauxp0C04_femRBVERlWlF6SHc/view?usp=sharing

AnalysisNamingConventioninAnalystinfofolderhttps://drive.google.com/file/d/0Byauxp0C04_fZGJ4d1Rfa2JsLXc/view?usp=sharing

OtherWebResourcesAlaskaFireScienceConsortiumFireModelingResources: https://www.frames.gov/partner-sites/afsc/partner-groups/fire-behavior-modeling-group/

CFFDRSYouTubeLearningResources:https://www.youtube.com/playlist?list=PLriyD21WeCtKRA2TsWwrInsRmuC0j86HG

GrowingSeasonIndexandLiveFuelMoisture: https://www.wfas.net/index.php/growing-season-index-experimental-products-96

AlaskaFireModelingWorkshop(2012): https://www.wfas.net/index.php/growing-season-index-experimental-products-96

13

KMLs:ActiveFireMappingKMLBundle:https://fsapps.nwcg.gov/afm/data/kml/alaska_latest_AFM_bundle.kml

AICCActiveFireBundle:https://afsmaps.blm.gov/imf/sites/help/AlaskaWildfires.kml

AvenzaMapsProductsAFSPDFMaps:https://fire.ak.blm.gov/predsvcs/geopdf.php

DOFPDFMaps:https://sites.google.com/site/alaskafiremaps/home

Reports,PresentationsandPublicationsCarlson,J.D.et.al.2007.ApplicationoftheNelsonmodeltofourtimelagfuelclassesusingOklahomafieldobservations:modelevaluationandcomparisonwithNationalFireDangerRatingSystemalgorithms.InternationalJournalofWildlandFire,2007,16,204–216 Fosberg,M.A.,andJ.E.Deeming.1971.Derivationofthe1-and10-hourtimelagfuelmoisturecalculationsforfire-dangerrating.ResearchNoteRM-207.FortCollins,CO,USDAForestService,RockyMountainForestandRangeExperimentStation.

Jolly,WilliamM.,Nemani,R.andRunning,S.W.2005.Ageneralized,bioclimaticindextopredictfoliarphenologyinresponsetoclimate.GlobalChangeBiology11(4):619–632.

Kidnie,S.K.,Wotton,B.M.andDroog,W.N.2010.FieldguideforpredictingfirebehaviourinOntario'stallgrassprairie.ElginCountyStewardshipCouncilSpecialPublication.OntarioMinistryofNaturalResources,Aylmer,Ontario.65p.

Miller,EricA.2009.FireIndicesforFSProinAlaska:AcomparisonofERCandBUIonthe2009TitnaRiverFire(420).UnpublishedReport.

Nelson,RalphM,Jr.2000.Predictionofdiurnalchangein10-hfuelstickmoisturecontent.CanadianJournalofForestResearch30,1071–1087.doi:10.1139/CJFR-30-7-1071

WottonB.M.2009.AgrassmoisturemodelfortheCanadianForestFireDangerRatingSystem.Paper3-2inProceedings8thFireandForestMeteorologySymposium.Kalispell,MTOct13-15,2009

Ziel,Robert,Wolken,Jane,St.Clair,Thomas,andHenderson,Marsha.2015.ModelingFireGrowthPotentialByEmphasizingSignificantGrowthEvents:CharacterizingAClimatologyOfFireGrowthDaysInAlaska’sBorealForest.ExtendedAbstractforpresentationattheAMS11thSymposiumonFire&ForestMeteorology.May5th,2015(https://ams.confex.com/ams/11FIRE/webprogram/Paper272864.html).