High wildfire damage in interface communities in California
Heather Anu KramerA,D,Miranda H. MockrinB,Patricia M. AlexandreCandVolker C. Radeloff AASILVIS Lab, Department of Forest and Wildlife Ecology, University of Wisconsin–Madison,1630 Linden Drive, Madison, WI 53706, USA.BNorthern Research Station,USDAForestService,5523ResearchParkDriveSuite,350Baltimore,MD 21228, USA.CForest Research Centre, School of Agriculture, University of Lisbon, Tapada da Ajuda,1349-017 Lisboa, Portugal.DCorresponding author. Email: hakramer@wisc.eduAbstract.
Globally,and intheUS,wildfires pose increasing risk to people and their homes. Wildfire management assumes that buildings burn primarily in the wildland–urban interface (WUI), where homes are either ignited directly (especially in intermix WUI areas, where houses and wildland fuels intermingle), or via firebrands, the main threat to buildings in the interface WUI (areas with minimal wildland fuel, yet close to dense wildland vegetation). However, even urban areas can succumb to wildfires. We examined where wildfire damages occur among urban, rural andWUI (intermixandinterface)areasforapproximately threedecadesin California(1985–2013).We foundthatinterfaceWUIcontained50%ofbuildings destroyedbywildfire,whereasintermixWUIcontainedonly32%.The proportionofbuildingsdestroyedbyfiresamongclasseswassimilar,thoughhighestininterfaceWUIareas(15.6%).OurresultsdemonstratethattheinterfaceWUIiswheremostbuildingsweredestroyedinCalifornia,despitelesswildlandfuel.Continuedadvancementofmodels,mitigationandregulationstailoredfortheinterfaceWUI,bothforCaliforniaandelsewhere,willcomplementthepriorfocusontheintermixWUI.Additionalkeywords:destruction,hazard,housingloss,plan,policy,wildland–urban interface.Received14July2018,accepted24June2019,published online 30 July 2019IntroductionOver the past several decades, catastrophic wildfires have caused mounting economic, social and ecological damage across the globe (Moritzet al.2014; Bowmanet al.2017). For example, Australia, Canada, Portugal and Chile have all seen record destruction and loss of life due to wildfire in recent years(Godoyet al.2019; Cruzet al.2012; Go ́mez-Gonza ́lezet al.2018; Oliveiraet al.2017; Boulianneet al.2018). The UnitedStates is similarly incurring high rates of wildfire damage,despite soaring expenditures on wildfire management (Fischeret al.2016; Steelman 2016). For example, in 2018 the Camp firedestroyed nearly 19 000 structures and killed 85 people – thedeadliest fire in California history (CAL FIRE 2018b, 2018c).In California, and the rest of the USA, wildfire managementhas become more complex, costly and dangerous as a result ofmultiple factors, including climate change, legacies of wildfiresuppression and housing growth (Fischeret al.2016; Steelman2016). One-third of all homes in the US are built in or nearwildland vegetation and constitute the wildland–urbaninterface, or WUI, a complex environment that increases thechallenges of wildfire management (Schoennagelet al.2017;Radeloffet al.2018). Because ignitions are overwhelminglyhuman-caused (Nagyet al.2018), wildfire ignitions are morelikely with WUI expansion (Syphardet al.2007; Syphardet al.2017) and even more so as the entire wildfire season lengthensowing to more ignitions and climate change (Balchet al.2017;Schoennagelet al.2017). At the same time, there are morehomes to protect in the event of wildfire (Radeloffet al.2018),resulting in greater wildfire suppression expenditures (Gudeetal.2013; Handet al.2016). Indeed, 69% of buildingsdestroyed by wildfire across the US are located in the WUI,and in California, that number rises to 75% (Krameret al.2018).Accordingly, wildfire policy and recommendations oftenfocus on the WUI and aim to reduce building ignition eitherfrom nearby vegetation or from firebrands (United StatesCongress 2003; Wildland Fire Executive Council 2014). Forexample, a recent presidential Executive Order intended toreduce the likelihood of wildfire risk to federal buildingsthrough retrofitting and maintaining defensible space (Obama2016). However, beyond these federal buildings, the US federalgovernment does not regulate or influence land-use planningand building practices in response to wildfire hazardsCSIROPUBLISHINGInternational Journal of Wildland Firehttps://doi.org/10.1071/WF18108Journal compilationÓIAWF 2019www.publish.csiro.au/journals/ijwf
(as opposed to othernatural hazards such as floods) (Burby 2001;Thomas and Leichenko 2011). Instead, the responsibility foradapting residential development to wildfire falls to localgovernments and communities. For example, city or countygovernments can use land-use regulations to guide or restrictresidential development to reduce risk of wildfire loss, andincorporate wildfire risk into community planning (Fire AdaptedCommunities Coalition 2014;FAC Learning Network 2016).At the level of the individual home, mitigation through buildingmaterials or vegetation can be required by building codes,overlay zoning, and other ordinances or regulations (Winteret al.2009;McCaffreyet al.2011). Voluntary efforts andeducation programs are also widespread (Krameret al.2018),such as the Firewise program, which provides homeowners withrecommendations on how to reduce risk from wildfires throughmitigation (National Fire Protection Association 2016).Management efforts commonly focus on all types of WUIuniformly, or define it loosely or with inconsistent definitions(Platt 2010). The 2001 Federal Register definition of the WUIdefines two separate types of WUI: interface WUI (Fig. 1a), i.e.developed areas that have sparse or no wildland vegetation, butare within close proximity of a large patch of wildland, andintermix WUI (Fig. 1b), i.e. the area where houses and wildlandvegetation directly intermingle (USDA and USDI 2001). Bothtypes of WUI have been widely mapped (Radeloffet al.2005;Radeloffet al.2018), based on the specifications of the FederalRegister WUI definition (USDA and USDI 2001). These mapsare used to inform federal policy, including a recent ExecutiveOrder (Obama 2016).Within this WUI framework, the intermix WUI is oftendescribed as more difficult to manage for wildland fire (Daviset al.2000).Indeed,thedesignationofintermixvsinterfaceWUIimplies that there is less flammable vegetation in the interfaceWUI. However, landscaping does not constitute a 'wildlandfuel', yet can carry fire, and, in the case of certain fire-pronevarieties such as eucalyptus, cypress and juniper, can propagatefire morerapidly than native fuels (CAL FIRE 2017). Among thetwo WUI types, there are far more houses in the interface WUI,but the intermix WUI is far more widespread (Radeloffet al.2018), and eight times more prevalent in area within destructivewildfires in the US (Krameret al.2018). However, interfaceWUI, and even urban areas, can also experience devastatinglosses from wildfires. For example, the 2017 Tubbs fire inCalifornia caused major damage within the city limits of SantaRosa, where it destroyed entire urban neighbourhoods and 5636buildings, making it one of the most destructive wildfires in UShistory. Furthermore, recent work bySyphardet al.(2019)suggests that rural destruction was high in a sample of firesthroughout California, further fuelling the debate on housingdensity and wildfire destruction. A further incongruence existsbetween research, which focuses more on the WUI, and man-agement and legislation, which in California is often based onmapped Fire Hazard Severity Zones, which only peripherallyconsider the WUI (Daviset al.2000; CAL FIRE 2007).An enhanced understanding of where wildfire losses occur,within and beyond the WUI, could help refine managementrecommendations. Although houses in the intermix WUI can bedirectly ignited from nearby wildland vegetation, other buildings,or firebrands (burning material that can be carried aloft), directignition from nearby wildland vegetation in the interface WUI islesscommon(becauseinterfaceareasaredefinedbyhavinglowerdensities of wildland vegetation), and firebrands, landscapingvegetation and other buildings are a more common source ofignition (Stewartet al.2007;Haaset al.2013). Inother words, thetwo types of WUI may require different management approaches.For example, maintaining defensible space and reducing fuelloads over large landscapes may be more effective in reducingwildfire losses in the intermix WUI than the interface WUI.Furthermore, fuel models and wildland fire behaviour modelsarelackingforareaswithanabundanceofnon-naturalfuelsuchaspropane tanks, vehicles and the homes themselves, as is com-monly the caseinurban and interface WUI areas (Anderson1982;Scott and Burgan 2005;Maranghides and Mell 2012).Our overarching research question was thus to determine theextent of wildfire destruction associated with different residen-tial development settings (urban, interface WUI, intermix WUIand rural, as defined by the Federal Register (USDA and USDI2001) and mapped byRadeloffet al.(2005)) in California – thestate with both the most building destruction in the WUI, as wellas the most building destruction from wildfire across the US(Krameret al.2018). We examined wildfires across Californiaover a 28-year period (1985–2013) and include the 2017 Tubbsfire as a recent case study of a notably destructive and urbanwildfire in order to fully characterise the challenges wildfireposes to homes and buildings in this densely developed andfire-prone state. Our research questions were: (1) how are(a)(b)(c)(d)Fig. 1.Images of (a) interface wildland–urban interface (WUI) (in the2003 Cedar fire); (b) intermix WUI (in the 2007 Witch fire); (c) urban non-WUI (in the 2017 Tubbs fire); and (d) rural non-WUI (in the 2008 BTULightning complex). Images were obtained from Google Earth (Google Inc.2016) for the year before each fire. Substantial destruction occurred tobuildings in each image during the subsequent fire.BInt. J. Wildland FireH. A. Krameret al.
building losses distributed between intermix WUI, interfaceWUI, non-WUI urban and non-WUI rural areas; and (2) do thesepatterns of building losses vary with the destructiveness (definedas the total number of buildings destroyed) of the wildfire, andamong individual wildfires? We expected both the total numberof buildings destroyed and the rate of destruction to be higherin the intermix WUI than the interface WUI because (i) theinterface WUI is smaller in area than the intermix WUI,(ii) firebrands, other buildings and landscaped vegetation arethe main source of ignitions in the interface WUI, and (iii) densewildland vegetation is mixed with buildings in theintermix WUI.MethodsStudy areaCalifornia is an apt location for our study. Owing to the abun-dance of wildfires and close proximity of people and wildfire-prone landscapes, the state has a long history of destructivewildfires. Between 2000 and 2013, more buildings weredestroyed by wildfire in California than in the other 47 conter-minous US states combined (Krameret al.2018). Hundreds ofmillions of dollars are spent each year on suppression alone, ontop of costs for mitigation, education and research into wayspeople and wildfire can coexist (California Department ofForestry and Fire Protection 2017). The state is highly diverse inclimate, vegetation, topography, land ownership, and develop-ment densities and patterns. California's natural systems rangefrom chaparral, most prominent in the south coast bioregion,redwood on the central coast, to oak woodland in the centralcoast and in a band across lower elevations of the Sierra Nevada,with mixed-conifer and alpine systems at higher elevations inthis bioregion (Sugiharaet al.2006). Building settlement pat-terns are also highly diverse, with large urban centres, bothwidespread interface and intermix WUI, and large swathes ofrural and agricultural lands. Furthermore, human communitiesvary broadly in income and socioeconomic status across thestate, with high racial and ethnic diversity (US Census Bureau2018). Wildfires that destroy buildings occur throughout Cali-fornia, and although most wildfires that destroy many buildingshave affected southern California, the three most destructivewildfires (in terms of building loss) in state history, the 2018Camp fire, 2017 Tubbs fire and 1991 Oakland Hills fire,occurred in the northern half of the state (CAL FIRE 2018c).DataTo determine the locations of buildings destroyed by wildfire, wedigitised all buildings before and after all wildfires that burnedbetween1985and2013thatcontainedatleastonebuildingandforwhich there was sufficient imagery (Fig. 2). Of 270 wildfiresconsidered, 89 destroyed at least one building. Data collectionmethods for the post-2000 wildfires (n¼78) are described byAlexandreet al.(2015)and methods for the pre-2000 wildfires(n¼11)aredescribedbyKrameret al.(2018),whichalsoincludesanaccuracy assessment for wildfiresfrom 2000 to2013. Althoughmethods are similar for these two datasets, image availability wasmore limited for older fires. We used wildfire perimeters from theMonitoring Trends in Burn Severity (MTBS) dataset, whichincludes all wildfires greater than 404 ha (1000 acres) in thewestern US (Monitoring Trends in Burn Severity 2016).As a case study, we acquired data on building location anddestruction status in the 2017 Tubbs fire. The wildfire perimeterwas obtained fromCAL FIRE (2018a). Building locations werecomposed of points from two sources. CAL FIRE providedpoint datafor all 5636 destroyedbuildings,aswellasfor damagedand surviving buildings, which we combined into a single non-destroyed class (n¼371) (CAL FIRE and Kephart 2018).Because some surviving buildings were skipped in the rapidpost-fire assessment, and did not appear in the CAL FIREdatabase, we used Sonoma County building footprints derivedfrom LiDAR flown in 2013 (Sonoma County Agricultural Preser-vationand OpenSpace District2018) tosupplementthe survivingbuildings dataset, adding (2519) additional surviving buildingpoints at the centroids of all footprint polygons with perimetersthat were.5 m from a CAL FIRE building point. Although thisapproach does not account perfectly for every building within thefire perimeter, quality check assured us that the accuracy wassufficient for the purpose of the present case study.For all wildfires studied, we used the 1990, 2000 and 2010WUI classifications created byRadeloffet al.(2018)to identifyinterface WUI, intermix WUI, urban and rural areas. WUI areaswere defined as census blocks with at least 6.17 homes km 2andeither (1) 50% wildland vegetation within the census block –classified as intermix WUI; or (2) a large (at least 5 km2)areaofdense wildland vegetation (at least 75%), within 2.4 km (based onthe distance firebrands may travel) – classified as interface WUI.These WUI types and thresholds are based on the 2001 FederalRegister definition of the WUI (USDA and USDI 2001;Radeloffet al.2005). Non-WUI areas were classified as urban, i.e. thosewith$6.17 homes km 2but with insufficient vegetation, evennearby, to be classified as WUI, or rural, i.e. those with,6.17homes km 2and with either wildland vegetation or not. Weapplied the 1990 WUI map to wildfires that burned between 1985and 1994, the 2000 WUI map to wildfires that burned between1995 and 2004, and the 2010 WUI map to wildfires that burnedbetween 2005 and 2013, as well as to the Tubbs fire case study.What is the distribution of building destruction in WUIand non-WUI classes?We calculated the area burned and buildings destroyed withininterface WUI, intermix WUI, non-WUI urban and non-WUIrural for all wildfires in our dataset that burned at least onebuildingbetween1985and2013.Wesummarisedthepercentageof each wildfire area in each class, where class represents urban,interface WUI,intermixWUI and ruralareas. We alsocalculatedtheproportionoftotaldestructionanddestructionrateasfollows:PðdcÞ¼DcDtRc¼DcDtþScP(dc)¼proportion of total destruction in a given classRc¼destruction rate for a given classDc¼number of buildings destroyed in a given classDt¼total number of buildings destroyedSc¼number of buildings that survived in a given classHigh wildfire damage in California interface WUIInt. J. Wildland FireC
We present findings for all wildfires together, as well asternary plots to examine variation among individual wildfires.Ternary plots show each point as a position within a triangle,where each side of the triangle forms an axis that ranges from0 to 100, and for which the three values for each point must sumto 100. As ternary plots have, by definition, three axes, wecombined urban and interface WUI into a single axis (the urbanarea was small and had little destruction). Intermix WUI andrural areas were represented by the other two axes. We createdtwo ternary plots showing how the (1) proportion of area, and(2) proportion of destruction in each class varied among firesand by fire size.How does overall wildfire destruction of buildings affectthe distribution of building destruction and area in WUIand non-WUI classes?We compared the building destruction and wildfire area inurban, interface WUI, intermix WUI and rural areas withthe overall destructiveness of the 89 fires in our dataset.We summarised these relationships in scatterplots with Pearsoncorrelation coefficients. We performed an exploratory analysisof how destruction rate changed over time in each WUI and non-WUI class using linear models, yet we caution that our data forolder fires is a small subset that may not capture some trends (seePart 4 in Supplementary material available online). Because ofthis difference in data availability for older fires, we did notperform an analysis of absolute destruction. We also describethe 2017 Tubbs fire as a case study, which was the mostdestructive California wildfire (destroyed the most buildings) atthat time. See Table A1.1 in Supplementary material for fulltable of wildfire area, destruction and survival.Robustness checks and caveatsInterface and intermix areas, by definition, differ in their resi-dential density and distribution of vegetation. Therefore, weconducted an additional exploratory analysis to test whetherbuilding density was significantly related to building survival.We compared the distance from each building to the nearestFig. 2.Map of California, showing the 89 wildfires that burned between 1985 and 2013, the 2017 Tubbsfire, and the total building destruction in each wildfire.DInt. J. Wildland FireH. A. Krameret al.building, nearest destroyed building and nearest survivingbuilding in urban areas, interface WUI, intermix WUI and ruralareas to see if our findings on building loss followed from thespatial configuration of buildings within these different settings(see Part 2 in Supplementary material). Because buildingsthemselves often serve as fuel, we expected shorter distancesbetween destroyed buildings in urban and interface WUI areascompared with buildings that survived. Although furtherinvestigation of building proximity to burned vegetation orhigh-severity fire would be valuable, such data collection wasout of the scope of this study.Because we used a national WUI map to investigate wildfirelosses and policy, we also examined the state-level policydesignations unique to California. California requires wildfiremitigation in designated hazard areas, as mapped by CAL FIRE(Daviset al.2000; CAL FIRE 2007). Starting in the 1980s,regulations for mitigation (building materials and defensiblespace) by wildfire hazard zones were first adopted for StateResponsibility Areas (SRAs), i.e. areas where the State ofCalifornia (CAL FIRE) has the responsibility for suppression.After the catastrophic Oakland Hills fire (Tunnel fire) of 1991,hazard zoning and mitigation requirements were expanded toLocal Responsibility Areas (LRAs), which includes incorpo-rated cities, densely populated areas and agricultural lands(1992 Bates Bill (California State Assembly 1992);CAL FIRE2018a). Because the efficacy of state-level hazard classifica-tions has been questioned in the past (Syphardet al.2012), weexamined destruction within these zones for all wildfires (seePart 3 in Supplementary material).ResultsWhat is the distribution of building destruction in WUIand non-WUI classes?From 1985 to 2013, we mapped 8722 buildings destroyed bywildfire in 89 individual wildfires that destroyed at least onebuilding, an overall destruction rate of 14% of all buildingsthreatened by these fires. Although only 32% of buildings inCalifornia are located in the WUI, 82% of the destroyed build-ings were located in the WUI (Table 1).Nearly all the area within wildfire perimeters was rural(89%), yet these rural areas contained only 14% of the buildingsthat were destroyed (Table 2;Fig. 3). Interface WUI accountedfor 50% of all buildings destroyed, despite covering only 1.8%of the total area burned and comprising only 27% of buildings inCalifornia (Table 1,Table 2). The destruction rate was alsohighest in the interface WUI (15.6% of all interface buildingswithin fire perimeters were destroyed;Table 2), though destruc-tion rates were similar across urban, interface WUI, intermixWUI and rural areas (11.3, 15.6, 11.6 and 14.1% respectively;Table 1). Ternary plots examined variation in destruction andWUI classes among individual wildfires. Nearly all of thewildfires occurred primarily in rural areas, but rural destructionwas variable (Table 2;Fig. 3b). For example, although ruralareas accounted for only 14% of all destroyed buildings (of allfires combined), 44 of 89 individual wildfires experienced morebuilding destruction in rural areas than all other areas combined,although the total number of buildings in these wildfires was low(these 44 wildfires only accounted for 632 (7.2%) buildingsdestroyed;Fig. 3b; Table A1.1).How does overall wildfire destruction of buildings affectthe distribution of building destruction and area in WUIand non-WUI classes?More destructive fires threatened and destroyed a higher pro-portion of buildings in the interface WUI and fewer in rural areas(Fig. 4). Destruction rate in more destructive wildfires washigher in urban, interface WUI and intermix WUI areas, yetTable 1. California wildfire area and building destruction in intermix wildland–urban interface (WUI), interface WUI, urban and rural areaswithin fires that burned between 1985 and 2013 (n589), as well as the proportion of area and buildings in each class in California statewideArea(km2)Area(%)Area inCalifornia (%)Survival(no. buildings)Destruction(no. buildings)Proportiontotalbuildings (%)Proportion totalbuildings inCalifornia (%)Proportion totaldestruction (%)Destructionrate (%)Non-WUI, urban110.11.027473494.957.14.011.3Interface WUI1551.81.923 658437444.026.950.115.6Intermix WUI8419.54.521 178279037.65.532.011.6Non-WUI, rural7836 88.692.67362120913.510.513.914.1Total8844 100.0100.054 9458722100.0100.0100.013.7Table 2. Wildfire area and building destruction in intermix wildland–urban interface (WUI), interface WUI, urban and rural areas forwildfires that burned between 1985 and 2013, as well as the 2017 TubbsfireAll California (n¼89)Tubbs (n¼1)Area (%)Non-WUI, urban0.11.3Interface WUI1.85.2Intermix WUI9.541.2Non-WUI rural88.652.3Proportion total destruction (%)Non-WUI, urban4.025.4Interface WUI50.134.9Intermix WUI32.035.7Non-WUI rural13.94.0Destruction rate (%)Non-WUI, urban11.375.7Interface WUI15.672.4Intermix WUI11.661.5Non-WUI rural14.135.2High wildfire damage in California interface WUIInt. J. Wildland FireE
destruction rate in rural areas was not related to the fire'soverall destruction (Fig. 4). However, more destructive wild-fires had a substantially higher proportion of area in interfaceWUI and lower proportion of rural area (Fig. 4). Destructionrate did not significantly change over time, either for all fires, orfor any individual WUI or non-WUI class (Part 4 in Supple-mentary material).In the Tubbs fire, similar to other California wildfires,destruction was primarily in the WUI (71 and 82% destructionrespectively;Table 2). Although more destructive fires showeda higher proportion of destruction in the interface WUI, theTubbs fire had approximately equal destruction in the interfaceand intermix WUI (35 and 36% respectively;Table 2). How-ever, 1/4 of destruction in the Tubbs fire occurred in urban areas,compared with 4% for California fires (Table 2). Indeed, only 5fires in our dataset of 89 had any destruction at all in urban areas,totalling 349 buildings, compared with 1430 urban buildingsdestroyed in the Tubbs fire alone (Table A1.1). The destructionrate was also very high in the Tubbs fire across urban areas, aswell as interface and intermix WUI (76, 72 and 62% respec-tively), matching trends of higher destruction rates in moredestructive fires (Fig. 4).Robustness checks and caveatsOur exploratory analysis of building proximity and destructionrevealed high variability. Although destroyed buildings weregenerally closer to other destroyed buildings than survivingbuildings, this relationship was not significant (owing to highvariability in distances). There was no significant differenceamong distances between buildings for the WUI and non-WUItypes (see Part 2 in Supplementary material).Fire Hazard Severity Zones accurately matched area burnedand destruction rate in most wildfires, with the exception of theTubbs fire. Of all area burned by destructive wildfires in oursample, 86% fell into the Very High hazard class and captured78% of destruction. Destruction rates were highest for High andVery High classes (13% in both; see Part 3 in Supplementarymaterial). In contrast, the Tubbs fire burned the most area (51%)in Moderate zones and the most buildings (39%) in Urban(unrated) areas, where destruction rate was also the highest(73%; see Part 3 in Supplementary material).DiscussionInterface WUI areas accounted for the majority of buildingdestruction in California wildfires that destroyed at least onebuilding. In fires with more overall building destruction, thepercentage of buildings in the interface WUI, as well as thedestruction rate in urban, interface WUI and intermix WUI washigher. More-destructive wildfires contained a larger proportionof interface and intermix WUI, and less rural area. In total, halfof all buildings destroyed by wildfire were located in theinterface WUI, which composed only 2% of the area burned bythese wildfires (though interface WUI includes 27% of allhomes in California). Within fire perimeters, buildings in theinterface WUI had the highest chance of destruction fromwildfire. This may have been due to non-wildland fuel in theseareas (e.g. homes, vehicles, propane tanks and landscapingvegetation) or other factors.Fig. 3.Proportion (a)area,and(b) total destruction in the intermixwildland–urban interface (WUI), non-WUI rural, interface WUI, and non-WUI urban areas for 89 destructive California wildfires (1985–2013) in blackand the 2017 Tubbs fire in red. Note that only six wildfires had any destructionin urban areas, which were combined with interface WUI. Each hollow circlerepresents a fire. Follow the centre of each point to each of the three axes tocalculate the relative proportion of that fire in each of the three classes. Forexample, the Tubbsfire(shown inred) had60% of its totaldestruction inurbanand interface WUI areas, 36% in intermix WUI and 4% in rural areas.FInt. J. Wildland FireH. A. Krameret al.
Urban wildfire building losses were rare, but concentratedwhen they did occur, and were nearly absent from the areaburned by destructive wildfire for all fires from 1985 to 2013.Indeed, urban areas accounted for a low proportion of totaldestruction, as well as a relatively low rate of destruction,presumably because these areas lacked wildland fuels (as inFig. 1c) and likely received increased suppression resources,making it rare for wildfire to even approach these areas (Gudeet al.2013). In contrast, in the Tubbs fire, there were extensivedamage and high destruction rates in urban areas (Table A1.1).The Tubbs fire also departed from the usual patterns observedfor Fire Hazard Severity Zones, with fewer losses in the highest-rated areas. Even the 1991 Oakland Hills (Tunnel) fire (the mostdestructive wildfire in our primary dataset) only experienced10% of destruction in urban areas, with an urban destruction rateof 34% (compared with 25% of destruction in urban areas and adestruction rate of 76% in the Tubbs fire; Table A1.1). The speedand destructiveness of the Tubbs fire were likely due to acombination of extreme winds that carried the fire and fire-brands long distances, and the high number of people and homeswithin the fire's perimeter (Keeley 2017). However, even for theTubbs fire, WUI areas contained the majority of the losses, anddestruction rates were high in all areas (interface WUI, intermixWUI and urban all had destruction rates of 60% or higher). Otherrecent and highly destructive fires including the 2018 Carr,Camp and Woolsey fires included no urban area within theirperimeters, exemplifying the rarity of building destruction bywildfire in urban areas.Rural areas encompassed the majority of the area burned bydestructive wildfires, and ternary plots revealed that half ofdestructive California wildfires destroyed buildings primarily inrural areas; yet these fires threatened and destroyed few buildingsoverall owing to low building density (Fig. 1d;Fig. 3). The rate ofdestruction in rural areas was also high. Buildings in non-WUI,rural areas with wildland vegetation certainly remain vulnerableto wildfires as evidenced by this high overall destruction rate (seealsoKrameret al.2018). In many cases, the reason why theserural areas with wildland vegetation are not mapped as WUI isthat their housing density is too low (Radeloffet al.2018).A consistent threshold that differentiates WUI from ruralareas is important, both in comparing results of different studies,and in relating results to management actions on the ground.Recent work bySyphardet al.(2019)suggested that destructionby wildfire in California was most prevalent in rural as opposedto WUI or urban areas. However, they found an overall meanhousing density of 0.08 to 2.01 structures ha 1for destroyedFig. 4.Grid of scatterplotsshowing howthe naturallogoftotal building destruction relates tothe proportionofarea, threatenedbuildings,destroyedbuildings and destruction rate across urban, interface wildland–urban interface (WUI), intermix WUI, and rural areas. Significant correlationcoefficients are indicated for each scatterplot in bold, with significance indicated forP,0.05 (*),P,0.01 (**) andP,0.001 (***).High wildfire damage in California interface WUIInt. J. Wildland FireG
structures, which is well above the WUI threshold of 0.062homes ha 1, suggesting that what we define as WUI matcheswhat they define as rural (Stewartet al.2007;Syphardet al.2019). Tracking these housing densities consistently acrossstudies is essential for comparing wildfire management, lossesand policy implications in the diversity of settings wherewildfire poses a threat to housing development (Stewartet al.2007;Platt 2010). Although other WUI definitions exist, thosedescribed by the Federal Register (USDA and USDI 2001) areused nationally and provide a consistent framework on which toevaluate the WUI. This does not mean that these definitions areset in stone, however, and future definitions could furtheradvance the ways that WUI is mapped and regulated.Our results highlight that wildfire can cause extensivedamage, even in areas with relatively little wildland vegetation.Although some wildland vegetation was present in urban andinterface WUI areas (Fig. 1cand1a), it was insufficient to mapthese areas as interface WUI. Indeed, using theScott and Burgan(2005)set of 40 fire models mapped by the LANDFIRE project(http://www.landfire.gov, accessed 4 March 2019), over half ofall destroyed buildings in our dataset (54%) were located in'unburnable' land-cover classes. These areas are consideredsusceptible to ignition primarily by firebrands, non-wildlandvegetation such as landscaping and agriculture and the buildingsthemselves, which become important fuel and sources of fire-brands in the interface WUI (Maranghides and Mell 2012;Syphardet al.2014;CAL FIRE 2017). Because buildingsthemselves often serve as fuel, we expected shorter distancesbetween destroyed buildings in urban and interface WUI areascompared with buildings that survived. Although this trendexisted, sample variance was high and the trend was notsignificant.CaveatsIn interpreting our findings, it is important to keep severalpossible sources of error associated with our analyses in mind.Our data on building destruction may not account for alldestruction. For example, a building may have been missed inour analyses owing to visual occlusion by overhanging vege-tation in the aerial image. Also, because imagery immediatelyafter the wildfire was not always available, some buildings mayhave been recorded as having survived when, in reality, theywere destroyed and rebuilt. However, only images under 3 yearsafter the wildfire were used to minimise these errors. Owing toimage and wildfire perimeter availability, we were not able tosample all wildfires in California between 1985 and 2013,especially older wildfires, where little information is available;however, the sampled wildfires should accurately representwildfires that destroy buildings in California. Although we didnot find significant temporal trends in destruction rates, oursmall sample may have been insufficient to identify potentialexisting trends. This sampling design meant that we could notmake an inference about the full population of all wildfires, andwere limited to statements about the fires that we analysed only.For these fires, however, we mapped all buildings and, as such,any differences are statistically significant when accounting forthe proportion of the population (of the buildings in our fireperimeters) that were sampled (which was all of them). Finally,WUI maps were based on work byRadeloffet al.(2005,2018),and represent areas that meet the housing density and vegetationcriteria as described by the Federal Register (USDA and USDI2001), but they do not predict wildfire risk based on other factors(vegetation, ignitions, topography, weather patterns, etc.).Although all of these factors may have affected the exactnumbers that we presented, none of them are likely to haveaffected our main conclusions.Management implicationsOur results have important implications for both wildfirepolicy and wildfire modelling. Wildfire models predict wildfirebehaviour and effects based on flammable natural fuels, yet ourresults indicate that wildfire can be highly destructive in theinterface WUI, where wildland fuels are sparse (see alsoMaranghides and Mell (2012)andSkowronskiet al.(2016)).Landscaping, agricultural vegetation and fuels other than veg-etation are rarely considered in sets of fuel models such as thoseofAnderson (1982)andScott and Burgan (2005). Wildfiremodels are used to predict the behaviour of specific wildfireevents, as well as gauge the relative wildfire risk in differentareas, yet without non-wildland vegetation, buildings, propanetanks, wood piles and vehicles included as potential vectors forwildfire, model performance in the WUI is likely to be poor.Although some WUI fire and fuel models exist (Haaset al.2013;Dietenberger and Boardman 2017), further research into thedynamics of wildfire spread and hazard in the WUI with itsdiverse natural and manufactured fuels could improve modelpredictions of wildfire behaviour and effects (Mellet al.2010;Mahmoud and Chulahwat 2018). In addition, continued focuson studying and mapping areas at risk of wildfire, as well asidentifying which mitigation strategies are most effective inthese areas are key components to reducing future buildingdestruction by wildfire. Such research could be helpful inimproving and updating the WUI definition, as well as hazardrisk ratings for the state of California (although many FireHazard Severity Zones mapped hazard accurately for wildfires,model results were a poor fit for the Tubbs fire). The Tubbs fire,in addition to other fires in our dataset, was characterised bystrong winds, which can lead to increased fire spread rate andeconomic damages and loss (Jinet al.2015). We did notexamine the relationship between wind and building destructionby fire, but it represents an important consideration to includein fire and risk models. Although nothing can eliminate riskentirely as long as people, buildings and fuel are present, manystrategies can decrease wildfire risk, and certain actions may beespecially beneficial in the interface WUI. For instance, forindividual buildings, using fire-resistant building materials andmaintaining defensible space in the home ignition zone (see alsoGibbonset al.(2018)), even when that zone extends acrossproperty boundaries (see California Public Resources Code4291;Schwarzenegger 2004), can reduce fire risk (Cohen 2008;Platt 2014;National Fire Protection Association 2016).ConclusionWe found that the interface WUI, i.e. settled areas with littlewildland vegetation that are near large blocks of wildland veg-etation, is where the greatest total amount of building destruc-tion has occurred in California (the state with more buildingHInt. J. Wildland FireH. A. Krameret al.
destruction by wildfire than all other states combined) in thecase of destructive fires that burned between 1985 and 2013.Wildfire and fuel models that include the broad range of thefuels present in the interface WUI (including landscaping,agricultural vegetation, vehicles and structures themselves) areimportant to understand fire behaviour and effects in these moredensely populated areas. A combination of improved modelling,research into wildfire risk in densely built areas, fuel reductionin the home ignition zone, use of fire-resistant landscaping andbuilding materials, strategic placement of fuel reduction treat-ments around communities, and community education andplanning of building locations in regards to wildfire could leadto policies and mitigation that reduce wildfire risk in the inter-face WUI.Conflicts of interestThe authors declare that they have no conflicts of interest.AcknowledgementsWe gratefully acknowledge support for this work by the Joint Fire ScienceProgram and the Rocky Mountain Research Station and Northern ResearchStation of the USDA Forest Service. We also acknowledge the ForestResearch Center, which is a research unit funded by Fundacao para a Cienciae a TecnologiaI.P. (FCT),Portugal (Grant UID/ABR/00239/2013).WethankV. Butsic, M. Hand, M. Grove and anonymous reviewers for helpful com-ments on this manuscript. D. Helmers provided guidance on geoprocessing,and C. Armer, A. Bontje, J. Canfield, A. Ciurro, B. Fendt, C. Frederick,S. Gordon, J. Paulson, T. Ng and C. Trott assisted with data collection.ReferencesAlexandre PM, Mockrin MH, Stewart SI, Hammer RB, Radeloff VC(2015) Rebuilding and new housing development after wildfire.Inter-national Journal of Wildland Fire24, 138–149. doi:10.1071/WF13197Anderson HE (1982) Aids to determining fuel models for estimating firebehavior. USDA Forest Service, Intermountain Forest and RangeExperiment Station, General Technical Report INT-122. (Ogden, UT).Balch JK, Bradley BA, Abatzoglou JT, Nagy RC, Fusco EJ, Mahood AL(2017) Human-started wildfires expand the fire niche across the UnitedStates.Proceedings of the National Academy of Sciences of the UnitedStates of America114, 2946–2951. doi:10.1073/PNAS.1617394114Boulianne S, Minaker J, Haney TJ (2018) Does compassion go viral? Socialmedia, caring, and the Fort McMurray wildfire.Information Communi-cation and Society21, 697–711. doi:10.1080/1369118X.2018.1428651Bowman DM, Williamson GJ, Abatzoglou JT, Kolden CA, Cochrane MA,Smith AM (2017) Human exposure and sensitivity to globally extremewildfire eventsNature Ecology & Evolution1, 0058. doi:10.1038/S41559-016-0058Burby RJ (2001) Flood insurance and floodplain management: the USexperience.Global Environmental Change Part B: EnvironmentalHazards3, 111–122. doi:10.1016/S1464-2867(02)00003-7CAL FIRE (2017) Fire-resistant landscaping. Available at http://www.readyforwildfire.org/prepare-for-wildfire/get-ready/fire-resistant-land-scaping/ [Verified 1 July 2019]CAL FIRE, Kephart A (2018) Tubbs fire spatial data. CAL FIRE.CAL FIRE (2018a) Incident information. Available at http://www.fire.ca.gov/incidents/ [Verified 01 07 2019]CAL FIRE (2018b) Top 20 deadliest California wildfires. Available athttps://calfire.ca.gov/communications/downloads/fact_sheets/Top20_Deadliest.pdf [Verified 27 11 2018].CAL FIRE (2018c) Top 20 most destructive California wildfires. Availableat http://www.fire.ca.gov/communications/downloads/fact_sheets/Top20_Destruction.pdf [Verified 27 11 2018]California Department of Forestry and Fire Protection (2017) Emergencyfund fire suppression expenditures. Available at http://www.fire.ca.gov/fire_protection/downloads/SuppressionCostsOnepage.pdf [Verified27 11 2018]California State Assembly (1992) California State Assembly Bill 337.(Sacramento, CA, USA).Cohen J (2008) The wildland–urban interface fire problem.Forest HistoryToday Fall2008, 20–26.Cruz M, Sullivan A, Gould J, Sims N, Bannister A, Hollis J, Hurley R(2012) Anatomy of a catastrophic wildfire: the Black Saturday KilmoreEast fire in Victoria, Australia.Forest Ecology and Management284,269–285. doi:10.1016/J.FORECO.2012.02.035Davis G, Nichols MD, Tuttle AE, Allshouse W (2000) Fire hazard zoningfield guide. University of California Berkeley Forest Products Labora-tory. (Berkeley, CA, USA).Dietenberger MA, Boardman CR (2017) EcoSmart Fire as structure ignitionmodel in wildland–urban interface: predictions and validations.FireTechnology53, 577–607. doi:10.1007/S10694-016-0632-0FAC Learning Network (2016) Fire-Adapted Communities self-assessmenttool. Available at http://fireadaptednetwork.org/resources/fac-assess-ment-tool/ [Verified 1 July 2019]Fire-Adapted Communities Coalition (2014) Guide to Fire-AdaptedCommunities. Available at http://www.fireadapted.org/,/media/Fire-%20Adapted/Files/FAC%20Reference%20Guide%202014%20FINAL%20reduced%202.pdf [Verified 19 April 2017].Fischer AP, Spies TA, Steelman TA, Moseley C, Johnson BR, Bailey JD,Ager AA, Bourgeron P, Charnley S, Collins BM (2016) Wildfire risk asa socioecological pathology.Frontiers in Ecology and the Environment14, 276–284. doi:10.1002/FEE.1283Gibbons P, Gill AM, Shore N, Moritz MA, Dovers S, Cary GJ (2018)Options for reducing house losses during wildfires without clearing treesand shrubs.Landscape and Urban Planning174, 10–17. doi:10.1016/J.LANDURBPLAN.2018.02.010Godoy M, Martinuzzi S, Kramer H, Defosse ́ G, Argan ̃araz J, Radeloff V(2019) Rapid WUI growth and wildfires in a natural-amenity rich regionin central-western Patagonia, Argentina.International Journal of Wild-land Fire. doi:10.1071/WF18097Go ́mez-Gonza ́lez S, Ojeda F, Fernandes PM (2018) Portugal and Chile:longing for sustainable forestry while rising from the ashes.Environ-mental Science and Policy81, 104–107. doi:10.1016/J.ENVSCI.2017.11.006Google Inc. (2016) Google Earth Pro, 7.1.7.2606. Available at https://www.google.com/earth/http://theconversation.com/why-were-californias-wine-country-fires-so-destructive-86043 [Verified 15 October 2016]Gude PH, Jones K, Rasker R, Greenwood MC (2013) Evidence for theeffect of homes on wildfire suppression costs.International Journal ofWildland Fire22, 537–548. doi:10.1071/WF11095Haas JR, Calkin DE, Thompson MP (2013) A national approach forintegrating wildfire simulation modeling into wildland–urban interfacerisk assessments within the United States.Landscape and UrbanPlanning119, 44–53. doi:10.1016/J.LANDURBPLAN.2013.06.011Hand MS, Thompson MP, Calkin DE (2016) Examining heterogeneityand wildfire management expenditures using spatially and temporallydescriptive data.Journal of Forest Economics22, 80–102. doi:10.1016/J.JFE.2016.01.001Jin Y, Goulden ML, Faivre N, Veraverbeke S, Sun F, Hall A, Hand MS,Hook S, Randerson JT (2015) Identification of two distinct fire regimesin southern California: implications for economic impact and futurechange.Environmental Research Letters10, 094005. doi:10.1088/1748-9326/10/9/094005Keeley JE (2017) Why were California's wine country fires so destructive?The Conversation.Available at http://theconversation.com/why-were-californias-wine-country-fires-so-destructive-86043 [Verified 25 June2019]High wildfire damage in California interface WUIInt. J. Wildland FireI
Kramer HA, Mockrin MH, Alexandre PM, Stewart SI, Radeloff VC (2018)Where wildfires destroy buildings in the US relative to the wildland–urban interface and national fire outreach programs.InternationalJournal of Wildland Fire27, 329–341. doi:10.1071/WF17135Mahmoud H, Chulahwat A (2018) Unraveling the complexity of wildland–urban interface fires.Scientific Reports8, 9315. doi:10.1038/S41598-018-27215-5Maranghides A, Mell W (2012) Framework for addressing the nationalwildland–urban interface fire problem – determining fire and emberexposure zones using a WUI hazard scale. US Department of Commerce,National Institute of Standards and Technology Technical Note 1748.doi:10.6028/NIST.TN.1748McCaffrey SM, Stidham M, Toman E, Shindler B (2011) Outreachprograms, peer pressure, and common sense: what motivates home-owners to mitigate wildfire risk?Environmental Management48, 475–488. doi:10.1007/S00267-011-9704-6Mell WE, Manzello SL, Maranghides A, Butry D, Rehm RG (2010) Thewildland–urban interface fire problem – current approaches and researchneeds.International Journal of Wildland Fire19, 238–251. doi:10.1071/WF07131Monitoring Trends in Burn Severity (2016) Monitoring Trends in BurnSeverity (MTBS). Available at http://www.mtbs.gov/ [Verified 19 April2017].Moritz MA, Batllori E, Bradstock RA, Gill AM, Handmer J, HessburgPF, Leonard J, McCaffrey S, Odion DC, Schoennagel T (2014)Learning to coexist with wildfire.Nature515, 58–66. doi:10.1038/NATURE13946Nagy R, Fusco E, Bradley B, Abatzoglou JT, Balch J (2018) Human-related ignitions increase the number of large wildfires across USecoregions.Fire1, 4. doi:10.3390/FIRE1010004National Fire Protection Association (2016) Firewise communities. Avail-able at http://www.firewise.org/ [Verified 24 February 2017].Obama B (2016) Executive Order No. 13728 – wildland–urban interfaceFederal risk mitigation. (The White House: Washington, DC, USA).Oliveira S, Zeˆzere JL, Queiro ́s M, Pereira JM (2017) Assessing the socialcontext of wildfire-affected areas. The case of mainland Portugal.Applied Geography88, 104–117. doi:10.1016/J.APGEOG.2017.09.004Platt RV (2010) The wildland–urban interface: evaluating the definitioneffect.Journal of Forestry108, 9–15.Platt RV (2014) Wildfire hazard in the home ignition zone: an object-oriented analysis integrating LiDAR and VHR satellite imagery.AppliedGeography51, 108–117. doi:10.1016/J.APGEOG.2014.03.011Radeloff V, Helmers DP, Kramer HA, Mockrin MH, Alexandre PM,Bar-Massada A, Butsic VA, Hawbaker T, Martinuzzi S, Syphard AD,Stewart S (2018) Rapid growth of the US wildland–urban interfaceincreases wildfire risk.Proceedings of the National Academy of Sciencesof the United States of America115, 3314–3319. doi:10.1073/PNAS.1718850115Radeloff VC, Hammer RB, Stewart SI, Fried JS, Holcomb SS, McKeefryJF (2005) The wildland–urban interface in the United States.EcologicalApplications15, 799–805. doi:10.1890/04-1413Schoennagel T, Balch JK, Brenkert-Smith H, Dennison PE, Harvey BJ,Krawchuk MA, Mietkiewicz N, Morgan P, Moritz MA, Rasker R(2017) Adapt to more wildfire in western North American forests asclimate changes.Proceedings of the National Academy of Sciences ofthe United States of America113, 4582–4590. doi:10.1073/PNAS.1617464114Schwarzenegger A (2004) California Public Resources Code (PRC) 4291 –The requirement of a defensible space is mandated by California statelaw, Public Resources Code (PRC) 4291. (Sacramento, CA, USA).Scott JH, Burgan RE (2005) Standard fire behavior fuel models: a compre-hensive set for use with Rothermel's Surface Fire Spread Model. USDAForest Service, Rocky Mountain Research Station, General TechnicalReport RMRS-GTR-153. (Fort Collins, CO, USA).Skowronski NS, Haag S, Trimble J, Clark KL, Gallagher MR, Lathrop RG(2016) Structure-level fuel load assessment in the wildland–urbaninterface: a fusion of airborne laser scanning and spectral remote-sensing methodologies.International Journal of Wildland Fire25,547–557. doi:10.1071/WF14078Sonoma County Agricultural Preservation and Open Space District (2018)Cartographic Building Footprints. Available at https://sonomaopen-space.egnyte.com/dl/1IkmiWqACl/ [Verified 5 February 2018]Steelman T (2016) US wildfire governance as social-ecological problemEcology and Society21, 3. doi:10.5751/ES-08681-210403Stewart SI, Radeloff VC, Hammer RB, Hawbaker TJ (2007) Defining thewildland–urban interface.Journal of Forestry105, 201–207.Sugihara NG, Van Wagtendonk J, Fites-Kaufman J, Shaffer K, Thode A(2006) 'Fire in California's ecosystems.' (University of California Press:Berkeley, CA, USA)Syphard AD, Radeloff VC, Keeley JE, Hawbaker TJ, Clayton MK, StewartSI, Hammer RB (2007) Human influence on California fire regimes.Ecological Applications17, 1388–1402. doi:10.1890/06-1128.1Syphard AD, Keeley JE, Massada AB, Brennan TJ, Radeloff VC (2012)Housing arrangement and location determine the likelihood of housingloss due to wildfire.PLoS One7, e33954. doi:10.1371/JOURNAL.PONE.0033954Syphard AD, Brennan TJ, Keeley JE (2014) The role of defensible space forresidential structure protection during wildfires.International Journal ofWildland Fire23, 1165–1175.Syphard AD, Keeley JE, Pfaff AH, Ferschweiler K (2017) Human presencediminishes the importance of climate in driving fire activity across theUnited States.Proceedings of the National Academy of Sciences ofthe United States of America114, 13750–13755. doi:10.1073/PNAS.1713885114Syphard AD, Rustigian-Romsos H, Mann M, Conlisk E, Moritz MA,Ackerly D (2019) The relative influence of climate and housingdevelopment on current and projected future fire patterns and structureloss across three California landscapes.Global Environmental Change56, 41–55. doi:10.1016/J.GLOENVCHA.2019.03.007Thomas A, Leichenko R (2011) Adaptation through insurance: lessons fromthe NFIP.International Journal of Climate Change Strategies andManagement3, 250–263. doi:10.1108/17568691111153401United States Congress (2003) Healthy Forests Restoration Act. Committeeon House Agriculture; House Resources; House Judiciary; SenateAgriculture, Nutrition, and Forestry; Senate Energy and NaturalResources; House Transportation and Infrastructure. 108th Congress,1st session, 2003. Public Law no. 108–148. (Washington, DC, USA)US Census Bureau (2018) 2010 census data. Available at https://www.census.gov/quickfacts/CA [Verified 25 February 2018]USDA and USDI (2001) Urban–wildland interface communities within thevicinity of federal lands that are at high risk from wildfire.FederalRegister66, 751–777.Wildland Fire Executive Council (2014) The National Strategy: the finalphase in the development of the National Cohesive Wildland FireManagement Strategy. Available at https://www.forestsandrangelands.gov/strategy/documents/strategy/CSPhaseIIINationalStrategyApr2014.pdf [Verified 23 May 2017]Winter G, McCaffrey S, Vogt CA (2009) The role of community policies indefensible space compliance.Forest Policy and Economics11, 570–578. doi:10.1016/J.FORPOL.2009.07.004www.publish.csiro.au/journals/ijwf
_._,_._,_
Groups.io Links:
You receive all messages sent to this group.
View/Reply Online (#32371) | Reply To Group | Reply To Sender | Mute This Topic | New Topic
Mute #destroyed
Your Subscription | Contact Group Owner | Unsubscribe [volcanomadness1@gmail.com]
_._,_._,_
No comments:
Post a Comment