CSIRO PUBLISHING
International Journal of Wildland Fire 2018, 27, 781–799 https://doi.org/10.1071/WF18026
Historical patterns of wildfire ignition sources in California ecosystems
Jon E. KeeleyA,B and Alexandra D. SyphardC
AUS Geological Survey, Western Ecological Research Center, Sequoia–Kings Canyon Field Station, 47050 Generals Highway, Three Rivers, CA 93271, USA. BDepartment of Ecology and Evolutionary Biology, University of California, 612 Charles E. Young Drive East, Los Angeles, CA 90095, USA. CConservation Biology Institute, 10423 Sierra Vista Avenue, La Mesa, CA 91941, USA. DCorresponding author. Email: [email protected]
Abstract. State and federal agencies have reported fire causes since the early 1900s, explicitly for the purpose of helping land managers design fire-prevention programs. We document fire-ignition patterns in five homogenous climate divisions in California over the past 98 years on state Cal Fire protected lands and 107 years on federal United States Forest Service lands. Throughout the state, fire frequency increased steadily until a peak c. 1980, followed by a marked drop to 2016. There was not a tight link between frequency of ignition sources and area burned by those sources and the relationships have changed over time. Natural lightning-ignited fires were consistently fewer from north to south and from high to low elevation. Throughout most of the state, human-caused fires dominated the record and were positively correlated with population density for the first two-thirds of the record, but this relationship reversed in recent decades. We propose a mechanistic multi-variate model of factors driving fire frequency, where the importance of different factors has changed over time. Although ignition sources have declined markedly in recent decades, one notable exception is powerline ignitions. One important avenue for future fire-hazard reduction will be consideration of solutions to reduce this source of dangerous fires.
Additional keywords: arson, debris burning, equipment, lightning, powerlines, smoking.
Received 20 February 2018, accepted 25 September 2018, published online 7 November 2018
Introduction illustrates the complexities of ascertaining relationships Increasing concern over wildfires has prompted a re-emphasis between different sources and how they change over time. by the federal government to stop this trend (https://www.doi. An important characterisation of anthropogenic ignitions is gov/pressreleases/secretary-zinke-directs-interior-bureaus-take- that the most abundant ignition sources are not always associ- aggressive-action-prevent-wildfires, accessed 1 April 2018; ated with the greatest area burned (Syphard and Keeley 2016). Bedard 2017). For many decades, the focus of fire management Thus, a topic in need of further study is how to sort out those has been on fuel modification with success on certain landscapes ignition sources that are most damaging, how those have (Kalies and Kent 2016) but limited improvement on others changed over time, and in light of future needs, how climate (Keeley and Safford 2016). Because humans are a dominant change is likely to affect different ignition sources and losses. ignition source over the majority of North America (Balch et al. For example, it has been demonstrated for the state of Victoria, 2017; Syphard et al. 2017) there is reason to believe improve- Australia, that some ignition sources, such as electrical distribu- ments in fire prevention may be a key to reducing fire impacts. tion lines, may be limited in number but result in much more Indeed, the United States Forest Service (USFS) has been severe fire consequences (Miller et al. 2017). In addition, these reporting fire causes since it began collecting systematic data on fires are more likely during periods of elevated fire danger. If fires in 1905, with the explicit purpose of helping land managers some ignition sources play a larger role in area burned, these design fire prevention programs (Donoghue 1982a). might be targets for closer scrutiny and fire-management plan- Effective fire-prevention requires a sound understanding of ning. This potential has been demonstrated for parts of southern the patterns and causes of fire ignitions, which are closely California over recent decades, where powerlines have been aligned with both human and biophysical-landscape character- shown to cause a substantial amount of area burned in both istics (Syphard et al. 2008). Prestemon et al. (2013) suggested a subregions in southern California (Syphard and Keeley 2015). conceptual model that linked ignitions to changes in biophysi- Other important factors were arson in one subregion and cal, societal, prevention and management variations that equipment in another.
Journal Compilation Ó IAWF 2018 Open Access CC BY-NC-ND www.publish.csiro.au/journals/ijwf 782 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
The goal of the present research is to expand that approach to USFS fire data covered 17 national forests (Fig S2) and include the entire written history of fires in the state. Our focus included the years 1910–2016 (see Table S2, two forests were was on the spatial and temporal patterns of different ignition created after 1910 carved out of area from adjoining forests). sources and the relationship between type of ignition and area Area protected has changed through this period of record and burned. We took a long-term historical approach utilising data thus data were normalised to the hectares protected for that year. from 1910 to 2016 for USFS lands and from 1919 to 2016 for Data for USFS lands through the 1980s were from annual fire state protected Cal Fire lands. First, we examine the spatial and statistics reports for Region 5, available in the Forestry Library temporal patterns of natural lightning-ignited fires v. human (most of which was transferred to the Bioscience Library) at fires in the state, and their contribution to area burned. Next, we the University of California, Berkeley. More-recent data, 1970 investigated which anthropogenic causes are most frequent, to 2016, were from the National Wildfire Coordinating Group their distribution within the state, their change over time and (see http://fam.nwcg.gov/fam-web/weatherfirecd/state_data. their contribution to area burned. Based on a study of state- htm, accessed 1 June 2017). protected lands in California, Syphard et al. (2007) found that Most investigators are unfamiliar with these historical fire fires increased from 1931 to the 1980s, but then decreased over records and are sometimes sceptical of their accuracy. For the subsequent decades. A similar pattern for the whole state was example, Stephens (2005) contended that USFS data before also reported by Keeley and Syphard (2017) on both Cal Fire and 1940 were inaccurate, but cited a source (Mitchell 1947) that USFS lands. Thus, the present study contrasts fire-ignition provided no evidence of this. Likely, the idea comes from patterns within climatically homogenous sub-regions for the Donoghue’s (1982b) comment ‘1940 marked the modern era period before 1980 and for the period 1980–2016.We also of fire reporting’. However, that comment was in reference to investigated the extent to which seasonal climate parameters the fact that ‘the report issued at this time was the first designed could explain patterns of ignitions and area burned for each type for automated data processing and easy readability’ and was not of ignition source. in reference to reliability. Historians have generally been confidant in these early California fire records (Brown 1945; Show 1945; Clar 1969; Methods Cermak 2005). The first author, J. E. Keeley, examined all of the Fire-history data for numbers of fires and area burned, by cause, California fire-related materials stored at the state and federal were analysed separately for state-protected Cal Fire and federal archives and believes collectively they show managers have USFS lands. Data for counties, forests and climate divisions always been conscientious about reporting accuracy and were all normalised by the area protected each year and within completeness. For example, beginning in 1905, USFS record- each unit and expressed as number of fires, or hectares burned, keeping required 15 items of information on the fire reporting per million hectares. Form 944, including the specific cause (Donoghue 1982b). Cal Fire data included 51 of the state’s 58 counties (see Fig On state-protected lands there was an incentive in that the S1, available as Supplementary material to this paper) as 7 1911 Federal Weeks Law provided fiscal aid to states based counties had limited fire activity or records. Fire statistics were on statistics of fire protection (see http://www.calfire.ca.gov/ from direct protection areas (DPA), which are mostly state- about/about_calfire_history2, accessed 23 May 2018). In 1919, responsibility lands with smaller amounts of federal lands, and the California state legislature appropriated money for fire included the years 1919–2016, summarised by county. The term prevention and suppression, and records in the state archive DPA was first used in 1986 and the area included was equivalent show that, by 1920, there were more than 400 fire wardens to what was called State Zone (1972–1985), Zones I and II distributed throughout the state who were charged with fire- (1945–1971) and Zones 1, 2, 3 (1919–1944). Cal Fire data from fighting and fire reporting. In 1920, there were 800 flights of the 1919 to 1930 are unpublished and were only available as typed Army’s 9th Aero Squadron fire patrol that covered 426 500 km reports at the California State Archives in Sacramento. Data during the 5-month California fire season (Cermak 1991). from 1931 to 2016 were available in annual reports variously One complication in studying ignition sources is that named, Forest Fire Summary, Fire Statistics, Fire Activity reported categories have changed over time. Certain causes Statistics, and Wildfire Activity Statistics, often referred to as have persisted over the entire period of record, including the Redbook series, available from research libraries or directly lightning, smoking and camping, but other categories have from the agency. Only 30 counties had complete data (excluding changed their names. For example, arson fires are a relatively 1927 for nearly all counties and a few additional years in other new category as intentionally set fires have, in the past, been counties) beginning in 1919, and an additional 21 counties had labelled as ‘incendiary’, and seem to have been more rural than continuous data beginning in 1945 (or slightly later in a few contemporary urban arson fires (Kuhlken 1999) and in this paper cases) (see Table S1 for years of records for each county). Area they are all recorded as arson fires. Other changes include the protected has changed through this period of record and thus term ‘brush burning’ being changed to ‘debris burning’, and the data were normalised to the hectares protected for that year categorisation of brush burning has been folded into debris presented with the annual reports. There was a period from 1941 burning. Causes that were unknown or represented minor to 1952 area where protected-area data were not included in categories have been included as miscellaneous fires (Donoghue annual reports and, as best we can determine, those data are no 1982a), but are not addressed here. longer available, so we used the areas protected in 1940. In all Cal Fire data were spatially explicit at the level of the county cases, the changes between 1940 and 1953, when such data were and USFS data at the level of the individual forest. However, for again available, were minor. analysis, these were grouped into climatically homogenous Fire ignition sources Int. J. Wildland Fire 783
areas as defined by the National Oceanic and Atmospheric precisely, counties or forests were placed in the climate division Administration’s (NOAA) National Climatic Data Center comprising the majority of land area in that unit. (NCDC) California Climate Divisions (Fig. 1), comprising the In 1919, Cal Fire-protected lands were 11.7 106 ha and main fire-prone landscapes in the state (see http://www.ncdc. increased to 12.5 106 ha in 2016. USFS lands comprised noaa.gov/temp-and-precip/time-series/index.php?parameter=pd- 9.8 106 ha in 1919 and decreased to 9.5 106 ha in 2016. si&month=1&year=2008&filter=p12&state=4&div=5, accessed Vegetation on state lands was dominated by grasslands and 15 June 2016). These include, from north to south, Division 1 shrublands in the south and with significant woodlands and (North Coast), 2 (North Interior), 5 (Sierra Nevada), 4 (Central coniferous forests farther north (see Keeley and Syphard 2017 Coast), and 6 (South Coast). Where boundaries did not match for more detailed vegetation data). USFS lands were dominated by coniferous forests, except in the southern part of the state where they were dominated by shrublands. To evaluate climate impact on fire activity, we utilised PRISM climate for each county on Cal Fire-protected lands and each forest on USFS lands (Fig. 1). For every year in the analysis, we extracted 2.5 arc-minute PRISM data (PRISM Climate Group, Oregon State University, see http://prism. oregonstate.edu, accessed 15 February 2017) for areas within the boundaries of the Cal Fire and USFS lands. For each county and forest, we computed area-weighted averages of monthly mean precipitation and temperature, summarised by season – winter being December (prior year), January and February, spring being March, April and May, summer being June, July and August, and autumn being September, October and November. Analysis was conducted with Systat software (ver. 11.0, Systat Software, Inc., San Jose, CA, http://www.systat.com/). For the climate analysis, we developed multiple regression models explaining area burned for USFS and Cal Fire based on seasonal temperature, precipitation and prior-season precipi- tation variables. To ensure multicollinearity would not be an issue, we calculated correlation coefficients among all potential explanatory variables and eliminated those that were strongly correlated (P , 0.05) with other variables in the model.
Results Long-term averages show that on a per-unit-area basis fires were N approximately twice as frequent or more on Cal Fire lands as on USFS lands in all five NOAA climate divisions (Table 1). However, the relationship between ignitions and area burned varied markedly between Cal Fire and USFS lands and between 0 45 90 180 270 360 divisions. In the North Coast division, Cal Fire dealt with twice Kilometers as many fires as the USFS but the average area burned was very Fig. 1. NOAA climate divisions and Cal Fire protected and USFS similar. In contrast, in the interior from the Sierra Nevada protected lands in California for the five climate divisions with long-term northward, Cal Fire experienced approximately double the fire history. number of fires and nearly double the area burned. In the coastal
Table 1. Fire frequency and area burned on state and federal lands in California
Cal Fire (1919–2016) USFS (1910–2016) NOAA division Fire frequency Area burned Fire frequency Area burned (n/year/106 ha) (ha/year/106 ha) (n/year/106 ha) (ha/year/106 ha)
North Coast 317 7780 150 7559 North Interior 421 9642 207 5914 Sierra Nevada 356 8436 169 4709 Central Coast 277 5496 66 18860 South Coast 656 15278 369 24442 784 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
Table 2. Cal Fire counties total fires, percentage due to human ignitions and regression coefficients for population density v. number of fires (per year per million ha) for years 1919–2016
Division County Total Percentage human ,1980 $1980 rPrP
North Coast Del Norte 330 97 0.39 0.007 0.07 0.667 Humboldt 219 93 0.31 0.018 0.62 0.000 Lake 446 98 0.72 0.000 0.66 0.000 Marin 1155 100 0.10 0.526 0.07 0.699 Mendocino 255 94 0.46 0.000 0.46 0.004 Napa 493 99 0.79 0.000 0.83 0.000 Siskiyou 256 63 0.44 0.000 0.07 0.681 Sonoma 488 99 0.90 0.000 0.71 0.000 Trinity 229 75 0.05 0.709 0.27 0.102 North Interior Butte 929 95 0.89 0.000 0.51 0.001 Colusa 115 95 0.44 0.003 0.60 0.000 Glenn 87 93 0.28 0.069 0.64 0.000 Lassen 178 50 0.55 0.000 0.39 0.017 Modoc 113 51 0.42 0.020 0.04 0.811 Nevada 936 97 0.72 0.000 0.80 0.000 Placer 1645 98 0.88 0.000 0.89 0.000 Plumas 489 74 0.67 0.000 0.08 0.654 Shasta 469 90 0.77 0.000 0.38 0.020 Solano 406 99 0.55 0.000 0.54 0.000 Tehama 196 93 0.90 0.000 0.48 0.000 Yolo 226 97 0.61 0.000 0.32 0.051 Yuba 723 97 0.49 0.000 0.26 0.118 Sierra Nevada Amador 507 99 0.60 0.000 0.47 0.003 Calaveras 337 99 0.59 0.000 0.57 0.000 El Dorado 213 98 0.77 0.000 0.26 0.123 Fresno 100 96 0.84 0.000 0.69 0.000 Inyo-Mono 255 98 0.67 0.002 0.63 0.000 Kern 804 99 0.84 0.000 0.51 0.002 Kings 321 99 0.29 0.146 0.32 0.107 Madera 823 99 0.65 0.000 0.48 0.003 Mariposa 617 97 0.70 0.000 0.45 0.005 Merced 602 95 0.21 0.180 0.42 0.010 San Joaquin 850 97 0.05 0.789 0.46 0.005 Stanislaus 237 95 0.35 0.024 0.20 0.244 Tulare 180 93 0.77 0.000 0.33 0.048 Tuolumne 280 93 0.89 0.000 0.23 0.174 Central Coast Alameda 117 98 0.65 0.000 0.68 0.000 Contra Costa 447 93 0.59 0.000 0.21 0.213 Monterey 306 93 0.84 0.000 0.83 0.000 San Benito 151 93 0.62 0.000 0.81 0.000 San Luis Obi 322 99 0.77 0.000 0.64 0.000 San Mateo 158 97 0.74 0.000 0.08 0.055 Santa Clara 242 93 0.38 0.002 0.63 0.000 Santa Cruz 814 97 0.80 0.000 0.57 0.000 South Coast Los Angeles 778 98 0.41 0.007 0.53 0.001 Orange 1198 100 0.85 0.000 0.62 0.000 Riverside 790 98 0.82 0.000 0.82 0.000 San Bernardi 791 96 0.73 0.000 0.81 0.000 San Diego 576 97 0.63 0.000 0.60 0.000 Santa Barbar 347 99 0.70 0.000 0.64 0.000 Ventura 713 99 0.84 0.000 0.42 0.010
areas from San Francisco to San Diego, there were substantially For Cal Fire-protected landscapes, the area-based average more fires on Cal Fire-protected lands but the area burned was number of fires per year (1919 to 2016) varied from 1645 in substantially greater on USFS lands. Placer County to 87 in Glenn County (Table 2). Humans were Fire ignition sources Int. J. Wildland Fire 785
(a) (b)(c) (d ) Lightning Arson Debris Smoking 2.7–3.4 2.4–5.8 2.7–8.0 3.3–9.6 3.5–4.2 5.9–11.2 8.1–15.5 9.7–15.9 4.3–4.7 11.3–23.0 15.6–26.3 16.0–23.7 4.8–7.2 23.1–39.0 26.4–32.1 23.8–32.2 7.3–10.6 39.1–47.0 32.2–44.2 32.3–41.0 10.7–15.2 47.1–62.1 44.3–55.9 41.1–58.0 15.3–22.0 62.2–80.1 56.0–73.8 58.1–74.1 22.1–32.9 80.2–97.9 73.9–125.9 74.2–103.0 33.0–56.5 98.0–162.2 126.0–189.2 103.1–137.0 56.6–126.9 162.3–261.6 189.3–250.2 137.1–215.8
(e)(f ) (g)(h) Playing Equipment Vehicle Powerline 0.9–3.0 14.9–27.2 2.7–7.3 0.4–2.0 3.1–8.5 27.3–43.2 7.4–14.8 2.1–3.2 8.6–12.6 43.3–57.2 14.9–20.8 3.3–4.6 12.7–19.5 57.3–81.5 20.9–23.5 4.7–5.8 19.6–24.8 81.6–97.5 23.6–40.8 5.9–7.8 40.9–47.7 24.9–31.2 97.6–118.9 7.9–9.6 47.8–64.7 9.7–13.8 31.3–54.6 119.0–162.2 64.8–85.9 13.9–18.2 54.7–79.6 162.3–205.8 86.0–120.6 18.3–22.9 79.7–136.8 205.9–259.4 120.7–201.3 23.0–35.2 136.9–270.2 259.5–528.9
Fig. 2. Fire frequency for different ignition sources on Cal Fire protected lands in California for the years 1919–2016 (n/year/106 ha); note change in scales for each source. responsible for most of fires, accounting for 95% or more of the most common in the north-east part of the state and declined ignitions in two-thirds of the counties. However, certain north- markedly in coastal central and southern California (Fig. 4a). ern California counties stood out as notable exceptions, e.g. in Area burned by lightning-ignited fires generally followed a Siskiyou, Trinity, Lassen, Modoc and Plumas counties lightning similar pattern with the exception that parts of the North Coast accounted for one-quarter to more than half of all ignitions, and Central Coast, despite having few such ignitions, had patterns illustrated in Fig. 2a. Regions with the lowest lightning- substantial area burned by this source (Fig. 5a). ignited fires extended through the coastal ranges from north of On both Cal Fire- and USFS-protected lands, humans played San Francisco to Santa Barbara. Area burned by lightning- a substantial role in fire ignitions. During the first two-thirds of ignited fires generally followed a similar pattern, although it the 20th century there was a very strong positive relationship was the source for significant burning in the San Bernardino between population density and fire frequency in nearly 90% of County of southern California (Fig. 3a). the counties (Table 2) and more than 75% of the forests For USFS lands, the area-based average number of fires per (Table 3). However, from 1980 to 2016, although population year (1910 to 2016) varied from 478 in San Bernardino to 67 in growth continued throughout the state, in most counties and Eldorado National Forest (Table 3). Humans accounted for far forests, population density exhibited a highly negative relation- fewer fires than on Cal Fire lands. In the South Coast division, ship with fire frequency (Tables 2, 3). humans were responsible for 74–88%; however, in half of the On both Cal Fire- and USFS-protected lands, human-ignited other forests, humans accounted for less than 50% of the fires. fires derived from both intentional and accidental causes. The As with Cal Fire landscapes, USFS lightning-ignited fires were highest number of fires was from equipment, arson, debris 786 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
(b) (c)(d ) (a) Lightning Arson Debris Smoking
0.8–14.1 0.6–206.0 5.7–69.0 5.7–171.6 14.2–38.0 206.1–521.6 69.1–198.9 171.7–386.4 38.1–85.2 521.7–769.1 199.0–393.2 386.5–613.1 393.3–505.8 85.3–177.7 769.2–1402.3 613.2–1049.0 1402.4–1929.2 505.9–621.9 177.8–282.5 1049.1–1190.7 1929.3–3040.5 622.0–723.9 282.6–486.3 1190.8–1394.0 3040.6–3871.0 724.0–889.2 1394.1–2074.5 486.4–693.5 3871.1–4830.5 889.3–1206.0 2074.6–2762.0 693.6–915.8 4830.6–6519.3 1206.1–1788.5 2762.1–3463.1 915.9–1448.1 6519.4–12775.1 1788.6–2436.8 3463.2–6316.6 1448.2–2059.2
(e) (f ) (g) (h) Playing Equipment Vehicle Powerline 0.6–3.1 0.0–8.1 0.3–3.1 0.7–117.2 3.2–15.5 8.2–21.5 3.2–7.6 117.3–286.7 15.6–36.8 21.6–38.7 7.7–11.6 286.8–412.7 36.9–90.6 38.8–95.5 11.7–23.9 412.8–523.9 95.6–174.8 524.0–649.4 90.7–156.5 24.0–46.2 174.9–263.8 649.5–877.7 156.6–303.9 46.3–79.3 263.9–658.2 877.8–1049.0 304.0–516.8 79.4–110.1 658.3–2060.9 1049.1–1353.1 516.9–1624.9 110.2–191.4 2061.0–3619.7 1353.2–3013.5 1625.0–3065.9 191.5–339.4 3619.8–5551.4 339.5–953.1 3013.6–4671.9 3066.0–4585.1
Fig. 3. Area burned by different ignition sources on Cal Fire protected lands in California for the years 1919–2016 (ha burned/year/106 ha); note change in scales for each source. burning, children playing with fire, smoking, vehicles and (Fig. 5c, f). Forests adjacent to high density metropolitan areas in powerlines (Tables S3–S6). Sources, such as railroads and Los Angeles and western San Bernardino counties had substan- lumber practices, did cause many fires in the early part of the tial burning due to smoking, children playing with fire, and record, but are of minor significance today (this change not powerlines (Fig. 5d, e, h). shown). These ignition sources exhibited marked geographical varia- tion in their importance (Fig. 2–5). On lower elevation Cal Fire Changing ignition patterns over time landscapes, arson was responsible for much of the area burned in The historical pattern of fire frequency on lower elevation Cal the northern Sierra Nevada Mountains. (Fig. 3b), whereas debris Fire-protected lands for 97 years and USFS lands for 107 years is burning was responsible for much of the area burned north and illustrated in Fig. 6 for the five climate divisions. There was a south of San Francisco (Fig. 3c), vehicles were the cause of common pattern across both Cal Fire and USFS lands and much of the burning in the central coastal ranges (Fig. 3g) and consistent within each of the five climate divisions – a highly equipment fires in southern California (Fig. 3f). Powerlines significant increase in fire frequency from the beginning of were responsible for significant number of fires in the north records to 1979, and a switch to a highly significant decline in bay area of San Francisco and coastal communities from Santa fires from 1980 to 2016 (Fig. 6), the single exception being the Barbara south to the border (Fig. 3h). South Coast USFS lands (Fig. 6r). Despite a significant fit of In USFS forests, area burned in the South Coast was most these data to the linear regression models, there were some heavily affected by arson and powerlines (Fig. 5b, h), but marked departures on Cal Fire lands during the early record. equipment and debris burning dominated in the Central Coast Plotting of linear regression residuals from 1919 to 1979 shows Fire ignition sources Int. J. Wildland Fire 787
Table 3. USFS forests total number of fires, percentage ignited by humans and regression coefficients for population density v. number of fires (per year per million ha) for years 1910–2016
Division Forest Total Percentage human ,1980 $1980 rPrP
North Coast Klamath 192 28 0.23 0.057 0.33 0.045 Mendocino 97 51 0.26 0.032 0.36 0.030 Six Rivers 137 65 0.46 0.007 0.11 0.528 North Interior Lassen 271 38 0.66 0.000 0.43 0.008 Modoc 118 21 0.42 0.000 0.06 0.730 Plumas 280 46 0.36 0.002 0.37 0.024 Shast-Trinity 188 50 0.15 0.142 0.41 0.012 Tahoe 257 54 0.22 0.067 0.44 0.007 Sierra Nevada Eldorado 67 58 0.43 0.000 0.25 0.129 Inyo-Mono 237 40 0.85 0.000 0.67 0.000 Sequoia 79 39 0.75 0.000 0.76 0.000 Sierra 213 48 0.63 0.000 0.68 0.000 Stanislaus 201 49 0.50 0.000 0.55 0.000 Central Coast Los Padres 203 82 0.51 0.000 0.50 0.002 South Coast Angeles 367 87 0.58 0.000 0.49 0.002 Cleveland 292 88 0.71 0.000 0.05 0.773 Sbernardino 478 74 0.89 0.000 0.50 0.002
(a) Lightning (b) Arson (c) Debris (d ) Smoking
11.7 0.9 1.6 7.3–8.3 11.8–36.7 1.0–8.5 1.7–3.3 8.4–11.5 36.8–48.2 8.6–8.9 3.4–4.7 11.6–13.1 48.3–93.2 9.0–9.9 4.8–6.2 13.2–17.9 93.3–93.6 10.0–10.7 6.3–7.7 18.0–20.2 93.7–99.8 10.8–12.0 7.8–8.0 20.3–24.7 99.9–103.2 12.1–13.1 8.1–11.9 24.8–29.7 103.3–127.6 13.2–18.8 12.0–12.6 29.8–31.3 127.7–149.5 18.9–45.1 12.7–13.7 31.4–33.7 149.6–165.1 45.2–51.9 13.8–17.0 33.8–50.5
(e) (f ) (g) (h) Playing Equipment Vehicle Powerline 2.5–3.1 1.6 0.4 0.3–0.4 3.2–5.6 0.5–2.1 1.7–3.3 0.5–0.9 5.7–5.9 2.2–3.6 3.4–5.1 1.0 6.0 3.7–4.2 5.2 1.1–1.3 6.1–7.1 4.3–5.1 5.3–6.0 1.4–1.7 7.2–8.7 5.2–5.8 6.1–6.9 1.8–1.9 8.8–11.8 5.9–8.0 7.0–8.2 2.0–2.2 11.9–12.9 8.1–9.7 8.3–10.2 2.3–2.5 13.0–25.1 9.8–40.7 10.3–54.9 2.6–6.6 25.2–53.6 40.8–59.0 55.0–95.4 6.7–10.4
Fig. 4. Fire frequency for different ignition sources on USFS protected lands in California for the years 1910–2016 (n/year/106 ha); note change in scales for each source. 788 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
(a)(Lightning b)(Arson c)(Debrisd ) Smoking
114.5 98.8 37.5–55.2 34.2–84.7 114.6–726.5 98.9–364.5 55.3–96.5 84.8–143.4 726.6–1378.6 364.6–558.0 96.6–104.2 143.5–373.9 1378.7–1634.9 558.1–773.2 104.3–115.8 374.0–473.6 1635.0–1738.0 773.3–1041.0 115.9–191.1 473.7–529.2 1738.1–1920.2 1041.1–1147.8 191.2–258.8 529.3–573.6 1920.3–2355.4 1147.9–1205.0 258.9–330.7 573.7–865.6 2355.5–3103.0 1205.1–3609.6 330.8–469.8 865.7–1245.5 3103.1–3816.3 3609.7–5990.1 469.9–859.7 1245.6–1465.6 1465.7–2828.6 3816.4–5525.1 5990.2–6830.1 859.8–1651.2
(e) Playing(f ) Equipment(g) Vehicle(h) Powerline
0.4–0.6 44.1 13.5–40.0 1.7 0.7–1.6 44.2–50.6 40.1–88.2 1.8–2.3 1.7–2.2 50.7–98.4 88.3–114.5 2.4–5.7 2.3–4.0 98.5–149.2 114.6–142.2 5.8–27.5 4.1–6.2 149.3–209.5 142.3–384.1 27.6–40.0 40.1–49.6 6.3–10.1 209.6–287.7 384.2–485.0 49.7–135.6 10.2–32.7 287.8–498.4 485.1–548.8 135.7–353.4 32.8–254.1 498.5–952.3 548.9–633.4 353.5–1157.6 254.2–324.4 952.4–1353.7 633.5–808.1 1157.7–15377.3 324.5–2539.6 1353.8–5006.0 808.2–1484.6
Fig. 5. Area burned by different ignition sources on USFS protected lands in California for the years 1910–2016 (ha burned/year/106 ha); note change in scales for each source. marked and consistent diversions in most divisions (Fig. 7). North Interior and Sierra Nevada) and in the south (Central Although the residuals early in the record are closely aligned Coast and South Coast), which is justified by the marked with the regression line, during the 1920s and 1930s from the similarities in ignition patterns within these two regions Sierra Nevada north there was a marked increase in ignitions. (see Fig. 2–5). This pattern was less obvious in coastal central and southern On Cal Fire-protected lands, it is noteworthy that changes in California. In the 1950s and 1960s, there was a marked number of ignitions for lightning-ignited fires matched that of depression in ignitions in all climate divisions. It is worth noting many human ignition sources, specifically increased ignitions that in the former period it was drier than the long-term average during the first part of the record and decreased ignitions in and in the latter period wetter (Fig. S3). recent decades (Fig. 8a, e). Numbers of lightning ignitions were Changes in area burned did not closely follow changes in fire more than double in the north than in the south, and substantially frequency (Fig. 6) – while fire frequency increased in the first fewer than the leading anthropogenic causes. On USFS forests, three-quarters of the 20th century, area burned declined or lighting fire frequency (Fig. 9a, e) followed a temporal pattern stayed more or less constant. USFS forests in the northern part similar to Cal Fire lands but were ,3 times more abundant than of the state showed a tendency for increased area burned in the on Cal Fire landscapes and were one of the dominant ignition last 4 decades (Fig. 6d, l) but in general there were no strong sources in forests. Despite changes in number of lightning- trends in area burned after 1980. ignited fires, the area burned by this source did not exhibit Of particular interest is how specific ignition sources consistent trends, although, in northern California forests, area have changed and, in order to simplify this presentation, we burned by lightning-ignited fires has increased since 1980 have consolidated climate divisions in the north (North Coast, (Fig. 8i, m, 9i, m). Fire ignition sources Int. J. Wildland Fire 789
Cal Fire USFS Cal Fire USFS (a)(b)(c)(d) 1000 North Coast500 North Coast 6 North Coast 6 North Coast 2 2 2 2 r 0.29 800 r 0.54 400 r 0.10 5 r 0 5 2 2 P 0.001 r 0.65 P 0.001 r 0.07 P 0.050 P 0.715 600 P 0.001 300 P 0.023 4 4
400 200 3 3 2 2 200 100 2 r 0.31 2 r 0.29 P 0.001 P 0.001 0 0 1 1 (e)(f )(g)(h) 1000 North Interior 500 North Interior 6 North Interior6 North Interior 2 2 2 2 r 0.09 800 r 0.60 400 r 0.25 5 r 0 5 2 2 r 0.69 P 0.001 r 0.21 P 0.002 P 0.800 P 0.070 600 P 0.001 300 P 0.001 4 4
400 200 3 3 2 2 200 100 2 r 0.28 2 r 0.19 P 0.001 P 0.001 0 0 1 1 (i)(j)(k)(l) 1000 Sierra Nevada 500 Sierra Nevada6 Sierra Nevada6 Sierra Nevada 2 2 2 2 800 r 0.62 400 r 0.50 5 r 0 5 r 0.14 P 0.001 P 0.001 P 0.924 P 0.021 2 2 600 r 0.64 300 r 0.63 4 4 P 0.001 P 0.001 400 200 3 3 2 2 r 0.34 r 0.16 200 100 2 2 P 0.001 P 0.001 0 0 1 1
(m)(n)(Area burned (log ha/yr/million ha) o)(p) Fire frequency (#/yr/million ha) 1000 Central Coast 500 Central Coast6 Central Coast6 (p) Central Coast 2 2 800 r 0.68 400 5 r 0.03 5 P 0.001 P 0.351 600 2 300 2 4 4 r 0.56 r 0.10 400 P 0.001 P 0.009 2 200 r 0.29 3 3 2 P 0.001 r2 2 r 0 200 100 2 0.39 2 r 0.05 P 0.001 P 0.058 P 0.878 0 0 1 1 (q)(r)(s)(t ) 2000 South Coast 1000 South Coast 6 South Coast6 South Coast 2 2 1600 r 0.80 800 r 0.04 5 5 2 P 0.001 2 P 0.254 r 0.49 r 0.58 1200 600 4 4 P 0.001 P 0.001 800 400 3 3 r2 0.03 2 2 2 400 200 2 r 0.03 2 r 0 r 0 P 0.130 P 0.299 P 0.973 P 0.985 0 0 1 1 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
Fig. 6. Fire frequency for Cal Fire protected lands (a, c, i, m, q) and USFS lands (b, f, j, n, r) and area burned on Cal Fire (c, g, k, o, s) and USFS (d, h, l, p, t) lands. Note for frequency the change in scale between the South Coast and other divisions and between Cal Fire and USFS lands.
North Coast North Interior Sierra Nevada Central Coast South Coast 3 3 3 3 3
2 2 2 2 2 1 1 1 1 1 0 0 0 0 0 1 1 1 2 1 1
Standardized residual 2 2 3 2 2
19101920193019401950196019701980 19101920193019401950196019701980 19101920193019401950196019701980 19101920193019401950196019701980 19101920193019401950196019701980 Year
Fig. 7. Ignition sources recorded throughout the period 1919–2016 on Cal Fire protected lands by frequency (a–h) and area burned (i–p) in the north (climate divisions North Coast, North Interior, and Sierra Nevada) and the south (Central Coast and South Coast) with lines for significant regressions from 1919 to 1979 and from 1980 to 2016; note the change in scale between lightning and anthropogenic sources for fire frequency. 790 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
(a) 100 North - Lightning(b)(200 North - Arsonc)(200 North - Debrisd) 200 North - Smoking
2 80 160 160 160 2 r 0.01 2 2 2 2 r 0.01 2 r 0.54 r 0.17 r 0.29 r 0.74 r 0.72 P 0.491 60 120 120 P 0.579 120 2 P 0.001 P 0.011 P 0.001 P 0.001 P 0.001 r 0.72 P 0.001 40 80 80 80
20 40 40 40
0 0 0 0 (e)100 (f)(200 g)(200 h) 200 South - Lightning South - Arson South - Debris South - Smoking 80 160 160 160 2 2 r 0.38 r 0.29 2 2 2 2 2 2 60 P 0.001 P 0.001 120 r 0.48 r 0.29120 r 0.09 r 0.71 120 r 0.02 r 0.38 P 0.001 P 0.001 P 0.020 P 0.001 P 0.001 Fire frequency (#/yr/million ha) P 0.348 40 80 80 80
20 40 40 40
0 0 0 0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
(i) 5.0 North - Lightning(j) 5.0 North - Arson(k) 5.0 North - Debris(l) 5.0 North - Smoking
2 2 2 2 2 4.2 r 0.06 r 0.01 4.2 r 0.14 4.2 r 0.32 4.2 r 0.61 P 0.066 P 0.590 P 0.021 P 0.001 P 0.001 3.4 3.4 3.4 3.4
2.6 2.6 2.6 2.6 2 2 2 1.8 1.8 r 0.09 1.8 r 0.33 1.8 r 0.34 P 0.030 P 0.001 P 0.001 1.0 1.0 1.0 1.0 (m) 5.0 South - Lightning(n)(5.0 South - Arsono)(5.0 South - Debrisp) 5.0 South - Smoking
2 2 2 2 4.2 r 0.12 r 0.06 4.2 r 0.44 4.2 r 0.09 4.2 P 0.010 P 0.152 P 0.001 P 0.075 2 3.4 3.4 3.4 3.4 r 0.27 P 0.001 2.6 2.6 2.6 2.6
2 Area burned (log ha/yr/million ha) 2 2 1.8 1.8 r 0.04 1.8 r 0.571.8 r 0.22 P 0.147 P 0.001 P 0.001 1.0 1.0 1.0 1.0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
Fig. 8. Ignition sources not reported separately before 1960 on Cal Fire protected lands by frequency (a–h) and area burned (i–p) in the north (climate divisions North Coast, North Interior, and Sierra Nevada) and the south (Central Coast and South Coast) with lines for significant regressions before 1980 and 1980–2016; note the change in scale for fire frequency.
The main anthropogenic ignition sources on Cal Fire lands years (Fig. 10, 11). Children playing with fire declined were arson, debris burning and smoking, and all showed significantly in both jurisdictions in the north and south a significant decrease in recent decades (Fig. 8b–d, f–h). Also, (Fig. 10a, b, 11b, f) as did area burned by this source area burned by these ignition sources mostly showed a marked (Fig.10i, m, 11j, n). Equipment-ignited fires increased decrease in recent decades (Fig. 8j–l, n–p). markedly between 1960 and 1979 on Cal Fire lands On USFS lands, arson and smoking were very important but (Fig. 10b, f) but, during the same period, declined on USFS camping was also a significant cause (Fig. 9). Arson fires lands (Fig. 11c, g). Since 1980, this source of ignitions has exhibited remarkable similarity in the south of both jurisdictions declined sharply on Cal Fire lands in the north and south with a marked decline in frequency and area burned since 1980 (Fig. 10b, f) but increased on USFS lands in the south (Fig. 9f, n). (Fig. 11g). In contrast to all other ignition sources, powerline On both Cal Fire and USFS lands, some ignition sources, fires on Cal Fire and USFS lands in both the north and south such as children playing with fire, equipment, vehicles and have not declined in the last 4 decades (Fig. 10d, h, 11d, h)nor powerlines, were not specifically recorded during the early has area burned by this ignition source (Fig. 10p, 11p). Fire ignition sources Int. J. Wildland Fire 791
(a) 300 North - Lightning(b)(150 North - Arsonc)(75 North - Campingd)75 North-Smoking
240 120 60 60 2 2 2 2 2 r 0.04 2 2 2 r 0.40 r 0.22 r 0.18 r 0.65 r 0.07 r 0.08 r 0.86 180 90 45 P 0.124 45 P 0.001 P 0.004 P 0.001 P 0.001 P 0.110 P 0.019 P 0.001 120 60 30 30
60 30 15 15
0 0 0 0 (e) 300 (f)(150 g)(75 h) 75 South - Lightning South - Arson South - Camping South - Smoking 240 120 60 60 2 2 r 0.37 r 0.06 2 2 2 2 2 2 r 0.64 180 P 0.001 P = 0.050 90 r 0.45 45 r 0.03 r 0.07 45 r 0.32 r 0.76 P 0.001 P 0.152 P 0.124 P < 0.001 P 0.001 Fire frequency (#/yr/million ha) P 0.001 120 60 30 30
60 30 15 15
0 0 0 0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
(i) 5.0 North - Lightning(j)(5.0 North - Arsonk)(5.0 North - Campingl)5.0 North - Smoking
2 2 2 2 2 4.2 r 0.04 r 0.21 4.2 r 0.08 4.2 r 0.02 4.2 r 0.52 P 0.090 P 0.005 P 0.100 P 0.436 P 0.001 3.4 3.4 3.4 3.4
2.6 2.6 2.6 2.6
2 2 2 1.8 1.8 r 0.47 1.8 r 0.23 1.8 r 0.24 P 0.001 P 0.001 P 0.001 1.0 1.0 1.0 1.0
(m)(5.0 n)(5.0 o)(5.0 p) 5.0 South - Lightning South - Arson South - Camping South - Smoking 2 2 2 r 0.24 2 4.2 r 0.01 r 0.00 4.2 4.2 r 0.00 4.2 P 0.494 P 0.778 P 0.002 P 0.921 2 3.4 3.4 3.4 3.4 r 0.32 P 0.001 2.6 2.6 2.6 2.6
Area burned (log ha/yr/million ha) 2 2 2 1.8 1.8 r 0.00 1.8 r 0.00 1.8 r 0.08 P 0.656 P 0.925 P 0.016 1.0 1.0 1.0 1.0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
Fig. 9. Ignition sources recorded throughout the period 1910–2016 on USFS protected lands by frequency (a–h) and area burned (i–p) in the north (climate divisions North Coast, North Interior, and Sierra Nevada) and the south (Central Coast and South Coast) with lines for significant regressions from 1910 to 1979 and from 1980 to 2016; note the change in scale between lightning and anthropogenic sources for fire frequency.
Climate relationships to ignitions ignition sources exhibited a significant climate mode and the Based on the sharp change in ignition patterns through the models determining fire frequency were not the same as those period of record, it is critical to understand to what extent for area burned. We note that, before 1980, the biggest driver of climate variation may have played a role. Considering the debris and railroad fires was prior-year precipitation. Since marked changes in climate over the period of this study 1980, there was a negative relationship with summer tempera- (illustrated as decadal anomalies in seasonal temperature and ture for debris burning, playing with fire, smoking and railroad precipitation in Fig S3), it is reasonable to expect climate fires. In the south, where lightning-ignited fires were uncommon variation has some explanatory value in understanding changes (Fig. 2a), before 1980, they were strongly associated with high in ignition sources. summer temperatures and autumn precipitation. Multi-variate models used mean temperature and total pre- Total area burned on Cal Fire lands in the north before 1980 cipitation for winter, spring, summer and autumn plus the prior- was significantly tied to low winter precipitation and high spring year winter–spring precipitation. Presented in Table 4 are those temperatures. In the south, area burned by both arson and ignition sources with a significant P , 0.05 model. Not all powerlines was significantly tied to climate variation. 792 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
(a) 200 North - Playing(b)(200 North - Equipmentc)(100 North - Vehiclesd)50 North - Powerlines r2 0.72 160 160 80 40 2 P 0.001 r2 0.77 r 0.88 r2 0.68 120 P 0.001 120 P 0.001 60 P < 0.001 30 r2 0.35 r2 0.23 2 r 0.88 P 0.001 P 0.002 80 80 40 20 P 0.001 40 40 20 10
0 0 0 0 (e) 200 (f)(200 g)(100 h) 50 South - Playing South - Equipment South - Vehicles South - Powerlines 160 160 80 40 r2 0.76 r2 0.57 2 2 r 0.64 r2 0.75 r2 0.67 r2 0.36 r 0.01 120 P 0.001 P 0.001 120 60 30 P 0.001 P < 0.001
Fire frequency (#/yr/million ha) P 0.001 P 0.001 P 0.682 80 80 40 20
40 40 20 10
0 0 0 0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
(i) 5.0 North - Playing(j)(5.0 North - Equipmentk)(5.0 North - Vehiclesl) 5.0 North - Powerlines
4.2 4.2 4.2 4.2 2 r2 0.43 r2 0.92 r 0.05 r2 0.17 3.4 3.4 3.4 3.4 2 2 P 0.001 P 0.001 P 0.181 P 0.034 r 0.18 r 0.10 P 0.193 P 0.052 2.6 2.6 2.6 2.6 r2 0.22 1.8 P 0.148 1.8 1.8 1.8
1.0 1.0 1.0 1.0 (m)(5.0 n)(5.0 o)(5.0 p)5.0 South - Playing South - Equipment South - Vehicles South - Powerlines 4.2 4.2 4.2 4.2 r2 0.44 2 2 2 3.4 P 0.001 3.4 r 0.91 r 0.15 3.4 r 0.04 3.4 r2 0.10 r2 0.01 2 r 0.81 P 0.001 P 0.019 P 0.314 P 0.354 P 0.665 2.6 P 0.001 2.6 2.6 2.6
Area burned (log ha/yr/million ha) 1.8 1.8 1.8 1.8
1.0 1.0 1.0 1.0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
Fig. 10. Ignition sources variously recorded throughout the period 1910–2016 on USFS lands by frequency (a–h) and area burned (i–p) in the north (climate divisions North Coast, North Interior and Sierra Nevada) and the south (Central Coast and South Coast) with lines for significant regressions before 1980 and 1980–2016; note the change in scale for fire frequency.
On USFS lands, total number of fires and area burned in both a very strong effect of prior-year precipitation and summer the north and south exhibited many significant climate models temperature. (Table 5). However, the patterns are complicated and not easily Since 1980, one of the strongest climate variables affecting summarised as the specific climate models varied both spatially both frequency and area burned was a positive relationship with and temporally, as well as being different for different ignition prior-year precipitation. Although higher summer temperatures sources. were associated with increased frequency of arson fires, it was For example, lightning fire frequency and area burned in noteworthy that lower summer temperatures were associated both the north and south and before and after 1980 were with an increased incidence of smoking, camping and children significantly associated with climate variation, but, before playing with fire in the north. Frequency of powerline fires were 1980 in the north, the frequency of lightning fires was posi- associated with elevated autumn temperatures and higher prior- tively associated with summer precipitation, but the area year precipitation. burned was negatively associated with summer precipitation. Cal Fire area-burned data were also presented by vegetation After 1980 in the north, the model switched and there was type and showed that in the north, forest and shrubland area Fire ignition sources Int. J. Wildland Fire 793
(a) 50 North - Debris(b)(50 North - Playingc)(50 North - Equipmentd) 25 North - Powerlines
40 40 40 20
30 30 r2 0.76 30 15 r2 0.17 r2 0.16 r2 0.35 P 0.011 r2 0.36 P 0.001 r2 0.45 r2 0.00 20 20 20 P 0.014 10 P 0.001 P 0.067 P 0.001 r2 0.03 P 0.735 10 10 10 5 P 0.650
0 0 0 0
(e) 50 (f)(50 g)(50 h) 25 South - Debris South - Playing South - Equipment South - Powerlines 40 40 40 20 r2 0.02 r2 0.07 2 2 2 2 2 30 P - 0.718 P 0.120 30 r 0.13 r 0.78 30 r 0.59 r 0.14 15 r 0.03 P - 0.303 P 0.001 P - 0.022 P 0.337
Fire frequency (#/yr/million ha) P 0.001 20 20 20 10 r2 0.02 P 0.729 10 10 10 5
0 0 0 0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
(i) 5.0 North - Debris(j)(5.0 North - Playingk)(5.0 North - Equipmentl) 5.0 North - Powerlines
4.2 4.2 4.2 4.2 r2 0.10 r2 0.03 r2 0.21 r2 0.38 r2 0.54 r2 0.01 3.4 3.4 3.4 3.4 2 P 0.011 P 0.305 P - 0.185 P 0.001 P 0.001 P 0.615 r 0.00 r2 0.01 P 0.938 P 0.491 2.6 2.6 2.6 2.6
1.8 1.8 1.8 1.8
1.0 1.0 1.0 1.0
(m)(5.0 n)(5.0 o)(5.0 p)5.0 South - Debris South - Playing South - Equipment South - Powerlines 2 4.2 r 0.17 4.2 4.2 4.2 P 0.001 r2 0.02 2 2 2 2 2 2 3.4 P 0.360 3.4 r 0.16 r 0.28 3.4 r 0.15 r 0.04 3.4 r 0.02 r 0.03 P 0.246 P 0.001 P 0.097 P 0.216 P 0.733 P 0.303 2.6 2.6 2.6 2.6
Area burned (log ha/yr/million ha) 1.8 1.8 1.8 1.8
1.0 1.0 1.0 1.0 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 1900 1930 1960 1990 2020 Year
Fig. 11. Ignition sources variously recorded throughout the period 1910–2016 on USFS lands by frequency (a–h) and area burned (i–p) in the north (climate divisions North Coast, North Interior, and Sierra Nevada) and the south (Central Coast and South Coast) with lines for significant regressions prior to 1980 and 1980-2016; note the change in scale for fire frequency. burned had significant relationships with climate variation departures from this linear model, with accelerated ignitions before 1980 but not afterwards (Table 6). In the south, grass- during the 1920s and 1930s and a marked drop in the 1950s lands had a significant climate model after 1980. and 1960s (Fig. 7). Climate may have had some role in these changes since the former decade was drier and the latter Discussion was wetter (Fig S3) and during this period total fires on USFS Particularly striking about California ignitions is the steady lands did have a significant climate model largely driven by increase in number of fires since the early 1900s until a peak c. high summer temperatures and low summer precipitation 1980, followed by a marked drop in fire frequency up to 2016. (Table 4). What is particularly striking is the disconnect This happened on both lower-elevation Cal Fire-protected between number of ignitions and area burned; during the first lands and higher-elevation USFS lands, and in most climate three-quarters of the 20th century, although ignitions were divisions (Fig. 6). Despite a significant increase in fires during increasing, area burned was steadily decreasing through much the first three-quarters of the 20th century, there were marked of the state. 794 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
In contrast, since 1980, ignitions have steadily declined, yet Table 4. Significant climate models (P , 0.05) explaining frequency of area burned has either not changed or, in some northern parts of ignitions and area burned for the period ,1980 and $1980 for Cal Fire the state, has increased. In short, the number of ignitions does protected lands not directly explain area burned. However, as discussed below, Models tested mean temperature and total precipitation in winter, spring, summer and autumn, and prior year winter þ spring precipitation this conclusion does not apply to individual ignition sources, , and, in this respect, there may be particular sources worth ( 1PptWinSpr) targeting for fire management purposes. 2 Factors that may have played a role in these historical Variable Era Adjusted R P Model patterns of ignitions and area burned are changes in: population Frequency in North density, infrastructure development, fire-prevention success, Debris ,1980 0.18 0.021 ,1PptWinSpr fire-suppression effectiveness, vegetation-management prac- Debris $1980 0.24 0.050 - TempSum tices, climate, and possibly record-keeping accuracy. The dri- Playing $1980 0.27 0.035 - TempSum þ TempWin vers behind changes in ignition patterns are quite possibly Smoking $1980 0.32 0.016 - TempSum Railroad ,1980 0.18 0.020 ,1PptWinSpr þ PptSpr different for different sources, different parts of the state and $ þ at different times. First, we consider the patterns for natural Railroad 1980 0.25 0.045 - TempSum TempWin Frequency in South lightning-ignited v. human-caused wildfires. Lightning ,1980 0.40 0.001 TempSum þ PptAut Debris ,1980 0.18 0.021 ,1PptWinSpr – PptAut Area in North: Lightning-ignited fires Total ,1980 0.23 0.007 - PptWin þ TempSpr In California, natural lightning-ignited fires decreased from Area in South: north to south and from high (USFS) to low (Cal Fire) elevation Arson $1980 0.25 0.045 - TempWin (Fig. 2, 4). On USFS lands, Lassen and Plumas forests in the Powerlines ,1980 0.99 0.046 - TempSum – PptSpr – north-east averaged over 150 lightning-ignited fires per year per PptAut million hectares, whereas the coastal Los Padres Forest aver- Powerlines $1980 0.24 0.050 - TempAut aged one-tenth as many (Table S5). In northern California for- ests, such as the Klamath, Lassen, Modoc, Inyo-Mono and Sequoia, lightning accounted for the majority of fires, and on many others it is about equally important as human-ignited fires (Stephens 2005; Van de Water and Safford 2011; Safford and Van (Table 3). Notable exceptions are the coastal Los Padres and de Water 2014). In coastal central and southern California, southern California Angeles, Cleveland and San Bernardino lightning accounts for very little area burned, in large part because forests, where lightning accounted for less than one-quarter of lightning strikes are very low, but also because human-ignited all fires. In contrast, on lower-elevation Cal Fire-protected fires often occur under weather conditions more conducive to fire lands, lightning accounted for less than 10% of all fires in most spread, contributing to a shorter fire-rotation interval, e.g. 40–50 counties, and in coastal areas from Sonoma County south, typ- years on southern California forest lands (Table 1). ically ,1% of all ignitions (Table 2). These patterns closely Lightning fires have increased markedly over most of the follow the distribution of lightning strikes in the state (van 20th century on both Cal Fire and USFS lands, in the north and Wagtendonk and Cayan 2008). In general, lightning-ignited south (Fig. 8, 9). A possible explanation for this pattern is fires in coastal California were substantially less than that improvement in detection, as lightning-ignited fires often occur observed over much of the USA (Prestemon et al. 2013). Thus, in remote areas and detection may have been less effective in the the report that extreme fire events driven by high winds are early part of the 20th century and improved in the latter part of commonly due to human ignitions and not lightning (Abatzo- the 20th century. However, there is there is reason to retain some glou et al. 2018) should not be too surprising in California level of scepticism that this pattern is an artefact of reporting because these extreme winds are largely restricted to coastal (see the ‘Methods’ section), primarily because state and federal areas in southern California and the San Francisco Bay Area. agencies have put in extraordinary effort at fire detection since Area burned by lightning-ignited fires approximately paral- the early 1900s (Clar 1969; Cermak 2005), including hundreds leled these geographical patterns with a couple noteworthy of thousands of kilometres of wilderness aircraft fire patrols exceptions. In the northern and central part of the state, more beginning in 1919 (Cermak 1991). coastal USFS forests had low lightning-ignited fire frequency Another reason for not simply dismissing historical patterns but these accounted for a substantial amount of area burned, as an artefact of reporting is that there are physical factors that although this was less evident on lower-elevation Cal Fire lands could account for such changes. For example, one potential (Fig. 2–5). For interior forests where lightning is the dominant factor for a 20th-century rise in lightning fires could be changes ignition source, fires have proven to be reasonably easy to in forest fuel structure, which has been shown to affect light- extinguish, in large part because they typically occur in forests ning-ignited fire frequency on other landscapes (Krawchuk et al. with a low-intensity surface-fire regime, and during lightning- 2006). In California, this would be expected based on the storm weather conditions (van Wagtendonk and Cayan 2008), marked drop in area burned following the burning peak in the are conducive to rapid fire control. As a consequence, less than 1920s – for both Cal Fire and USFS lands, three times more area 1% of these forest lands burn each year and these landscapes burned in that decade relative to the decadal average burned have a fire-rotation interval of 100–200 years (Table 1), very from 1950 to 1980 (Keeley and Syphard 2017). Thus, during the different with what is believed to be the natural fire interval mid-20th century there was potentially an increase in fuels that Fire ignition sources Int. J. Wildland Fire 795
Table 5. Significant climate models (P , 0.05) explaining frequency of ignitions and area burned for the period ,1980 and .1980 for USFS protected lands Models tested mean temperature and total precipitation in winter, spring, summer and autumn, and prior year winter þ spring precipitation (,1PptWinSpr)
Variable Era Adjusted R2 P Model
Frequency in North Total ,1980 0.27 0.001 TempSum – PptSum þ PptAut Lightning ,1980 0.21 0.005 TempSum þ PptSum þ PptAut Arson ,1980 0.13 0.040 –PptAut – PptSpr Arson $1980 0.28 0.030 TempSum Smoking $1980 0.46 0.001 –TempSum – PptSum Debris $1980 0.43 0.002 – PptWin – TempAut – PptSum Camping $1980 0.26 0.039 –TempSum Playing $1980 0.47 0.001 –TempSum – PptSum Railroad , 1980 0.28 0.007 PptAut Railroad $1980 0.32 0.017 –PptWin þ PptSpr Equipment $1980 0.32 0.016 –TempSum – PptWin – PptSum Powerlines $1980 0.35 0.010 ,1PptWinSpr Frequency in South Total $1980 0.29 0.026 Ppt–1WinSpr Lightning ,1980 0.18 0.012 TempSum þ PptAut Powerlines $1980 0.28 0.029 TempAut þ PptWin Area in North: Total ,1980 0.35 ,0.001 –PptWinSpr – PptSum Total $1980 0.63 ,0.001 ,1PptWinSpr þ TempSum – PptAut Lighting ,1980 0.17 0.018 –PptSum Lighting $1980 0.64 ,0.001 ,1PptWinSpr þ TempSum Arson ,1980 0.38 ,0.001 –PptWin – PptAut – TempAut – TempWin Debris ,1980 0.15 0.031 – PptWin –PptAut Smoking ,1980 0.16 0.022 TempAut þ TempSpr Playing $1980 0.34 0.010 –TempSum –TempAut Equipment $1980 0.28 0.030 ,1PptWinSpr Vehicles $1980 0.37 0.007 ,1PptWinSpr Powerlines $1980 0.47 0.001 –TempWin þ ,1PptWinSpr Area in South Total ,1980 0.20 0.006 – PptAut – PptWin Total $1980 0.24 0.048 TempSum – TempWin Lighting $1980 0.25 0.043 –TempSpr – PptSpr
may have contributed to a greater chance of lightning strikes The future projections are that lightning strikes will igniting fires. increase 50% over this century (Romps et al. 2014), but this Changes in reporting standards is also not likely to explain is not easily translated into future lightning fire risks in the pattern of decreased lightning-ignited fires from 1980 to California. Some landscapes, such as forests in the north- 2016 (Fig. 8, 9). On Cal Fire and southern USFS lands, this did eastern part of the state, may already be saturated with not produce any significant trend in area burned by this fire lightning ignitions and coastal landscapes have very few source, although northern California (including the Sierra strikes and thus a 50% increase may not significantly change Nevada) USFS lands showed an increase in area burned by lightning-ignited fire risk. In addition, changes in lightning- this source, a pattern also seen in the northern Rocky Moun- strike frequency will have very different impacts dependent on tains (Stephens 2005). Climate is strongly implicated in this which season those changes occur in as well the state of future change (Table 5) as two-thirds of the annual variation in area fuel conditions. burned by lightning-ignited fires is explained by a combination of prior-year precipitation and current-year summer tempera- ture. Although the latter variable most likely affects fuel Human-ignition sources moisture at the time of fire, the former is thought to increase The fact that in all climate divisions, the number of ignitions is fires through its effect on herbaceous fuels in the following not a monotonic function of time over the past 100 years sug- year (Littell et al. 2009; Crimmin and Comrie 2011; Keeley gests a complex model of how ignition sources affect burning and Syphard 2016). A similar conclusion was drawn by Knapp activity. Prestemon et al. (2013) presented a conceptual model (1995) for the climatic control of lightning-ignited fires in the of biophysical, social, prevention and management drivers in Intermountain West. controlling human ignition sources. These factors are not static, 796 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
as illustrated by Guyette et al.’s (2002) dynamic anthropogenic fragmentation of fuel surely has affected ignition trends and area fire regime model for the Ozark Mountains in Missouri. In their burned. It may therefore be important to monitor areas that are model, they found that the landscape changed over time from becoming newly developed, as these may be the most fire-prone being ignition limited to fuel limited followed by stages areas on the landscape, with sufficient people to start fires and dependent on fuel fragmentation and ultimately a culture- wildland vegetation to carry fires (Radeloff et al. 2018). dependent stage. These temporal changes in drivers could Patterns such as these have been interpreted as indicating explain a lot about the temporal changes observed in California fires are not limited by human ignitions (Knorr et al. 2013; ignitions. Moritz and Knowles 2016). This has prompted some to conclude It may be that the marked rise in ignitions during the first that fire activity during the last several decades has been driven three-quarters of the 20th century in California is the result of largely by climate change (Westerling et al. 2011). It is apparent increasing effectiveness of reporting, but this seems unlikely that in mid–high-elevation forests in California seasonal climate because the steepest rise in ignitions was in the latter part of the variation has been an important factor in determining annual 20th century, i.e. 1960–1980 (Fig. 6). The 20th-century increase area burned (Table 5) and that global warming may exacerbate in ignitions was very strongly correlated with population growth the fire situation on those landscapes (Keeley and Syphard (Tables 2, 3), but we believe that more is involved than just 2016). However, in coastal California, climates are capable of increasing population growth translates into more fires. This generating large fire events most years (Keeley and Syphard early–mid-20th-century growth spurt was correlated with road 2017), with one exception being years with anomalous late expansion throughout the state, which was bringing more people spring rains (e.g. Dennison et al. 2008). In these coastal loca- in contact with highly flammable fuels (Show 1945; Lockmann tions, big fire events occur during extreme wind events, how- 1981; Keeley and Fotheringham 2003). In addition, because of ever, these Santa Ana, Diablo or North Wind events occur migration patterns, growth included populations from less fire- predictably every year and yet big fires occur at unpredictable prone parts of the US, and thus a population relatively naı¨ve intervals, being determined by the coincidence of a human about the dangers of fire use in wildland areas (e.g. Zahn 1944; ignition with a wind event (Keeley and Zedler 2009). Show 1945). In addition, fire-prevention education was in its During the first two-thirds of the 20th century more people infancy and the population was slow to recognise their role in the translated into more fires, and greater fire activity. However, in fire problem. Included too is the widespread use of outdoor recent decades the relationship between human population equipment that contributed to the sharpest rise in fires on Cal growth and fire activity has become more complex, nicely Fire landscapes between 1960 and 1980 (Fig. 10). On top of that, captured in Prestemon et al.’s (2013) model. In California in development of fire-response actions were far from perfect (Clar recent decades, increasing population density has increased the 1969). Also, during the period from 1940 to 1970 the State probability of ignitions under the worst weather conditions, either Resources Agency was actively involved in promoting burning intentionally by arson for example or accidentally by powerline of chaparral shrublands for the express purpose of type- failures. This appears to be a widely seen situation throughout the converting native shrublands to exotic grasslands of greater USA where human-related ignitions are associated with condi- economic value as rangelands (unpublished records in the State tions resulting in large wildfires (Nagy et al. 2018). Archives). Indeed, the state was funding type conversion of Decreasing ignitions over the last 4 decades is potentially private lands as this was perceived as a fire hazard reduction reflective of increasing efficiency of fire prevention. However, strategy, with an economic incentive of increasing rangeland. it also likely reflects changes in human infrastructure; new roads Not directly related to changing demography is the signifi- in this era were tied to development projects that required cant decline in fires in the last several decades – while popula- demonstration of adequate fire response capabilities. In addi- tions continued to grow after 1980, fire frequency was tion, an important factor behind declining ignitions is quite negatively related to population density (Tables 2, 3). This is possibly the emergence of the California Fire Safe Council in consistent with the pattern of fire activity peaking under inter- the early 1990s (http://www.cafiresafecouncil.org/about-us/, mediate population density (Syphard et al. 2009). That is, the accessed 11 August 2016), which made significant contributions relationship between population density and ignition frequency to fire-safety education. is likely a function of finer-scale spatial processes regulating the Arson has long been a major source of intentional human degree of interspersion between development patterns and ignitions on both Cal Fire (Fig. 8) and USFS (Fig. 9) lands and on wildland vegetation. In other words, as both population and both jurisdictions arson ignitions increased during the first part development expand into wildland areas, ignitions increase up of the 20th century and then dropped markedly in recent to a point at which the area of development, or, impervious decades. Arson fires have always been one of the largest sources surface, far exceeds the area of wildland, and at that point, the of area burned, although it was much higher in the early 20th relationship becomes negative. However, the timing of this century than in recent decades. This category comprises igni- switch varies with regions, e.g. south-east Australia continues tions motivated for diverse reasons. Early in the 20th century, to see a positive relationship in between population density and these were termed incendiary fires and were often motivated by fire frequency (Collins et al. 2015). goals of maintaining traditional burning practices (Coughlan Thus, these broad-scale patterns observed across the state may 2016). As such practices became less socially (and legally) be reflecting macro-scale urbanisation trends over time. Massive acceptable, the category was labelled arson fires. Arson fires areas of wildland vegetation have been developed and fragmen- exhibit interesting distribution patterns. On low-elevation Cal ted in California over the course of the 20th century (Hammer Fire lands they are a major ignition source in the northern Sierra et al. 2007; Syphard et al. 2017), and the resulting extent and Nevada (Fig. 2, 3) but on USFS lands they dominate in the Fire ignition sources Int. J. Wildland Fire 797
southern part of the state (Fig. 4, 5), suggesting a need for more that powerline fires are so dangerous is that they commonly concentrated anti-arson prevention measures in those regions. occur during high winds and there are three effects of these This clustering of arson fires has been observed in parts of the winds: tree contact, line arcing, and metal fatigue resulting in Mediterranean basin and has prompted an early alert system lines down (Mitchell 2009). These winds create extremely (Gonzalez-Olabarria et al. 2012). dangerous fires capable of rapid spread over long distances. One of the real success stories illustrated by these data is the This is a serious problem in other regions such as southern marked decline since 1980 in frequency and area burned by Australia where it was found that electricity-caused wildfires arson fires on both Cal Fire and USFS lands (Fig. 8, 9). This are over-represented when fire danger is high (Miller et al. reduction in arson fires is a pattern observed for other parts of the 2017) and similar conclusions were drawn by Ganteaume country (Prestemon et al. 2013). In California, this may be and Guerra (2018). Powerline distribution tends to follow attributed to better neighbourhood-watch programs, which roads and this may be part of the reason burning patterns include patrols during red-flag warnings, but broadcasted fire are closely correlated with road distribution in southern prevention messaging may also be a factor. Another factor may California (Faivre et al. 2014). Also, they burn larger areas be increased penalties for arson; e.g. the person found guilty of than fires ignited by most other causes and are associated with starting the 2003 Old Fire in southern California was sentenced more significant impacts on lives and property (Collins et al. to death, as was the arson convicted of the 2006 Esperanza Fire 2016). (Gabbert 2012). Because these powerline failures typically occur in known Another source of burning on both Cal Fire and USFS lands extreme-wind corridors, it has been proposed that wiring these has been smoking. This was a significant cause in the earliest corridors with underground power could minimise the problem records, recording even ignitions from cigarettes thrown from (Keeley et al. 2009). However, utility companies have shown a open cockpit planes. Throughout the first half of the 20th reluctance to accept this solution. One company in southern century, smoking was a major cause of wildfires and was the California, San Diego Gas & Electric, has opted for an alterna- focus of one of the earliest fire prevention campaigns. In 1942, tive plan whereby they monitor weather throughout the county over 100 000 ‘fag bags’ were distributed to persons entering the and use these data to shut down portions of the power grid when Angeles National Forest, bright red bags designed to carry that area experiences high winds (https://www.cnbc.com/2017/ smoking materials and with a prominent fire-safety reminder 12/13/southern-california-utilities-shut-off-power-to-prevent- stamped on them (Show 1945). The late 20th-century decline in wildfires.html, accessed 1 June 2017). In initial attempts to deal smoking caused such fires to decline at a much faster rate (Fig. 8, with fire hazards there have been significant complaints about 9) than due to simple reduction in smoking (Prestemon et al. the process of shutting down the power grid as it creates many 2013). Reductions in smoking-caused fires are due to a combi- unanticipated problems (https://www.nbcsandiego.com/news/ nation of less smoking, more fire-resistant cigarettes, and local/Supervisor-Demands-State-Investigation-of-Power-Shut- improved fire prevention (Butry et al. 2014). offs-During-Lilac-Fire-467782743.html, accessed 1 June 2017). Children playing with fire has been an important ignition Other approaches have been to replace wooden poles with metal source and it has exhibited a marked decline in frequency in poles, however, this seems to be a distraction since wooden poles recent decades on both Cal Fire (Fig. 10) and USFS (Fig. 11) have not been blamed for starting fires (https://www.voiceof- lands. Increased fire-prevention effectiveness through better sandiego.org/topics/science-environment/sdge-environmentalists- messaging and development of childproof lighters are potential are-at-opposite-poles-on-one-fire-prevention-method/, accessed 1 factors. Perhaps stricter ordinances in power-tool usage in June 2017). wildlands under red-flag warnings may be a factor as well as Climate change impacts on anthropogenic ignitions is rather requirements for more effective spark arrestors. difficult to parse out because climate affects both fire behaviour Vehicles present another accidental fire source that has and human behaviour. For example, in forests, fire activity is declined sharply on Cal Fire protected lands. Catalytic conver- enhanced by higher spring and summer temperatures through ters, which were first required in 1975, are thought to have been effects on fuel moisture (Westerling et al. 2006; Littell et al. a significant ignition factor (Bertagna 1999; http://www.cbs8. 2009; Keeley and Syphard 2017). However, in the present study, com/story/35871110/how-a-cars-catalytic-converter-can-spark- fires started by camping, children playing with fire, and smoking a-massive-fire, accessed 1 June 2017) when they overheated, were negatively correlated with summer temperatures, suggest- igniting roadside vegetation. However, modern vehicles have ing the possibility that cooler temperatures may have encour- warning lights when they overheat, which has the potential aged greater outdoor activity. for reducing vehicle fires and could be a factor in the decline In general, climate variation exhibited a closer relationship of such fires. Another factor potentially reducing vehicle fires is with fire activity in the higher-elevation USFS lands in the improved vegetation treatment along roadside verges. northern part of the state, consistent with the flammability limited Electrical powerlines have been reported ignition sources fire regimes in these regions (Keeley and Syphard 2017). Of since 1905 (Show 1945). In the present study, this source of particular significance is the importance of prior-year rainfall ignition stands out in that, unlike many other human ignition as this is well known to be due to increased fuel production sources, powerline fires and area burned by this ignition source in grass dominated ecosystems (Crimmins and Comrie 2004; have not declined in recent decades (Fig. 10, 11). Although Keeley and Syphard 2016). We found that this climate variable powerlines do not account for many fires, they often account was strongly tied to powerline fires, suggesting perhaps that for substantial area burned, and some of substantial size fine flashy fuels may be a marked hazard in association with (Keeley et al. 2009; Syphard and Keeley 2015). One reason powerlines and may be an additional management target. 798 Int. J. Wildland Fire J. E. Keeley and A. D. Syphard
Conclusions Coughlan MR (2016) Wildland arson as clandestine resource management: Throughout California, fire frequency has increased steadily a space-time permutation analysis and classification of informal fire management regimes in Georgia, USA. Environmental Management 57, until a peak c. 1980, followed by a marked drop to the present. 1077–1087. doi:10.1007/S00267-016-0669-3 There was not a tight link between frequency of ignition sources Crimmins MA, Comrie AC (2004) Interactions between antecedent climate and area burned by those sources and the relationships changed and wildfire variability across south-eastern Arizona. International over time. Natural lightning-ignited fires decreased from north Journal of Wildland Fire 13, 455–466. doi:10.1071/WF03064 to south and from high to low elevation. Throughout most of the Dennison PE, Moritz MA, Taylor RS (2008) Evaluating predictive models state human-caused fires dominated the record and were posi- of critical live fuel moisture in the Santa Monica Mountains, California. tively correlated with population density for the first two-thirds International Journal of Wildland Fire 17, 18–27. doi:10.1071/ of the record, but this relationship reversed in recent decades. WF07017 Most ignition sources have declined markedly in recent decades Donoghue LR (1982a) Classifying wildfire causes in the USDA forest service: with one notable exception, powerline ignitions. One important problems and alternatives. USDA Forest Service, North Central Forest Experiment Station, Research Note NC-280. (St. Paul, MN, USA) avenue for future fire hazard reduction will be consideration of Donoghue LR (1982b) The history and reliability of the USDA Forest solutions to reduce this source of dangerous fires. Service wildfire reports. USDA Forest Service, North Central Forest Experiment Station, Research Paper NC-226. (St. Paul, MN, USA) Conflicts of interest Faivre N, Jin Y, Goulden ML, Randerson JT (2014) Controls on the spatial The authors declare that they have no conflicts of interest. pattern of wildfire ignitions in southern California. International Journal of Wildland Fire 23, 799–811. doi:10.1071/WF13136 Gabbert B (2012) Man gets death sentence for starting the Old Fire in 2003. Acknowledgements In Wildfire Today, 28 September 2012. Available at http://wildfiretoday. We thank colleagues for helpful feedback; Hugh Safford (USFS), Mike com/2012/09/28/man-gets-death-sentence-for-starting-the-old-fire-in- Rohde (Orange Co Fire Authority, retired), Dave Passovoy, Dave Sapsis and 2003/ [Verified 25 October 2018] Tadashi Moody (Cal Fire), Gary Gilbert (Cal Fire, retired) and Stephen Pyne Ganteaume A, Guerra F (2018) Explaining the spatio-seasonal variation of (Arizona State University). Thanks also go to Anne Pfaff, Kate Dobrinsky fires by their causes: the case of southeastern France. Applied Geography and Ken Ferschweiler for assistance with data gathering. 90, 69–81. doi:10.1016/J.APGEOG.2017.11.012 Gonzalez-Olabarria JR, Brotons L, Gritten D, Tudela A, Teres JA (2012) References Identifying location and causality of fire ignition hotspots in a Mediter- Abatzoglou JT, Balch JK, Bradley BA, Kolden CA (2018) Human-related ranean region. International Journal of Wildland Fire 21, 905–914. ignitions concurrent with high winds promote large wildfires across the doi:10.1071/WF11039 Guyette RP, Muzika RM, Dey DC (2002) Dynamics of an anthropogenic USA. International Journal of Wildland Fire 27, 377–386. doi:10.1071/ WF17149 fire regime. Ecosystems 5, 472–486. Balch JK, Bradley BA, Abatzoglu JT, Nagy RC, Fusco EJ, Mahood AL Hammer RB, Radeloff VC, Fried JS, Stewart SI (2007) Wildland–urban (2017) Human-started wildfires expand the fire niche across the United interface housing growth during the 1990s in California, Oregon, and Washington. International Journal of Wildland Fire 16, 255–265. States. Proceedings of the National Academy of Sciences of the United States of America 114, 2946–2951. doi:10.1073/PNAS.1617394114 doi:10.1071/WF05077 Bedard P (2017) Interior’s Zinke demands ‘aggressive’ war on fires, stop Kalies EL, Kent LLY (2016) Tamm Review: are fuel treatments effective at letting ‘nature take its cours’. In Washington Examiner, 14 September achieving ecological and social objeectives? A systematic review. Forest Ecology and Management 375, 84–95. doi:10.1016/J. 2017. Available athttp://www.washingtonexaminer.com/interiors-zinke- demands-aggressive-war-on-fires-stop-letting-nature-take-its-course/arti- FORECO.2016.05.021 cle/2634376 [Verified 25 October 2018] Keeley JE, Fotheringham CJ (2003) Impact of past, present, and future fire regimes on North American Mediterranean shrublands. In ‘Fire and Bertagna PJ (1999) Catalytic converter caused fires. In ‘Industrial Opera- Climatic Change in Temperate Ecosystems of the Western Americas’. tions Fire Prevention Field Guide’. pp. 135–141. (California Department of Forestry and Fire Protection: Sacramento, CA, USA) (Eds TT Veblen, WL Baker, G Montenegro, TW Swetnam) pp. 218–262. Brown WS (1945) History of Los Padres National Forest. Manuscript on file, (Springer: New York, NY, USA) Keeley JE, Safford HD (2016) Fire as an ecosystem process. In ‘Ecosystems USFS, Los Padres National Forest. Butry DT, Prestemon JP, Thomas DS (2014) Investigation of the decline in of California’. (Eds H Mooney, E Zavaleta) pp. 27–45. (University of reported smoking-caused wildfires in USA from 2000 to 2011. Interna- California Press) tional Journal of Wildland Fire 23, 790–798. doi:10.1071/WF13146 Keeley JE, Syphard AD (2016) Climate change and future fire regimes: examples from California. Geosciences 6, 37. doi:10.3390/ Cermak RW (1991) Pioneering aerial forest fire control: the Army air patrol in California, 1919–1921. California History 70, 290–305. doi:10.2307/ GEOSCIENCES6030037 25158571 Keeley JE, Syphard AD (2017) Different historical fire-climate patterns in California. International Journal of Wildland Fire 26, 253–268. doi:10. Cermak RW (2005) Fire in the forest. A history of forest fire control on the 1071/WF16102 national forests in California, 1898–1956. USDA Forest Service, Pacific Southwest Region, R5-FR-003. (Albany, CA, USA) Keeley JE, Zedler PH (2009) Large, high intensity fire events in southern Clar CR (1969) ‘Evolution of California’s Wildland Fire Protection California shrublands: debunking the fine-grained age-patch model. Ecological Applications 19, 69–94. doi:10.1890/08-0281.1 System.’ (California State Board of Forestry: Sacramento, CA, USA) Collins KM, Price OF, Penman TD (2015) Spatial patterns of wildfire Keeley JE, Safford H, Fotheringham CJ, Franklin J, Moritz M (2009) The ignitions in south-eastern Australia. International Journal of Wildland 2007 southern California wildfires: lessons in complexity. Journal of Fire 24, 1098–1108. doi:10.1071/WF15054 Forestry 107, 287–296. Knapp PA (1995) Intermountain West lightning-caused fires: climatic Collins KM, Penman TD, Price OF (2016) Some wildfire ignition causes pose more risk of destroying houses than others. PLoS One . doi:10.1371/ predictors of area burned. Journal of Range Management 48, 85–91. JOURNAL.PONE.0162083 doi:10.2307/4002510 Fire ignition sources Int. J. Wildland Fire 799
Knorr W, Kaminski T, Arneth A, Weber U (2013) Impact of human Safford HD, Van de Water KM (2014) Using fire return interval departure population density on fire frequency at the global scale. Biogeosciences (FRID) analysis to map spatial and temporal changes in fire frequency on Discussions 10, 15735–15778. doi:10.5194/BGD-10-15735-2013 national forest lands in California. USDA Forest Service, Pacific Krawchuk MA, Cumming SG, Flannigan MD, Wein RW (2006) Biotic and Southwest Research Station, Research Paper PSW-RP-266. (Albany, abiotic regulation of lightning fire initiation in the mixedwood boreal CA, USA) forest. Ecology 87, 458–468. doi:10.1890/05-1021 Show SB (1945) History of the Angeles National Forest. Unpublished ms, on Kuhlken R (1999) Settin’ the woods on fire: rural incendiarism as protest. file in USDA Forest Service, Arcadia, CA. Geographical Review 89, 343–363. doi:10.2307/216155 Stephens SB (2005) Forest fire causes and extent on United States Forest Littell JS, McKenzie D, Peterson DL, Westerling AL (2009) Climate and Service land. International Journal of Wildland Fire 14, 213–222. wildfire area burned in western U.S. ecoprovinces, 1916–2003. Ecologi- doi:10.1071/WF04006 cal Applications 19, 1003–1021. doi:10.1890/07-1183.1 Syphard AD, Keeley JE (2015) Location, timing and extent of wildfire vary Lockmann RF (1981) ‘Guarding the Forests of Southern California.’ (Arthur by cause of ignition. International Journal of Wildland Fire 24, 37–47. H. Clark Co.: Glendale, CA, USA) doi:10.1071/WF14024 Miller C, Plucinski M, Sullivan A, Stephenson A, Huston C, Charman K, Syphard AD, Keeley JE (2016) Historical reconstructions of California Prakash M, Dunstall S (2017) Electrically caused wildfires in Victoria, wildfires vary by data source. International Journal of Wildland Fire 25, Australia, are over-represented when fire danger is elevated. Landscape 1221–1227. doi:10.1071/WF16050 and Urban Planning 167, 267–274. doi:10.1016/J.LANDURBPLAN. Syphard AD, Radeloff VC, Keeley JE, Hawbaker TJ, Clayton MK, Stewart 2017.06.016 SI, Hammer RB (2007) Human influence on California fire regimes. Mitchell JA (1947) Forest fire statistics: their purpose and use. Fire Control Ecological Applications 17, 1388–1402. doi:10.1890/06-1128.1 Notes 8, 14–17. Syphard AD, Radeloff VC, Keuler NS, Taylor RS, Hawbaker TJ, Stewart Mitchell JW (2009) Power lines and catastrophic wildland fire in southern SI, Clayton MK (2008) Predicting spatial patterns of fire on southern California. In ‘Fire and Materials Conference Proceedings, 11th Inter- California landscape. International Journal of Wildland Fire 17, 602– national Conference and Exhibition’, 26–28 January 2009, San Fran- 613. doi:10.1071/WF07087 cisco, CA, USA. Available at http://citeseerx.ist.psu.edu/viewdoc/ Syphard AD, Radeloff VC, Hawbaker TJ, Stewart SI (2009) Increases in download?doi=10.1.1.469.2877&rep=rep1&type=pdf [Verified 24 fire frequency in Mediterranean-climate ecosytems. Conservation Biol- October 2018] ogy 23, 758–769. doi:10.1111/J.1523-1739.2009.01223.X Moritz MA, Knowles SG (2016) Coexisting with wildfire. Promoting the Syphard AD, Keeley JE, Pfaff A, Ferschweiler K (2017) Human presence right kind of fire – and smarter development – is safer and more cost- diminishes the importance of climate in driving fire activity across the effective than fighting a losing battle. American Scientist 104, 220–225. United States. Proceedings of the National Academy of Sciences of the Nagy RC, Rusco E, Bradley B, Abatzoglou JT, Balch J (2018) Human- United States of America 114, 13750–13755. doi:10.1073/PNAS. related ignitions increase the number of large wildfires across U.S. 1713885114 ecoregions. Fire 1. doi:10.3390/FIRE1010004 Van de Water KM, Safford HD (2011) A summary of fire frequency Prestemon JP, Hawbaker TJ, Bowden M, Carpenter J, Brooks MT, Abt KL, estimates for California vegetation before Euro-American settlement. Sutphen R, Scranton S (2013) Wildfire ignitions: a review of the science Fire Ecology 7, 26–58. doi:10.4996/FIREECOLOGY.0703026 and recommendations for science and recommendations for empirical van Wagtendonk JW, Cayan DR (2008) Temporal and spatial distribution of modeling. USDA Forest Service, Southern Research Station, General lightning strikes in California in relation to large-scale weather patterns. Technical Report SRS-GTR-171. (Asheville, NC, USA) Fire Ecology 4, 34–56. doi:10.4996/FIREECOLOGY.0401034 Radeloff VC, Helmers DP, Kramer HA, Mockrin MH, Alexandre PM, Bar- Westerling AL, Hidalgo HG, Cayan DR, Swetnam TW (2006) Warming Massada A, Butsic V, Hawbaker TJ, Martinuzzi S, Syphard AD, and earlier spring increase western US forest wildfire activity. Science Stewart SI (2018) Rapid growth of the US wildland–urban interface 313, 940–943. doi:10.1126/SCIENCE.1128834 raises wildfire risk. Proceedings of the National Academy of Sciences of Westerling AL, Bryant BP, Preisler HK, Holmes TP, Hidalgo HG, Das T, the United States of America 115, 3314–3319. doi:10.1073/PNAS. Shrestha SR (2011) Climate change and growth scenarios for California 1718850115 wildfire. Climatic Change 109, 445–463. doi:10.1007/S10584-011- Romps DM, Seeley JT, Vollaro D, Molinari J (2014) Projected increase in 0329-9 lightning strikes in the United States due to global warming. Science 346, Zahn C (1944) The San Diego fires y an inquest. American Forests 50, 851–854. doi:10.1126/SCIENCE.1259100 161–164.
www.publish.csiro.au/journals/ijwf