Abrupt, climate-induced increase in wildfires in British Columbia since the mid-2000s

British Columbia (BC) has recently experienced unprecedented wildfire seasons (2017, 2018, 2021, 2023), contrasting with a 20th-century decline in fire activity attributed to cooler/wetter climate and suppression. Projections have long indicated increases in fire under warming, but the observed surge began earlier than anticipated (~2000s) and with great magnitude. The study asks whether climate change has pushed BC into a new fire epoch. It aims to compare century-scale trends in wildfire activity (area burned, number of fires, season timing) with annual climate variables (temperature, precipitation, climatic moisture deficit—CMD), evaluate regional differences across BC’s Central, Coastal, and Northern zones, and consider the roles of bottom-up factors (fuels, insects, land use, suppression). It further interprets findings in the context of future climate projections and discusses implications for communities and ecosystems.

Historical work shows a decline in BC wildfires from mid-20th century linked to wetter conditions and suppression (e.g., Meyn et al. 2010; 2013). Multiple studies project substantial increases in fire potential across BC with warming this century (e.g., Nitschke & Innes 2008; Wang et al. 2017). Globally and across western North America, recent increases in extreme fire weather and fire activity have been tied to rising temperatures and moisture deficits (e.g., Jolly et al. 2015; Abatzoglou & Williams 2016; Jain et al. 2022). Regional disturbance histories—including Indigenous fire stewardship, land-use change, logging, and insect outbreaks (notably mountain pine beetle)—have altered fuels and fire regimes across BC. Comparisons with western U.S. indicate similar climate trends but different timing and moisture regimes, with BC’s area-burned increase lagging U.S. regions such as California and the Interior West.

Study area included BC province and three nested zones (Central, Coastal, Northern). Climate data (historical and CMIP6 SSP projections) were downscaled/interpolated with ClimateNA v7.3 at 50-km resolution for 1900–2100. Selected variables (|Spearman r| < 0.7) included seasonal (spring/summer) mean temperature, total precipitation, Hargreaves climatic moisture deficit (CMD), and annual number of frost-free days (NFFD). Associations accounted for serial correlation using a modified Spearman correlation test (astrochron package). Wildfire perimeters: Canadian National Fire Database (1919–1984) and National Burned Area Composite (1985–2021); polygon areas were adjusted to correct for unburned islands and perimeter inaccuracies; fires <20 ha were excluded. Annual metrics computed: total area burned, number of fires ≥20 ha, and day-of-year when cumulative area burned reached thresholds (1%, 2%, 5%, 10%; 2% used in analyses). Time series (1919–2021) were modeled with segmented (piecewise) linear regression (segmented R package) to estimate breakpoints and slopes; significance assessed with a modified Mann–Kendall trend test adjusted for autocorrelation (modifiedmk package). Additional datasets summarized included suppression costs (inflation-adjusted), evacuation statistics, and disturbance/land cover layers (harvested cutblocks, mountain pine beetle impacts from CanLAD; land cover at 30 m). Comparative fire statistics for neighboring jurisdictions were compiled from jurisdiction-specific fire polygon datasets.

- After a century-long decline in fire activity, BC experienced a significant increase beginning mid-2000s; the segmented regression identified a breakpoint in annual area burned around 2008, coincident with shifts in climate variables. - CMD trends reversed from a wetting trend to drying: spring CMD breakpoint circa 2011 and summer CMD breakpoint circa 1999. Despite high precipitation, rapid warming increased evaporative demand, elevating moisture deficits. - Precipitation increases per degree of warming were insufficient to offset drying: approximately 3.34% (spring) and 5.74% (summer) per °C versus an estimated minimum ~15% needed to compensate for fuel flammability increases. - Fire season lengthening and earlier onset: number of frost-free days increased by about 26.7 days since early 20th century; the date when 2% of annual area had burned occurred 27.1 days earlier on average. - Significant correlations were observed between fire activity (annual area burned, number of fires ≥20 ha) and climate variables (notably CMD) for 1919–2021. - Zonal climate and fire responses differed: Central zone showed ΔTemp (1970–2021) ~0.98°C, abrupt decline in summer precipitation, and increased summer CMD, aligning with increased fire activity; Coastal zone showed the largest ΔTemp (~1.25°C) and increased CMD but did not exhibit a recent increase in fire activity due to inherent lower flammability and ignition limitations; Northern zone showed the smallest ΔTemp (~0.61°C) with increasing summer wetness and lower CMD historically, yet 2023 saw nearly 1 Mha burned by Aug 24 (exceeding previous maxima). - 2017, 2018, 2021, and 2023 rank among BC’s most severe fire seasons; 2023 surpassed the provincial area-burned record with ~1.75 Mha by August 24 and the largest recorded wildfire (Donnie Creek ~550,000 ha). - Attribution studies link 2017 BC fire season and the 2021 heat dome to human-induced climate change; the heat dome event was estimated to be ~150 times less likely absent anthropogenic warming. - CMIP6 projections (SSP2-4.5, SSP3-7.0, SSP5-8.5) indicate continued warming across all zones through 2100 and increasing summer CMD, with the highest deficits projected in the Central zone. - Compared with western U.S., BC’s area-burned increase occurred later (breakpoint ~2008 vs mid-1980s for many U.S. regions), reflecting different moisture regimes (BC CMD much lower on average than California). - Bottom-up factors have substantially modified fuels and fire behavior: extensive mountain pine beetle mortality (~15 Mha affected; e.g., 53.7% of forested area in Central zone impacted 2001–2021) increased fire spread rates in recently attacked stands; logging, land-cover changes, and long-term suppression (including reductions in prescribed/cultural burning) contributed to fuel accumulations and altered fire regimes.

Findings indicate that BC has undergone an abrupt transition to a more fire-conducive climate since the late 1990s–2000s, with drying CMD trends and longer fire seasons aligning with the observed resurgence in area burned after a century-long decline. The strong association of fire activity with CMD, coupled with insufficient compensatory precipitation increases per °C of warming, supports the hypothesis that climate change has pushed BC into a new fire epoch. Regional contrasts underscore the interaction of climate with topography, ignition regimes, and fuels: while the Central zone’s drying and fuel legacies facilitated fire increases, the Coastal zone’s inherent ignition limitations and fuel structure moderated fire activity despite rising CMD. The Northern zone’s historically lower CMD did not preclude extreme 2023 fire activity, illustrating the role of episodic weather extremes. Bottom-up controls—beetle outbreaks, logging, suppression, and disruptions to Indigenous fire stewardship—have shaped fuel structure and continuity, modulating how climate translates to fire outcomes. Projections show continued warming and drying, implying sustained or increasing wildfire potential even under moderate emissions scenarios. The convergence of climate forcing and altered fuels challenges suppression capacity, increases societal exposure, and necessitates adaptation strategies that integrate climate-smart, place-based fuel and land management and expanded use of prescribed and cultural fire.

The study documents an abrupt, climate-driven escalation in wildfire activity in BC since the mid-2000s, linked to reversals in moisture trends, lengthening fire seasons, and strong relationships between CMD and fire metrics. Regional analyses reveal heterogeneous responses governed by climate, topography, ignition regimes, and fuel legacies. Projections indicate continued warming and increased CMD, suggesting sustained elevated fire potential through the 21st century. The work underscores the need for adaptation and mitigation strategies, including landscape-scale fuel management, Indigenous-led cultural burning, optimized harvesting and species conversions, and enhanced community resilience and planning. Future research should refine attribution across seasons and regions; improve understanding of interacting disturbances (insects, drought, pathogens) and their effects on fuels and fire behavior; enhance projections of CMD and fire weather at actionable scales; evaluate effectiveness and scalability of fuel treatments across BC’s diverse ecosystems; and address data limitations in early historical climate and fire records.

- Early 20th-century climate observations are sparse in inland BC, introducing uncertainty in historical CMD and temperature estimates and trend breakpoints. - Attribution of specific fire seasons to climate change is complex due to the influence of large-scale climate oscillations, episodic weather, and interacting bottom-up factors. - Indirect climate effects on fuels (e.g., drought-induced mortality, insect/pathogen dynamics) and their net impacts on fire occurrence and behavior remain uncertain and potentially non-linear. - Variability among zones in ignition regimes, topography, and vegetation complicates generalization and mechanistic attribution. - Exclusion of fires <20 ha (due to inconsistent reporting across the century) may omit informative small-fire dynamics. - Adjustments to historical fire perimeters mitigate but may not eliminate mapping inaccuracies (e.g., unburned islands, perimeter error).