Projected Changes in Both Mean Climate and Climate Variability Drive Substantial Increases in Extreme Fire Weather in the Western United States

Wildfire frequency, extent, and duration in the western United States are projected to increase throughout the twenty-first century as dry and warm conditions become more frequent, widespread, and persistent. However, there is limited knowledge about how changes in mean climate and changes in climate variability, individually and together, contribute to these increases. To disentangle these effects, we use a 100-member ensemble of climate simulations produced with the Community Earth System Model, version 2, under historic and shared socioeconomic pathway (SSP) 3-7.0 forcing from 1980 to 2100. Using the Canadian forest fire weather index (FWI) to quantify fire-related meteorological conditions in the simulations, we select extreme FWI thresholds relative to a baseline distribution centered at 1980 (“fixed threshold”) and an evolving distribution (“moving threshold”) to identify spatiotemporally connected extreme fire weather events. By 2100, the frequency, area, and number of event days increase when considering a fixed threshold, and events are up to 48C warmer on average. Moreover, events over the Pacific Coast expand northwestward, while those over the Four Corners region expand westward and northeastward. Changes in event frequency are small and not significant when considering a moving baseline but become spatiotemporally more connected and widespread, reflecting nonthermodynamically driven changes. Increases in the mean of maximum temperature, as well as changes in both the variability and mean of relative humidity, drive the largest increases in extreme fire weather event area and number of event days. By quantifying how changes in the mean and variability of climate variables impact wildfire conditions in the western United States, our study has implications for increasing resiliency to wildfire risk under climate change.