HAPpy Hour Seminar - Pre-fire climate and snow drought explain most of western U.S. burned area variability

This study uses machine learning models to quantitatively relate pre-fire climate conditions to summer fire burned area across the snow-dominant western United States (U.S.) from 1984 to 2020. Predicted burned area based on winter and spring climate conditions alone captures 87% of the interannual variability of observed summer burned area, which indicates that climate variability in antecedent seasons has been the primary driver to broad-scale changes in summer wildfire activity in the western U.S. over the recent four decades. For snow-dominant regions, the pre-fire snow drought area is found to be the most valuable predictor for summer burned area relative to other antecedent climate predictors (e.g., vapor pressure deficit, temperature, precipitation, evapotranspiration). Namely, spring snowpack memory is realized through anomalies in land and atmospheric moisture during the summer which modulates fire hazard. This study highlights the important role of snow drought in seasonal-to-subseasonal forecasts of summer burned area.