Climate Change and Future Wildfire in the Western United States: An Ecological Approach to Nonstationarity

TLDR: The authors used a finer ecological classification and fire-relevant climate predictors, and created statistical models linking climate and wildfire area burned for ecosections, which are geographic delineations based on biophysical variables.

Abstract: We developed ecologically based climate-fire projections for the western United States. Using a finer ecological classification and fire-relevant climate predictors, we created statistical models linking climate and wildfire area burned for ecosections, which are geographic delineations based on biophysical variables. The results indicate a gradient from purely fuel-limited (antecedent positive water balance anomalies or negative energy balance anomalies) to purely flammability-limited (negative water balance anomalies or positive energy balance anomalies) fire regimes across ecosections. Although there are other influences (such as human ignitions and management) on fire occurrence and area burned, seasonal climate significantly explains interannual fire area burned. Differences in the role of climate across ecosections are not random, and the relative dominance of climate predictors allows objective classification of ecosection climate-fire relationships. Expected future trends in area burned range from massive increases, primarily in flammability limited systems near the middle of the water balance deficit distribution, to substantial decreases, in fuel-limited nonforested systems. We predict increasing area burned in most flammability-limited systems but predict decreasing area burned in primarily fuel-limited systems with a flammability-limited (“hybrid”) component. Compared to 2030–2059 (2040s), projected area burned for 2070–2099 (2080s) increases much more in the flammability and flammability-dominated hybrid systems than those with equal control and continues to decrease in fuel-limited hybrid systems. Exceedance probabilities for historical 95th percentile fire years are larger in exclusively flammability-limited ecosections than in those with fuel controls. Filtering the projected results using a fire-rotation constraint minimizes overprojection due to static vegetation assumptions, making projections more conservative. Plain Language Summary Most people, including many familiar with fire ecology and future climate, assume that the area burned by wildfire will increase in a warmer climate. This depends a lot on what kind of ecosystem wemean. In all ecosystems, fuels must be available to fire for fires to get very big, but the climate controls on those fuels vary widely with vegetation. In wetter forests, it takes an abnormally warm, dry year to make normally wet fuels available. But in many drier ecosystems, fuels are dry enough to burn most years—whether fires get big depends also on whether there is sufficient fuel available to carry fires over large areas. In this kind of vegetation, abnormally wet years in the year prior to fire can create larger or more connected fuels that then lead to larger fires. In this study, we use this concept to investigate how future area burned might be affected by climate change. We found that some ecosystems will burn much more, just as expected. But some will actually burn less. We characterized these futures for 70 different ecosystems around the West. The similarities and differences illustrate the range of futures that might be expected under climate change.