Abstract
Despite widespread concern over increases in wildfire severity, the mechanisms underlying this trend remain unclear, hampering our ability to mitigate the severity of future fires. There is substantial uncertainty regarding the relative roles of extreme weather conditions, which are exacerbated by climate change, and forest management, in particular differences between private industrial timber companies and public land agencies. To investigate the effects of extreme weather and forest management on fire severity, we used light detection and ranging (LiDAR) data to characterize pre‐fire forest structure across five large wildfires which burned 460,000 ha in the northern Sierra Nevada, California, USA. We found that the odds of high severity fire occurrence in these fires were 1.45 times higher on private industrial land than in publicly owned forests, an effect equivalent to a three standard deviation decrease in fuel moisture. Next, we quantified the relationships between key forest structure metrics and the probability of high severity fire, as well as how these relationships were modified by extreme weather. We found that dense, spatially homogeneous forests with high ladder fuels were more likely to burn at high severity. Extreme weather magnified the effect of density, suggesting that treatments which remove overstory trees are especially important in extreme conditions. Forests managed by private industry were more likely to be dense, spatially homogeneous, and contain high ladder fuel loads than publicly owned forests, offering a potential explanation for the increase in high‐severity fire occurrence on private industrial land. Overall, these results illustrate the need for comprehensive forest management to mitigate fire severity in a warmer future.