In fire-prone forests, self-reinforcing fire behavior may generate a mosaic of vegetation types and structures. In forests long subject to fire exclusion, such feedbacks may result in forest loss when surface and canopy fuel accumulations lead to unusually severe fires. We examined drivers of fire severity in one large (>1000 km2) wildfire in the western United States, the Rim Fire in the Sierra Nevada, California, and how it was influenced by severity of 21 previous fires to examine the influences on (1) the severity of the first fire since 1984 and (2) reburn severity. The random forest machine-learning statistical model was used to predict satellite-derived fire severity classes from geospatial datasets of fire history, topographic setting, weather, and vegetation type. Topography and inferred weather were the most important variables influencing the previous burn. Previous fire severity was the most important factor influencing reburn severity, and areas tended to reburn at the same severity class as the previous burn. However, areas reburned in <15 yr burned at lower severity than expected. Previous fire severity and Rim Fire severity were higher on ridges, at intermediate elevations (~750–1250 m), and on slopes <30°, indicating a consistent effect of topography on fire severity patterns in these forests. Areas burned with low severity prescribed fires burned at low severity again in the Rim Fire, and areas with long fire-free periods burned at higher severity. This fire history effect suggests that prescribed burning was an effective management tool, leading to lower fire severity in the previous burns and the subsequent reburn. Our results show that self-reinforcing fire behavior results mainly from effects of vegetation structure and fuels on fire severity and that this behavior is mediated by topographic setting and the time since last fire.
All Science Journal Classification (ASJC) codes
- Ecology, Evolution, Behavior and Systematics