Local-Scale Predictors of Fire Spread Across the U.S. (2001 - 2019)
Abstract
Over the last few decades, wildfires have increased in frequency and intensity across several land cover classes in the United States. This is due in part to land management through time including recent growth of the wildland urban interface (WUI), and a climate trending towards drier, hotter conditions in many regions. Understanding and anticipating wildfire behavior is critical to mitigating negative impacts, such as loss of life and property, decline of local economy, soil degradation, and burgeoning invasive species. Recent wildfire research, particularly studies focused in California and the Western US, have suggested that understanding trends and regimes of wildfire drivers at a local scale is necessary for effective wildfire management. Coupling meteorological data, terrain features, and fuel characteristics with daily fire progression perimeters for over 50,000 wildfire events across CONUS, we fit a random forest model to the predictors of wildfire spread for each land cover type. This work defines the primary predictors of wildfire spread at a local level, and demonstrates that several land cover types support multiple regimes of wildfire behavior, each regime defined by a hierarchy of drivers of wildfire spread. Specifically, this research quantifies the conditions at which one driver will supersede another for each land cover type. The results contribute to our understanding of wildfire behavior at a local scale, bridging the gap between experiment-based and global wildfire knowledge.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMA143.0015D
- Keywords:
-
- 3307 Boundary layer processes;
- ATMOSPHERIC PROCESSES;
- 3322 Land/atmosphere interactions;
- ATMOSPHERIC PROCESSES;
- 3379 Turbulence;
- ATMOSPHERIC PROCESSES;
- 3390 Wildland fire model;
- ATMOSPHERIC PROCESSES