Wildfire Risk Assessment in Yellowstone National Park

Abstract

This study investigates wildfire risk in Yellowstone National Park, as informed by fuel and terrain attributes. From 2022 remote sensing data, we identify the standing dead to live tree ratios and investigate their impact on fire ignition, spread, and intensity. The abundance of standing dead trees, which can be the result of competition with other plant species, drought, insects, or diseases, is assessed for causation by the terrain features of slope and aspect and for vulnerability to fire by proximity to roads, buildings, waterways, and trails. Spatial data and a Random Forests machine learning algorithm inform a comprehensive understanding of wildfire risk. The study area covers 2.2 million acres in northwest Wyoming. Data sources include a forest assessment map, hydrography data from the National Hydrography Dataset, national park infrastructure data from the National Park Service, and terrain data from the USGS 3D Elevation Program. Python ArcPy, Pandas, and NumPy libraries, along with ArcGIS Pro for spatial visualization, are employed in data analysis. Addressing key questions on the pattern of standing dead trees, the relationship between terrain and forest state, the comparison to terrain within historic fire perimeters, and the clusters of standing dead trees that warrant priority management, our findings contribute valuable insights. Exploratory spatial data analysis reveals clustering of standing dead trees, and a two-way ANOVA identifies significant effects of slope and aspect on tree ratio. This research informs adaptive fire management strategies in Yellowstone National Park, providing essential knowledge for mitigating wildfire risks in this dynamic ecosystem.