| Description |
The relationship between wildfire ignition and environment was investigated in the state of Utah with consideration of ignition source and fire size. The dataset consisted of 24 years (1992-2015) of geospatial fire records consolidated by Karen Short of the US Forest Service. Environmental suitability influenced all sources of ignition and was modeled in this study with the machine learning algorithm Maximum Entropy. Within the state, the spatial distribution of lightning and human-caused ignition was found not to be random. Results highlighted these spatial trends as well as the degree to which environmental variables influenced ignition potential. Lightning-caused ignition was found to be most influenced by climate, while human-caused ignition was found to be most influenced by proximity to roads. The results indicated vegetation type and condition immediately surrounding an ignition point influence fire spread and size. These findings are congruent with related research but are unique to Utah and applicable to policy development and long-term management of state, federal, and private lands |
