|dc.description.abstract||Ecologists and wildlife managers alike have explored the role of fire as an ecosystem disturbance for decades and, yet, the role of scale remains poorly understood in pyrogeography. Understanding how wildfire occurs on the landscape and, furthermore, how these trends will change in the future provides an enhanced understanding of vegetative patterns, successional changes and biome distributions. As scientific research begins to account for the effects of climate change, predictive modeling will remain one of the foremost tools in understanding how present-day trends will begin to change. This study employs a series of spatial modeling techniques to examine which factors are most influential on the presence of wildfire hotspots on the landscape and which factors may be influential on areas devoid of wildfire occurrence entirely. Clustering algorithms were used to identify wildfire hotspots across the study area and targeted pseudo-absence points were created outside the bounds of these clusters.
The resulting presence/absence points were analyzed within physiographic regions and a predictive model was fit to the data. Analysis of common covariates, such as climatic variables, land use, and topography allowed this study to not just fit a model to wildfire distribution, but inform comparable studies conducted anywhere similar data are available. As different aspects of climate change begin to exert influence on ecosystems globally, this study sheds light on how fire regimes may change with it.||en