Investigating the Spatio-Temporal Relationships Between Snowmelt Timing and Wildfire Area Burned in the US Mountain West

Wildfire is a growing problem in the US mountain west, with suppression costs exceeding $2 billion in 2015. Wildfire outbreaks occur in climate-driven synchronous events, and by studying the climate patterns that lead to dangerous fire conditions scientists have been able to identify numerous climatic factors that contribute to large fires. Low snow years and early snowmelt have long been hypothesized as indicators of large fire years, though there are few papers that identify this link explicitly, and those that do show great variation between the different mountainous areas of the west. In this thesis I, along with my co-investigators, explore the relationship between snowmelt timing and wildfire area burned among the many ecological systems of the US mountain west. We begin by defining a new way to identify snowmelt timing using time-series satellite imagery. We then form a theoretical and statistical framework for comparing snowmelt timing with area burned, modeled from previous climate/fire investigations. We further refined the snowmelt timing data using the complete MODIS record for the northern hemisphere. Finally, we use the MODIS-derived snowmelt timing data to investigate snowmelt/fire relationships at a moderate scale across hundreds of ecological systems of the US mountain west. Once we identified specific ecological systems that exhibit a link between snowmelt timing and wildfire we discussed the ecological implications of this relationship as well as fire-management strategies for land managers and public officials. This body of work demonstrates a substantial contribution to the fields of cryosphere studies, remote sensing, and fire ecology.