UCLA

This thesis explores an approach to analyzing wildfire risk in forests, with a focus on Mount Rainier National Park. Monte Carlo simulations utilize terrain and elevation data, point process methods, and existing wildfire spread knowledge to generate risk heatmaps at a fine-grained scale. The research is driven by the following questions: How can we accurately and efficiently simulate wildfire start points and spread using terrain and elevation data and various statistical methods? What can we learn about risk factors, drawing from the simulation results in conjunction with existing wildfire knowledge?The study aims to capture the inherent uncertainties and complexities associated with wildfire behavior. Grounded in the relationship between topographical features, forest density, and climatic conditions, the simulation model effectively integrates the key factors contributing to wildfire ignition and spread. The probabilistic nature of the Monte Carlo method allows for the exploration of a wide range of wildfire scenarios, providing a nuanced understanding of risk and potential impacts. Results from the simulation identified areas of increased wildfire susceptibility and gave rise to Bayesian conclusions about risk in variables such as seasonality, moisture levels, and terrain.