A mathematical model of spot fires and their management implications

In spite of considerable effort to predict wildfire behavior, the effects of firebrand lift-off and the resulting spot fires on fire propagation are still poorly understood. Horizontal discontinuities termed "fuelbreaks" can impede surface fire spread, however long-distance spot fires allow wildfires to ignite and propagate beyond (i.e., jump) fuelbreaks. Current fire behavior programs aid fire managers in predicting potential wild fire behavior, but these programs lack appropriate methods to determine firebrand and spot fire behavior. In this thesis, we developed a cellular automata model integrating key mathematical models and a recent model that determines firebrand landing patterns. Using our model we varied values of wind speed 6 meters above treetop, surface fuel loading, surface fuel moisture content, and canopy base height to examine two scenarios: the probability of a spot fire igniting beyond a fuelbreak of various widths, and how spot fires affect the surface fire's rate of spread. This was tested in the context of a pine forest. Canopy base height had the greatest influence for both scenarios, followed by wind speed 6 meters above treetop, surface fuel moisture content and surface fuel loading. Our results suggest that the average rate of spread with spot fires is constant, and fire managers would benefit from a mathematical model that determines this rate. Since spot fires impact fire propagation updating fire behavior programs utilized by fire managers, or creating a new fire behavior program, is essential to improve wild fire prediction.