Fire seasons in the western United States (US) during the years 2000--2006 have put issues surrounding structure protection from wildfires squarely in the public land management limelight with large amounts of money and firefighter exposure needed to protect residences from wildfire. No single modeling tool is currently capable of predicting wildfire ignition risk to WUI residences. This dissertation demonstrates the construction of an existing thirty-year wildfire hazard estimate at each house by pioneering a modeling system that combines results from a structure ignition assessment model (SIAM) with wildfire probability results from an ecological disturbance model (SIMPPLLE). Analysis of western Montana study area data reveal that nearly all of the structure ignition probability results modeled with extreme wildfire weather for visited homes are 1.0. This contrasts the low probabilities (0-0.05) that wildfire will reach the vegetation polygons hosting these houses. The result of the modeling system equation is that the average existing thirty-year ignition hazard estimate across the study area is roughly half of one percent.

Two suites of mitigation options are then designed, one concentrating on structural modification and fuels removal/replacement within 100 feet of each home (generally homeowner responsibility), and one using an optimization tool (MAGIS) to schedule thinning and prescribed burning treatments within 1.5 miles of homes (generally land management agency responsibility). The effectiveness of the mitigation options on both the individual model results and the combined average hazard estimates range from zero to 63 percent. While both home ignition zone mitigations and silvicultural treatments can markedly reduce wildfire hazard estimates, the former appear to provide a more linear reduction in hazard as correlated with budgets. Future work should focus on upgrading SIAM, stabilizing SIMPPLLE predictions or substituting a wildfire behavior model, and integrating the modeling system into a user-friendly GIS tool.