Predicting forest fire occurrence in Ontario

This dissertation documents the investigation and identification of factors influencing forest fire occurrence and describes the development and structure of models of forest fire occurrence in the forests of Ontario. An exploratory analysis of lightning strike density and rainfall relationships was carried out to determine if such a relationship could be used to improve inter-station rainfall estimation: only weak relationships between lightning and rainfall could be identified. A new moisture index, the Sheltered Duff Moisture Code (SDMC), was developed to improve lightning fire prediction capability across the country. This new index tracks forest floor moisture content at heavily sheltered areas of a stand (where lightning ignitions generally occur) better than the currently used Duff Moisture Code of the Canadian Forest Fire Weather Index (FWI) System. The SDMC can be integrated easily with the FWI System, and be adopted immediately by fire management agencies in Ontario and across the country. Logistic regression models of the probability of a lightning strike igniting a fire were developed for Ontario's fire region. The SDMC was shown to be a strong and consistent predictor in each of these models, while the presence of positive lightning strikes in an area, the timing of a lightning storm, and weather and fuel moisture conditions following a storm were also shown to have significant influence on the probability of lightning fire ignition. These daily models of lightning fire occurrence capture both low and extremely high periods of lightning fire activity well. Poisson regression methods were used to develop people-caused fire occurrence prediction models for each of the ecoregions in Ontario's fire region. Fuel moisture indices and an index of the probability of sustained flaming were identified as significant predictors of people-caused fire activity. The application of these newly developed predictive models of lightning- and people-caused fire occurrence are discussed in detail in terms of both the current daily operations of a fire management agency and the development of climate change impacts scenarios.