Grant: Discovery Grants Program - Individual

The main objectives of my research program are to develop tools and methodology for detecting nonlinear trends in non-negative spatiotemporal data, emphasizing the analysis of spatially referenced time series of counts and proportions. My research is motivated by the analysis of historical forest fire, fire-weather and lightning data recorded by Canadian forest fire management agencies. Statistical and Industrial Engineering considerations touch on areas covering mixtures, analysis of extremes, nonlinear trend analysis, queueing theory and estimation, and other topics related to the modelling and analysis of stochastic processes. A long term goal is the construction of a framework for investigating forest fire and climate-change related hypotheses across a provincial or national study region, consisting of an irregular lattice with polygons that can be assumed relatively internally homogeneous with respect to ecological characteristics and fire-management strategies. I expect to develop and make available a comprehensive suite of tools to test for nonlinear trends under various spatial, temporal and spatiotemporal scenarios. In addition, I plan to construct a spatial-temporal queueing system for forest fire management. There, wildfires are regarded as the customers arriving randomly in space-time, while suppression resources are the available servers. Fire characteristics, such as intensity or perceived risk, will influence its suppression priority in the queue. Results from my research program are expected to be of interest to statistical and operational research communities. Applied results are expected to impact research and planning decisions within Canadian forest fire management agencies. Collaborations with individuals therein are ongoing.