By Mitchell Zimmer
PhD student Jonathan Lee's presentation "Parallel computation of stochastic lattice models" received The 2011 Student Research Presentation Award at the Statistical Society of Canada's 2011 meeting at Wolfville, Nova Scotia. Using the spread of forest fires as an example, Lee’s work developed modeling tools to represent spatial locations in a grid structure so that each cell of the lattice evolves according to not only what random variables are contained within each cell (such as the availability of fuel) but also information communicated from neighbouring cells (whether that neighbouring cell is on fire) as well as external factors for each cell (such as weather and topography). These transition rules can describe the phenomenon of the growth of a forest fire front. Lee has incorporated parallelization into the calculations so that there is a decrease in computation time making it is easier to increase the resolution of the lattice, to work with larger lattices and to work with more complicated models. Lee's work has important ramifications in determining insurance rates.