Fire propagation is a major concern in the world in general and inArgentinian northwestern Patagonia in particular where every year hundreds ofhectares are affected by both natural and anthropogenic forest fires. Wedeveloped an efficient cellular automata model in Graphic Processing Units(GPUs) to simulate fire propagation. The graphical advantages of GPUs wereexploded by overlapping wind direction maps, as well as vegetation, slope andaspect maps, taking into account relevant landscape characteristics for firepropagation. Stochastic propagation was performed with a probability model thatdepends on aspect, slope, wind direction and vegetation type. Implementing agenetic algorithm search strategy we show, using simulated fires, that werecover the five parameter values that characterize fire propagation. Theefficiency of the fire simulation procedure allowed us to also estimate thefire ignition point when it is unknown as well as its associated uncertainty,making this approach suitable for the analysis of fire spread based on maps ofburned areas without knowing the point of origin of the fires or how theyspread.
展开▼
