Stochastic model for spatial process: a case of fire arrival process

Disturbance is critical in determining how diverse ecological systems function. Fires are one of major agents of disturbance. In this research, the fire arrival process is modeled assuming that events are generated by some underlying random mechanism and simulated based on the developed stochastic model. This process was applied to the fire succession process in the Great Victoria Desert of Australia. The fire arrival points were mapped using Landscape Multi-Spectral Scanner (MSS) data covering the period from 1972 to 1989. In this study, nonhomogeneous planar Poisson process where the event processes are dependent on the location of the events is proposed as a model for fire arrival process and rejection sampling thinning the homogeneous Poisson process is used for its simulation