Forest fire spread simulation algorithm based on cellular automata

Traditional models result in low efficiency and poor accuracy when simulating the spread of large-scale forest fires. We constructed an improved model that couples cellular automata with an existing forest fire model to ensure better time accuracy of forest fire spread. Our model considers the impact of time steps on simulation accuracy to provide an optimal time step value. The model was tested using a case study of forest fire spread at Daxing’an Mountain in May 2006. The results show that the optimal time step for the forest fire spread geographic cellular automata simulation algorithm is 1/8 of the time taken for cellular material to be completely combusted. When compared with real fire data from Landsat Thematic Mapper (TM) images, our model was found to have high temporal and spatial consistency, with a mean Kappa coefficient of 0.6352 and mean accuracy of 87.89%. This algorithm can be used to simulate and predict forest fire spread and is also reversible (i.e., it can identify fire source points).