Dynamic simulation of fire propagation in forests and rangelands using a GIS-based cellular automata model

Prediction of the way wildfires propagate in forests and rangelands is one of the critical issues in environmental protection and disaster management. This research aims to simulate wildfire propagation using a geographical information system (GIS)-based cellular automata (CA) model. The model considers the most effective spatial and temporal drivers of wildfire propagation, including wind speed and direction, type and density of vegetation, and topographic conditions. Wind speed and direction were considered changeable over the simulation process. A genetic algorithm (GA) was used to calibrate the model developed. Validation of the model was performed using an independent fire case assessed by the overall accuracy (OA) criterion and the Kappa coefficient index. The mean values obtained for the OA and the Kappa coefficient in 100 runs for two wildfire cases indicated that the proposed model could be effectively used in the simulation of wildfire propagation. The results of this model can assist fire managers in predicting fire propagation to better control wildfires.