Modelling of forest fire spread using Cellular Automata

This research was done to model the fire spread using Alexandridis model and cellular automata method in part of District Three of Neka Zalemroud forests. The effective factors on fire spread were elevation, slope, wind velocity and direction, vegetation type and vegetation density, based on the mentioned model (Alexandridis model). To simulate of the fire spread and evaluate its accuracy, the required data of actual fire including the fire start point, fire area, wind velocity and direction in fire time was provided. All effective parameters for actual fire confine were provided and coded in GIS. Finally the fire spread model was implemented using programming of Alexandridis model in MATLAB, regarding to Moore neighborhood and loading of digital layers of all effective parameters for actual fire confine. Some cellular situations and fire transfer rules were considered to predict the fire spread based on cellular automata method. The dynamic exhibition of the fire spread was presented by selecting the fire start point and number of burned and unburned cells was determined. Results of the model output including fire spread direction and form was compared with the actual fire confine to evaluate the accuracy of the used model qualitatively. The total precision and Kappa index were used to evaluate the accuracy of the used model quantitatively. The total precision and Kappa index were 0.88 and 0.74 respectively that it can show the precision of prediction of the fire spread model and the cellular automata method in modeling of fire spread in current research.