Using cellular automata and imperialist competitive algorithm for modeling the spreading of forest fire

The forest fire phenomenon is one of the natural disasters that threatens a large part of the forest. Due to the devastating effects of fire on the forest, it is very important to carry out research using modern methods for modeling forest fire spreading. In this research, we aimed to model the Golestan forest fire spreading based on cellular automation, which was used to calibrate it using imperialist competitive algorithm. In order to achieve more accurate results, neighboring filters and different spatial resolution were used to obtain optimal values for these parameters. The results showed that the proposed method of this study for the forest fire of July 15, 2011 is, at best, has kappa index and overall accuracy equal to 0.841 and 0.893. This result was obtained by using the Von Neumann Neighborhood Filter and the spatial resolution of 5 meters. According to the results, it can be seen that the use of data with a smaller spatial resolution capability will provide a higher accuracy and as far as smaller neighborhoods are used, the accuracy of modeling will be improved.

Keywords:

Modeling , Forest Fire Spreading , Cellular Automata , Imperialist Competitive Algorithm