Fire risk zoning of forest areas using an integrated method of artificial neural network and spatial information system, Murray study: Shimbar protected area, Khuzestan province

Simultaneous use of Geographic Information System (GIS) and methods based on artificial intelligence has always led to good results in research in the field of natural resources. This study was conducted in the same format and in order to prioritize the factors affecting the spread of fire and identify high risk areas in the forests of the Shimbar protected area, in this regard, factors for the artificial neural network method were considered. In implementing the artificial neural network method with factors affecting forest fire, a fire risk zoning map with five classes of very low risk, low risk, medium risk, high risk, very high risk with a total accuracy of 0.83 and RMSE error is prepared. It is 0.75. The results showed that 20% of the area in the middle class, 11% of the area in the upper class and 10% of the area in the very upper class, the potential for fire, also the most important variables affecting the occurrence of fire including distance from the river, type Lands, altitude and minimum temperature. It is concluded that according to the considered factors, the integrated models of artificial neural network (ANN) and spatial information system have a high efficiency in preparing a fire risk zoning map and it is suggested that these models be used to prevent, control and manage fires elsewhere. The country should also be used on a large scale.