Evaluating the efficiency of the Dong model in determining fire vulnerability in Iran’s Zagros forests

Abstract

Recent fires in Iran’s Zagros forests have inflicted heavy, extensive losses to the environment, forests, villages, and forest inhabitants, resulting in a huge financial loss to the country. With the increasing risk of fire and the resulting losses, it has become ever more necessary to design and develop efficient fire control and prediction procedures. The present study utilizes the Dong model to develop a map of areas vulnerable to fire in the Zarivar lake forests as a representative sample of Zagros forests. The model uses as its inputs some of the most significant factors (such as vegetation, physiographic features, and the human component) that affect the fire occurrence and spread. Having assigned weights to each factor based on the model, all maps were overlapped in the ArcMap and then the region was divided into five zones. The results showed that 74% of the region was located in three classes: highly vulnerable, vulnerable, and medially vulnerable. To validate the proposed zoning map it was compared with a map based on real data obtained from previous fires. The results showed that 81% of fire incidents were located in highly vulnerable, vulnerable and medially vulnerable zones. Furthermore, the findings indicated a medium to a high degree of fire vulnerability in Zarivar Lake forests.