Landslide Hazard Mapping Using AHP and Shannon Entropy Models in Cherikabad of Urmia Watershed

Understanding the factors, conditions of occurrence, and development of landslides in order to assess and zoning the risk of its occurrence provides the possibility of achieving methods by which the dangers and damage caused by landslides can be prevented. The purpose of this study is to prepare a landslide risk potential map with AHP and Shannon entropy models in the Cherikabad watershed of Urmia. Initially, ninety-five landslide points were recorded using GPS through field visits and Google Earth. Layers of mean annual rainfall, slope degree, slope Aspect, elevation classes, land use, Lithology, distance to River, distance to road, distance to fault, distance to village, and normalized index of vegetation difference (NDVI) As factors affecting the occurrence of landslides in the basin were selected and maps of these layers were prepared in the GIS environment and digitized. After combining the layers of factors affecting the occurrence of landslide, the landslide risk potential map was prepared for two models of Analytic hierarchy process (AHP) and Shannon entropy. The results of preparing risk maps using the Areas Under the Receiver Operating Characteristic Curve with the area below the 99% significance curve showed that both models have good performance in evaluation. Shannon entropy model with 87.9% AUC and very good performance compared to AHP model with 70.6% AUC and a good performance class, has a higher accuracy in preparing the maps. According to Shannon's entropy model, 32% of the basin is located in high and very high-risk classes, which should be considered by the relevant authorities to implement relief measures.