Information value based landslide susceptibility zonation of Dharamshala region, northwestern Himalaya, India

This study investigates the application of statistical information value method (In V) for landslide susceptibility zonation of Dharamshala region, Kangra valley of Himachal Pradesh, India. The study area witnesses a number of landslides due to the prevailing factors such as slope angle, aspect, lithology, soil type, land use pattern, drainage density, and fault density. A landslide inventory map was prepared for the study area to understand the spatial distribution of landslides and their correlation with the prevailing causal factors. The mapped landslides covered an area of approximately 1.1 km 2 (landslide training data 0.66 km 2 and testing data 0.44 km 2 ) out of the total study area (39.3 km 2 ). Degree of correlation of the causal factors with the mapped landslides was inferred using the bivariate statistical information value (In V) method. The results show that, VHS zone has 0.65 km 2 landslide affected area whereas, the HS zone has 0.01 km 2 landslide affected area which means that the complete landslide training data (0.66 km 2 ) falls in the HS and VHS zones. The performance of the landslide susceptibility zonation map for predicting the future landslide events was inferred based on the prediction rate curve which gave 0.96 area under curve value.