<scp>GIS</scp> ‐Based Landslide Susceptibility Mapping of Almora District in India: A Comparison of Different Techniques

ABSTRACT Landslide susceptibility maps at the regional scale are a very efficient tool for landslide planning and management. The Himalayan region in India is highly prone to the occurrence of landslides that lead to huge losses. In this paper, we prepared the landslide susceptibility maps for the Almora district in Uttarakhand, India using different techniques. A landslide inventory was prepared using the available historical data of landslides in the region. Eight factors including slope, aspect, curvature, relative relief, distance from lineament, drainage density, geology, and distance from faults are designated as causative factors of landslides. The contribution of the causative factors and sub‐factors is modelled using different techniques. In this study, frequency ratio, statistical index, analytic hierarchy process, and landslide numerical risk factor are applied for calculating the weightage of causative factors. A hybrid method by combining the landslide numerical risk factor and analytic hierarchy process termed as AHP‐LNRF method is also proposed. Five landslide susceptibility maps of the study area were prepared in a GIS environment. The landslide susceptibility maps are divided into five categories according to the susceptibility of different zones to the occurrence of landslides. The accuracy of the output of the study is predicted by receiver operating characteristics (ROC) curve technique. It is observed that the statistical index technique had the highest prediction rate (0.789) whilst frequency ratio showed a prediction rate of 0.764. Landslide numerical risk factor technique showed the least accuracy with a prediction rate of 0.719 which is improved to 0.738 in the AHP‐LNRF technique. The AHP technique had a prediction rate of 0.735. The models and their output can be improved by improving the landslide inventory and optimising the causative factors of landslides. The results of the study can be used by the planner, managers, and researchers for landslide risk planning and management.