Landslide Inventory and Susceptibility Assessment Using GIS and Weighted Overlay Model (WOM) Along a Part of National Highway-1 (NH-1), Northwestern Himalaya

The present study aims to create a historical landslide inventory database and identify landslide susceptibility zones along a part of NH-1 from Sonamarg to Kargil in the Northwestern Himalayas. A comprehensive landslide-prone map is created using landslide-influencing geo-environmental elements such as elevation, curvature, slope, distance to drainage, aspect, distance to lineament, geomorphology, lithology, land use/land cover, and rainfall. Furthermore, a Geographic Information System (GIS) based weighted overlay model (WOM) was employed to generate a landslide susceptibility map using ten contributing parameters. All the parameters were weighted based on their corresponding importance in the landslide occurrence. These weights were normalized so that the sum of normalized weights was equal to one. Further, the generated landslide susceptibility map identified four vulnerable zones viz: Very High (67.36 km2), High (456.96 km2), Moderate (147.96 km2), and Low (6.30 km2). The study also used secondary sources and multiple-field surveys to locate and map past landslides in a study region. Landslide data was also used to validate the model’s accuracy. Thus, the model’s accuracy was evaluated using the receiver operating characteristic (ROC) approach. Likewise, the AUC value, which measures the area under the curve, was 0.686 (68.60% of prediction accuracy). Furthermore, the WOM’s satisfactory prediction performance yielded a reliable landslide susceptibility map. Moreover, the landslide susceptibility model can be used in other Himalayan regions to mitigate landslides and plan long-term land use.