GIS-Based Multi-temporal Analysis of Landslide Susceptibility Mapping Along the Ramban-Banihal Road Section of National Highway-44, Jammu and Kashmir

Abstract

One of the most landslide-prone areas in the Jammu and Kashmir Himalaya is the Ramban-Banihal road section of National Highway - 44, which is characterized by frequent occurrences of landslides. These events cause damage to infrastructure and fatalities. To manage this landslide hazard, it is important to prepare landslide susceptibility map (LSM) by taking into account the significant causative factors in the region. In this study, a frequency ratio model was applied to assess the impact of causative factors to landslides and to prepare a landslide susceptibility map of the study area on the Geographic Information System. A total of 81 and 106 landslide events were identified from the Google Earth Image for the years 2017 and 2020 respectively, among which 70% of the landslide incidences were utilized for training the model, while the remaining 30% were utilized for testing. Thirteen factors, including slope, aspect, curvature, lithology, geomorphic, LULC, distance from fault, distance from streams, distance to road, relative relief, stream power index, topographic wetness index, terrain ruggedness index was analyzed and integrated with landslide occurrences. These factors were weighted based on the presence of landslides in their respective class and integrated in maps for the year 2017, and for the year 2019 were generated using a GIS platform. The obtained maps were categorized into three different landslide susceptibility classes, i.e., low, moderate and high. The results for the years 2017 and 2019–20 demonstrate a decrease in the region's landslide susceptibility i.e., high zone from 22.39% to 14.19% and moderate zone from 44.95% to 38.34% while an increase in the low zone landslide susceptibility from 32.67% to 47.48%. This indicates a decrease in the proportion of high and moderate susceptible zones over the previous two years because of majority impact by the anthropogenic activities. Area Under Curve (AUC) of the Receiver Operating Characteristic curve (ROC) have been used to assess the accuracy of the maps and have revealed an excellent accuracy of 0.918 and 0.942 of maps for the years of 2017 and 2019–2020 using a frequency ratio technique. These LSM’s can be used to by communities, engineers, and land-use authorities about different landslide susceptibility zones and also for future land use planning in order to reduce the damage caused by landslides. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.