Landslide detection and susceptibility mapping using geological and remote sensing data: A case study of Azad Kashmir, NW Sub-Himalayas

Abstract

Azad Kashmir has undergone sever landslide episodes in the past. The main reasons behind the occurrence of landslides in this area are heavy rainfall, weak geological strata, seismic zones and active faults. Several episodes of roads blockage due to landslide is one of the major problem in this area, therefore, this study is carried out for landslide susceptibility analysis along the roads of Kotli to Trar Khel, Azad Kashmir, NW Sub-Himalayas. Geographic information system (GIS) is used to interpret remote sensing, geological strata and topographical data. Total ten anthropogenic and physical factors (aspects, fault, lithology, drainage density, curvature, distance to roads, landuse landcover, slope, seismic and elevation) were examined by Multi-criteria Decision Analysis (MCDA) technique. The thematic layers of all factors were prepared for the preparation of the final susceptibility map by using a weighted overlay method (WOM). The susceptibility map classified the study area into four susceptible classes of low, moderate, high and very high. The spatial analysis and result outcomes depicted that active faults, slope gradient and landuse landcover may facilitate landslide phenomena. The results were verified by landslide density analyses (LDA), receiver operator characteristics (ROC) and area under curve (AUC) with predictive accuracy of 77 % that is rated as satisfactory by many researchers.