Ayanika RAY, Department of Geography, Delhi School of Economics, University of Delhi, India
Ashis KUMAR SAHA, Department of Geography, Delhi School of Economics, University of Delhi, India
Landslides are one of the most frequent and widespread hydrogeological hazards in the Himalayan region, where every year, landslides claim many lives and result in economic losses adding up to a billion dollars (Mandal and Maiti, 2014). Therefore, the proper identification of landslide-prone areas through landslide susceptibility zonation (LSZ) mapping becomes a necessity, as it can help in minimizing landslide-associated damages and fatalities. A wide range of quantitative methods may be applied along with the techniques of remote sensing and GIS, for the preparation of accurate landslide susceptibility maps (Saha et al., 2005). The main objective of this study is to determine the applicability of three bivariate statistical methods, namely, frequency ratio (FR), information value (InfoVal) and weights-of-evidence (WoE) for landslide susceptibility mapping of the highway road section between Rangpo, Mangan and Gangtok in Sikkim, India. The bivariate methods compare a landslide distribution map with maps of different landslide causative factors, in order to determine the role of each factor in landslide formation. Here, eleven landslide causative factors have been considered for the purpose of susceptibility analysis - lithology, slope, aspect, relative relief, plan curvature, profile curvature, drainage density, distance to streams, distance to faults and thrusts, distance to roads, and land use/land cover. For susceptibility assessment, the causative factor maps were overlaid and compared with the landslide distribution map within a GIS environment. The final LSZ maps were validated and their overall performance was compared using known landslide locations within the study area and the area under the curve (AUC) method. The results show that all three methods may be used to identify areas of very high and high landslide susceptibility with a reasonable degree of accuracy.
Mots clés : Landslide Susceptibility|GIS|Frequency Ratio|Information Value|Weights-of-Evidence
A104477AR