リモートセンシングとGISのジャーナル

概要

Landslide Susceptibility Zonation Using Bivariate Models, Around Tehri Reservoir, Uttarakhand, India

Ghildiyal B, Champati ray PK, Bisht MPS and Rawat GS

Uttarakhand, the 27^th state of India, is highly susceptible to landslides probably owing to its 86% area in Himalayan terrain. In recent times, however, the landslide incidents have increased leaps and bounds mainly due to unprecedented human interventions in the form of settlements, farming, road construction, myriad of hydroelectricity projects. One such case study is done in the current study around the Tehri Dam reservoir, Uttarakhand, India. Landslide causative factors such as slope, aspect, lithology, geology and geomorphology are derived using remote sensing techniques. Thereafter, two methods, Information Value (IV) and weight of evidence (WofE) model were applied and the output was reclassified into five zones viz. very low, low, moderate, high and very high. The validation of these models was performed using area under curve (AUC) analysis, which shows the accuracy of WofE model was 83% while that of IV model was 81%. Both WofE and IV susceptibility map shows 1.95% area under very high susceptibility zone which mostly covers the area bordering the reservoir hence implementing reservoir rim to be most prone for landslides.