Exploring landslide susceptible zones by analytic hierarchy process (AHP) for the Gish River Basin, West Bengal, India

Landslide is a major threat in the Darjeeling Himalaya within sub-humid climate. Hence proper identification of landslide susceptible zone (LSZ) is very much essential. In this case a multi criterion evaluation approach is applied using thirteen selected indicators. The parameters are categorized into five categories viz. anthropogenic factor, surface causal factor, lithological causal factor, triggering factor and protective factor. Weighted composite model is prepared adopting weighting base as Analytic Hierarchy Process. The obtained result shows that near about 19.92 sq. km (approximately 7.52%) area within the basin is highly susceptible for landslides. High drainage density (avg. 4.31 km/sq. km), relatively steeper slope (>10 degrees) accelerate this process. Beside the main landslide susceptibility layers five separate models of five causal factor groups are prepared and correlated with final LSZ for understanding the priority cluster. Lithological factors cluster appears as a dominant factor group (correlation value 0.95). This LSZ model is also validated by frequency as well as areal density of historical landslides. Beside this, the validation by ROC curve shows 84.00% area under the curve. So, the model can be treated as relevant.