Landslide susceptibility and risk: a micro level study from the Balason River basin in Darjeeling Himalaya

Abstract

This present study primarily focuses on the susceptibility analysis of the settlement area and metalled road based on the susceptibility index maps/susceptibility zonation maps. Susceptibility index maps were prepared using thee recognized techniques, i.e., frequency ratio (FR), analytical hierarchy process (AHP), and logistic regression (LR) models. Elevation, slope angle, slope aspect, slope curvature, stream power index (SPI), topographic wetted index/compound topographic index (TWI/CTI), land use/land cover (LULC), normalized differential vegetation index (NDVI), lineament number density, distance from lineament (D2L), drainage density (DD), distance to drainage (D2D), geology, soil, and rainfall were incorporated for the preparation landslide susceptibility zonation maps. The prediction accuracy of three aforesaid models was 93.6, 89.3, and 94.7%, respectively. The result indicated that the total settlement area and metalled road lengths were 48.84 km2 and 314.38 km, respectively, within basin boundary. Settlement areas of 33.11, 35.28, and 60.39% were located on the very high susceptibility zones based on the FR, AHP, and LR models, respectively. Similarly, 57.61, 17.25, and 3.08% metalled road out of total road length were found on the very high susceptibility zones based on the aforesaid three models, respectively. Total 130 mouzas fall within the basin boundary; out of these, the most vulnerable mouza was Bara Adalpur Dwitiya Khanda under Matigara block because 100% of the settlement area falls in the category of very high susceptibility zones on the basis of FR model. Dhajea Tea Garden and Manjua Forest mouza under Jorebunglow Sukiapokhri and Mirik block, respectively, were most susceptible to landslide hazard in respect of road based on the AHP model. The findings of the research can be utilized to prioritize risk management so that the losses and damages of properties and lives can be minimized due to future landslide event.