Landslide Susceptibility Assessment Using Frequency Ratio Model in Turung Mamring, South District of Sikkim, India

The Sikkim state is part of the Himalayan hill ranges in north-eastern India, and landslide is one of the most frequent hazards in this region, which affects every year. The present study has been focused on preparing a landslide susceptibility zonation (LSZ) map using geospatial technology and statistical analyses. Eight landslide influencing factors were identified, including slope, aspect, curvature, elevation, lithology, land use and land cover, proximity to drainage, and proximity to lineament. The detailed landslide inventory database was prepared using high-resolution satellite imageries and extensive fieldwork. The independent variables’ spatial database was prepared using high-resolution satellite imageries, a digital elevation model, published maps, and field data. The LSZ map was developed using a frequency ratio model by establishing the association between landslide-influencing factors and past landslide regions. Furthermore, the LSZ map was classified into five susceptibility zones: very low, low, medium, high and very high. The success rate of the LSZ map was validated using existing landslide inventory data through Area Under the Curve (AUC) method. The frequency ratio model has shown a fair success rate of 79.45%. The final LSZ map can be used for landslide hazard prevention, proper infrastructure planning, and geo-environmental development.