Data-driven evidential belief function (EBF) model in exploring landslide susceptibility zones for the Darjeeling Himalaya, India

Abstract

In the present study, data-driven evidential belief function model (belief function) was employed to generate landslides susceptibility index map of Darjeeling Himalaya considering 15 landslide causative factors, which grouped into six categories, i.e. geomorphological factors (elevation, aspect, slope, curvature), lithological factors (geology, soil, lineament density, distance to lineament), hydrologic factors (drainage density, distance to drainage, stream power index, topographic wetted index), triggering factor (rainfall), protective factor (normalized differential vegetation index) and anthropogenic factor (land use and land cover). Total 2079 landslide locations were mapped and randomly divided it into training datasets (70% landslide locations) and validation datasets (30% landslide locations). The resultant susceptibility map was divided into five different susceptibility zones i.e. very low, low, moderate, high and very high which covered 5.60%, 25.65%, 34.47%, 24.67% and 9.61% area respectively of the Darjeeling Himalaya. Receiver operating characteristics curve suggested that 80.20% prediction accuracy of the prepared map whereas frequency ratio plot indicated towards the ideal landslides susceptibility index map.

Keywords:

• Darjeeling Himalaya
• factors group models
• landslide susceptibility index map
• receiver operating characteristics curve
• frequency ratio plot

Acknowledgement

The authors would like to express their sincere thanks to Survey of India (SOI), Geological Survey of India (GSI) and National Bureau of Soil Survey and Land Use Planning (NBSS&LUP) for providing necessary data, facilities and support during the study period.

Disclosure statement

No potential conflict of interest was reported by the author(s).