Evaluation of landslide susceptibility models: A comparative study on the part of Western Ghat Region, India

Abstract

The present research focuses on evaluating the effectiveness of landslide susceptibility using fuzzy gamma operator (FGO) and analytical hierarchy process (AHP) model in part of Western Ghat, India. The landslide inventory data and various causative factors were used to generate the landslide susceptibility index (LSI) map using geospatial technology. The LSI map classified into five classes and finally validated by Area Under Curve (AUC) method for accuracy assessment. The AUC result of success and prediction rates depicts that FGO model has high accuracy value than AHP model. However, both models are highly correlated (r = 0.976). Among the all causative factors slope, lithology, land use land cover, geomorphology and lineament proximity are strongly correlated (r = >0.9) with the landslide susceptibility between the two models. The research outcome will be helpful for future spatial planning in the study region.