Evaluating predictive efficiency of analytical hierarchy process (AHP) for landslide susceptibility mapping in relation to unparalleled rainfall in 2018: A case study of Kuttiyadi river Basin, Kerala, India

Kuttiyadi River Basin (KRB) which falls in the Western Ghat ranges of South India, is prone to frequent landslides which are mostly triggered by continuous heavy rainfall for a short period. The landslides which happened in 2018, were the worst in the state's history.As Kerala Western Ghats is thickly populated and has unique geological and ecological characteristics, the need for Landslide Susceptibility Mapping (LSM) is an essential requisite. In this study, an attempt is made to identify landslide-prone areas using Analytical Hierarchy Process (AHP). Seven key causative factors such as slope, relative relief, drainage density, distance from drainages, land use/land cover (LULC), distance to roads, and road density were used to generate Landslide Susceptibility Index (LSI) and hence, an LSM of the basin. Thirty-two paleo landslide scars of the 2018 deluge were identified through inventory mapping which was used as test data to determine the accuracy of the AHP model using the Receiver Operating Characteristics (ROC) curve approach. The total basin area is 666 km² and out of which 503 km² (76%) falls in the stable region validated two landslides, 83 km² (13%) in moderate 13 landslides validated in this class, 57 km² (9%) in high 12 landslides have validated, and 23 km² (3%) in very high susceptible zones validated 5 landslides. The ROC curve model demonstrated 84% accuracy for thesusceptible map. The AHP method-based landslide susceptible map accurately identified possible landslide-susceptible zones in the KRB, demonstrating the efficacy of the AHP tool in LSM.