COMPARATIVE LANDSLIDE SUSCEPTIBILITY MAPPING IN PITHORAGARH DISTRICT, UTTARAKHAND: A MULTI-MODEL APPROACH USING FREQUENCY RATIO, SHANNON ENTROPY AND ANALYTICAL HIERARCHY PROCESS

Abstract

Pithoragarh District, located in the Kumaon division of the Uttarakhand Himalaya, remains highly vulnerable to heavy and dangerous landslides resulting from fractured bedrock, steep slopes, long-standing tectonic activity, and heavy seasonal rainfall. As a first step toward facilitating hazard targeting and management, this paper primarily carries out spatially comparative susceptibility assessments of three frequency-ratio based methods, Frequency Ratio (FR), Shannon Entropy (SE), and Analytical Hierarchy Process (AHP), in this complex geological mountainous setting. Location data for 366 past slope failures were gathered and integrated into an inventory database and linked to ten topographic and environmental factors: slope, aspect, hillshade, curvature, rock type, distance to roads, distance to drainage, distance to structural lineaments, Topographic wetness index, and Terrain ruggedness index. The sample was split 70:30 for model fitting and independent testing, respectively. Among the three methods that were tested, AHP gave the best result with the AUC value of 0.839, whereas SE and FR produced AUC values of 0.836 and 0.835, correspondingly, all three having the very good accuracy range comfortably. Expert assignments of weights in the AHP schema were corroborated by a Consistency Ratio of 0.009, which indicates a high level of internal consistency of the judgment matrix. The two major factors that influence the occurrence of slope movements and kept revealing their dominant roles were the proximity to roads and the lithology of the bedrock. Final hazard maps provide spatially differentiated hazard zones for the district, which can be used by planners and disaster management offices as decision-making tools in at-risk mountainous areas of Uttarakhand.