Statistical Modelling of Landslide Susceptibility along National Highway 7 from Rudraprayag to Joshimath, Indian Himalaya

ABSTRACT This study investigates landslide susceptibility over a 115 km stretch of National Highway 7 (NH-7) from Rudraprayag to Joshimath in the Indian Himalaya. Two statistical methods, Frequency Ratio (FR) and Information Value (IV), were utilized to assess eleven causal components, including geological, topographical, and land-use elements, to generate Landslide Susceptibility Zonation (LSZ) maps. The study employed a landslide inventory comprising 122 landslides and 11 preparatory factors. The characteristics of these factors were examined by generating thematic layers utilizing Geographic Information Systems (GIS). This study integrates eleven diverse parameters, including geological, topographical, hydrological, and anthropogenic factors, ensuring a more comprehensive analysis. The FR model designated 6.48% of the area as very high susceptibility and 15.56% as high susceptibility, whereas the IV model categorised 11.18% and 23.90% in these classifications, respectively. Both models were evaluated by landslide density analysis and success rate curves, with the IV model demonstrating marginally superior performance. The IV model detected 53% of landslides in the highest susceptibility zones, whereas the FR model identified 50%. Significant findings indicated that slope angles exceeding 35° and nearness to roadways, faults, and drainage systems demonstrated robust relationships with landslide occurrences. The resultant LSZ maps serve as an essential component for hazard reduction and infrastructure design in this geologically dynamic area, presenting a comprehensive framework for landslide risk evaluation in mountainous transportation routes characterized by intricate geological conditions.