Enhancing Landslide Hazard Assessment Using Monte Carlo Simulations and Improved Soil Thickness Mapping in Uttarakhand, India

Landslides are a major threat to communities and infrastructure in mountainous regions, highlighting the need for precise landslide hazard assessments to enhance risk management strategies. This study aimed to enhance landslide hazard assessment methodologies by incorporating Monte Carlo simulations (MCS) and refining soil thickness mapping techniques. The study area, a crucial road section situated in the Joshimath region of Uttarakhand, India, is characterized by steep terrain and high rainfall intensity, making it prone to landslides triggered by transient rainfall infiltration. Previous studies have identified limitations in existing methodologies, particularly regarding the characterization of uncertainties in geotechnical and hydrological factors and the accuracy of soil thickness mapping. To address these limitations, the present study applied MCS separately for each layer within the TRIGRS model, allowing for a detailed understanding of uncertainties associated with key factors such as cohesion, soil unit weight and internal friction angle. Additionally, the study developed the geomorphologically indexed soil thickness (GIST) – MCS model for soil thickness mapping, leveraging its higher accuracy compared to traditional models like the Z-model. The findings of the study demonstrate that the MCS-based approach significantly improves the accuracy and reliability of landslide hazard assessments by providing more precise estimations of factor of safety (FoS) values. Furthermore, the GIST-MCS model offers enhanced accuracy in soil thickness estimation, leading to more reliable landslide hazard maps. Comparative analyses between hazard maps generated using MCS-derived input parameters and those based on traditional approaches highlight the superiority of the MCS-based methodology.