LANDSLIDE HAZARD EVALUATION OF HILL SLOPES IN AND AROUND MUSSOORIE TOWNSHIP, INDIA

Abstract

Landslides are one of the most prominent geo-environmental hazards, which also cause severe socio-economic damages. Planning of developmental and environmental conservation strategies has always been considered as an arduous task in mountainous terrains due to the occurrence of such geo-environmental hazards. The Himalayan terrain witnesses a greater number of landslide events than other physiographic divisions of India. The inherent proneness of this terrain gets aggravated due to human interference. Past incidents suggest that the inadequately planned anthropogenic activities resulted in large scale slope instability issues in the Himalaya, particularly in the Himalayan urban centres. In this context, the present research investigates the stability of hill slopes in and Mussoorie Township, a township where urbanization being taken place at a rapid pace and slope stability issues are common. Mussoorie is one of the most famous tourist destinations in the Indian State of Uttarakhand, which was established in the year 1826 by British militants. The township of Mussoorie is situated on the first range of hills running east-west parallel to the Dehradun valley and Siwaliks. Geologically, the Mussoorie area is comprised of the Mussoorie Group of rocks, which along with the Jaunasar Group constitutes the Krol Nappe. In the study area, two Formations of the Mussoorie Group are exposed, namely Krol and Tal Formations. A systematic landslide hazard evaluation of hill slopes in and around Mussoorie Township has been carried out for planning future urbanization in the area. For that, landslide susceptibility mapping and detailed stability evaluation of hazard prone slopes had been carried out. Based on detailed stability evaluation, suitable control measures were designed for each site. Further, favourable slopes suitable for future expansion of the township is also identified as part of the investigation. Landslide susceptibility mapping is the process of classifying the land surface into different categories of stability relatively, based on an estimated significance of landslide causative factors (CF). A wide range of artificial intelligence (AI) techniques of fuzzy logic and machine learning (ML) models were adopted as the methods for geographic information system (GIS) based landslide susceptibility mapping. The adopted AI techniques of fuzzy logic domain are fuzzy set procedure (FSP) and fuzzy expert system (FES), and the ML models are artificial neural network (ANN), extreme learning machine (ELM), support vector machine (SVM) and extreme learning adaptive neuro fuzzy inference system (ELANFIS). A total of 24 landslide susceptibility maps (LSM) were generated using deferent AI techniques and were validated spatially and statistically to pick the most accurate LSM of study area. Based on the LSM of study area and site observations, six very important sites, namely Surabhi landslide, Santura Devi landslide, Lal Bahadur Shastri National Academy of Administration (LBSNAA) slope, Civil Hospital landslide, Landour landslide and Woodstock landslide were chosen for the detailed stability evaluation. Each site was analysed using the finite element method (FEM) with shear strength reduction (SSR) technique to derive stability status in terms of factor of safety (FOS). Further, type of failure and causes failure were also identified as part of the investigation. Based on the analytical results and site observations, suitable control measures were suggested for the sites. Further, favourable slopes for future expansion of the township were identified. For that, a practical concept of slope favourability assessment technique was developed using AI and GIS, which relatively classifies the land surface into different classes of favourability. Here, it also takes into account various important favourability factors (FF) for the overall assessment of the area. This is one of the significant outcomes of the present research. Accordingly, a slope favourability map (SFM) was prepared, which portrays the spatial distribution of various slope favourability classes.