Bivariate landslide susceptibility analysis for parts of Kumaon Himalayas: A case study of Nainital town and its surroundings, India

Abstract

Landslides are viewed as a persistent problem in the Nainital and Almora districts of Kumaon Himalayas, since long. In this region, landslides have not only caused damage to property and life, but also affected the society by disrupting the utility services and economic activities. This study investigates application of eight crucial geo-factors that affect the frequency and distribution of landslides in Nainital town and its surroundings using the weighted multiclass index overlay method in geographic information system (GIS). The macro-scale landslide inventory map was prepared using the landslide locations identified from multi-temporal google imageries, field checks and the old landslide reports of the area. A total of 981 landslides were identified, mostly characterized under shallow translational rock and debris slides. For landslide susceptibility analysis, 70% of landslides were used, while the remaining 30% of landslides were considered for validation. Association between landslides and geo-factors were computed by means of Yules co-efficient (Yc) values and predictor ratings. The integrated landslide susceptibility map (LSM) was classified into two distinct categories through natural break method, a) three and b) five. Both these categorized maps reveal that nearly one-tenth of the study area is extremely susceptible to slope failures. The validation and accuracy assessment of maps display a score of more than 78% through receiver operating characteristic (ROC) curve. Besides, the landslide density index (R) also indicate a strong positive association of more than 65%.