Efficacy of Landslide Susceptibility Maps Prepared Using Different Bivariate Methods: Case Study from Mussoorie Township, Garhwal Himalaya

Abstract

Four bivariate methods viz frequency ratio, weight of evidence, Yule’s coefficients and information value were utilized for the preparation of the landslide susceptibility map of the hilly township of Mussoorie. Two scenarios, one with partitioning landslide inventory prepared till 2019 with 70% landslides, and another with all the active landslides till 2019 were used for the preparation of landslide susceptibility maps. In order to understand the efficacy and the reliability of each of the bivariate approach used under both the scenarios, the maps thus obtained were overlaid with the landslides that occurred in the area during excessive rainfall of 2020. It has been noted that the landslide susceptibility maps prepared using four different bivariate methods exhibit more or less similar results, nevertheless of all the four methods used, information value method indicate that more than twice the area (∼38%) fall in high and very high landslide susceptible zones in scenario-I. The scenario-II exhibits higher percentage of area falling in high and very high landslide susceptible zones for all the methods as compared to scenario-I, still the Information value indicate the highest percent of area (∼31%) falling in the high and very high landslide susceptibility zones. The validation of the maps prepared using scenario-I exhibit 58–75% of the 2020 landslides occur in high and very high landslide susceptibility zones, whereas in scenario-II, 57–72% of the 2020 landslides falls in high and very high landslide susceptibility zones. Finally the weight of evidence method and information value method indicate the higher prediction accuracy under both the scenarios