Mapping the landslide susceptibility considering future land‑use land‑cover scenario

Abstract

The landslide susceptibility (LS) of any mountainous region is signifcantly affected by the land-use land-cover (LULC) change. Recently, LULC change effects on landslides have been investigated by many researchers. However, the future prediction of the LS using these LULC changes has not been quantifed. The main objective of this study is to predict the future LS map considering the future LULC change scenario for the Tehri region, India. To achieve this objective, we frst prepared a geospatial database of past landslide events. These events data were clustered into three major temporal categories, 2005–2010, 2010–2015, and 2015–2020. Second, the artifcial neural network (ANN) approach was adopted to prepare LS maps for the years 2010, 2015, and 2020. Then, for the same years, LULC maps were also developed. Third, the future scenario of LULC for the year 2030 was simulated using the ANNcellular automata model, and the future LULC changes were derived using the change detection technique. Finally, the future LS map for 2030 was projected using derived future LULC changes. The LULC change results reveal that the region is expected to see a signifcant growth in the built-up area by 34.1%, water body by 6.3%, and agriculture land by 1%. Further, a shrink in dense forest area by 2.4% and sparse forest area by 0.9% is expected in the future. Additionally, the projected LS results reveal a 33% increment in the very high LS zone. This information about the increase in future LS due to rapid urban growth in the mountains can help the various government agencies to scientifcally plan the various developmental activities.