Development of Landslide Early Warning Using Rainfall Thresholds and Field Monitoring: A Case Study from Kalimpong

Abstract

Landslides in hilly regions cause lives and property loss and are considered as highly destructive natural disasters. Urbanization of highlands due to population rise makes this geomorphological process highly risky. In India, the Himalayan belt is experiencing hazardous landslides every year and the major triggering factor for such landslides is rainfall. Hence a satisfactory method for risk reduction is the development of Landslide Early Warning System (LEWS), which can help in issuing alert to the public regarding any possible landslides. This study explores in detail the different landslide forecasting methods for Kalimpong town in Darjeeling Himalayas, a highly susceptible landslide zone. Multiple rainfall thresholds are defined for the study area and have been validated using the real time filed monitoring observations using Micro Electro Mechanical Systems (MEMS) tilt sensors installed in the region. It was observed that an algorithm-based approach, called SIGMA is the best suited approach among the different rainfall thresholds. SIGMA can be used to issue multiple levels of warning based on the severity of landslides. The rainfall threshold can be used as the first line of action and warnings can be issued after verifying the field monitoring data. Thus, the combination of rainfall threshold and filed monitoring can be used to develop an efficient LEWS for Kalimpong. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.