Applicability of complex rainfall thresholds for Territorial Landslide Early Warning Systems (Te-LEWS) in the Himalayas

Increasing occurrences of landslides as compound events under extreme-rainfall demand the development of landslide early warning systems at the territorial scale. Traditionally, rainfall Intensity-Duration (ID) thresholds are adopted in these Territorial Landslide Early Warning Systems (Te-LEWS). However, the usage of complex rainfall thresholds in Te-LEWS has proven successful across the globe since these thresholds account for the terrain conditions in direct or indirect ways, unlike the ID thresholds which consider only the rainfall characteristics. In Japan, the Soil Water Index (SWI), a complex rainfall threshold, is implemented successfully to predict rainfall-induced sediment disasters. Implementing such complex thresholds in new geological settings is very challenging. Here, we present a framework to test the SWI threshold to predict shallow landslides and debris flows in the Himalayas. By analysing the landslides' historical occurrences and their corresponding triggering rainfall intensities from satellite-derived hourly precipitation data, we determine the critical threshold of SWI. The framework considers the SWI calculation within the desired region at every 5 km * 5 km grid. The approach is found applicable for selected rainfall-induced landslides and debris flows in the Himalayas, especially in Uttarakhand, India. This framework may be useful for establishing complex rainfall threshold-based Te-LEWS in new geological settings.