Comparative analysis of rainfall thresholds for landslide pre- and post- the 2011 Sikkim earthquake at regional scale

While rainfall thresholds for landslide early warning and the exacerbating effect of post-earthquake rainfall on landslides are well documented, understanding the changes in rainfall thresholds for post-earthquake landslides and their application to regional-scale early warning systems remains challenging. This study proposes an integrated framework combining statistical and mechanistic modeling to quantify rainfall threshold variations pre- and post-earthquake, using historical landslide data, geological maps, and seismic records from Yadong County, Xizang, China. We first derived a 3-day cumulative rainfall threshold curve for pre-earthquake landslide. The Newmark model was subsequently applied regionally to back-analyze landslides triggered by the 2011 Sikkim earthquake, establishing post-seismic rainfall thresholds validated by a cumulative displacement distribution map (AUC = 0.81). Results demonstrate a 31.7% average reduction in post-earthquake cumulative rainfall thresholds, highlighting a more pronounced reduction in rainfall requirements for landslide initiation following seismic events. This study underscores the necessity of dynamic adjustments to rainfall thresholds considering seismic disturbances when devising early warning systems for landslide risk management across active tectonic regions in flood seasons.