FOR MOTHER EARTHAmrita landslide research
Site investigations · mega disasters
Amrita teams go where landslides and cascading disasters strike, and bring back the evidence that shapes the next warning.
Amrita’s response combined GPM IMERG satellite rainfall, KSDMA/IMD rain-gauge data, ERT investigation of the water table and SRT bedrock surveys. Trigger analysis applied Caine (1980) thresholds with regional rainfall accumulation models.
The lesson: treating each hazard in a separate silo would have missed the chain.
Intense rainfall triggered an ice avalanche that breached the terminal moraine. The flash flood ran down the Teesta River and destroyed the Chungthang (Teesta III) Dam.
Building on a 2013 Current Science study by Dr Remya S. N., researchers analysed Landsat 5/7 imagery from the 1990s to 2022 and CORONA data. Graph-based hazard-network analysis confirmed dam collapse and downstream flooding as the highest-risk nodes (Ekkirala & Ramesh, Scientific Reports, 2026).
Multi-directional cracking, road subsidence and barrage damage were documented across the three sites.
The findings inform future Himalayan early warning deployments and dynamic risk maps that combine geological, hydrological and human stress factors.
By analysing sub-basins, drainage patterns and geology, researchers identified high-risk zones, especially sub-basin G. The study shows how geology and human activity combine, and offers a systematic framework for hazard assessment and mitigation planning.
Read in Publications →
The research goal is an AI-driven early warning system that integrates real-time and historical rainfall, soil and terrain analysis, and machine-learning prediction, targeting the coupling between cyclones, rainfall and slope stability on India’s east coast.
With the Karnataka State Natural Disaster Monitoring Centre, researchers developed and rigorously tested rainfall-threshold-driven dynamic landslide maps. They predict and monitor real-time landslide occurrence across the Kodagu region, a markedly more precise approach to forecasting.