Context
Kodagu district in Karnataka lies in the Western Ghats, one of India's most landslide-prone regions. Steep terrain, high annual rainfall, and dense settlement patterns make early warning critical. The district has experienced repeated landslide events during the monsoon season, with significant impacts on roads, settlements, and health infrastructure.
Methodology
AmritaKripa was deployed across Kodagu with a database of 79 recorded landslide events and 309 mapped points of interest including settlements, roads, and health facilities. Five rainfall models were run across six time windows and validated against the historical record. The incident tool was configured to search a 20 km radius around any reported or predicted slide location.
Findings
The platform produced hourly risk updates at Gram Panchayat level throughout the study period. The 30 model combinations were evaluated against the 79 recorded events to identify which combinations performed best for Kodagu's specific terrain and rainfall patterns. The incident tool successfully identified exposed settlements and health facilities within the search radius for each recorded event.
Outcomes
The Kodagu deployment demonstrated that panchayat-level risk mapping is operationally feasible with existing meteorological data. The exposure search capability gave officials a structured basis for evacuation planning that was not previously available. The case study is being documented jointly with Amrita Vishwa Vidyapeetham and will be presented at the 7th World Landslide Forum in Faridabad, November 2026.
Bring AmritaKripa to your district
We are working with district administrations and state disaster management authorities to expand the platform to new regions.