The Kyrgyz Republic faces increasing climate risks from flash floods and mudflows caused by glacier retreat, irregular snowmelt and changing precipitation patterns. This project, supported by CTAF under GGGI, assesses the feasibility of deploying an NPU-enabled Climate Risk Intelligence System that combines earth observation data, geospatial analytics, artificial intelligence and Edge Computing technologies to strengthen climate risk forecasting and early warning capabilities. The project is the first of its kind in Kyrgyzstan and evaluates how NPU-powered AI models can automatically process multiple weather and climate datasets, compare precipitation forecasts against locally calibrated risk thresholds and generate near real-time risk alerts for selected pilot river basins. The result is a low-cost, scalable forecasting solution for data-scarce mountainous river basins, where conventional monitoring infrastructure is limited. The primary beneficiaries are communities exposed to flash floods and mudflows in the Osh and Jalal-Abad regions, who gain more accurate flood warnings and are better prepared to respond.

Climate Adaptation, Disaster Risk Reduction, Edge AI, Early Warning

The project establishes a technical and institutional roadmap for future deployment of AI-enabled early warning systems across priority river basins in Kyrgyzstan. The lessons learned can be replicated across other Central Asian countries facing similar climate-induced flood, mudflow and landslide risks, positioning Kyrgyzstan at the forefront of real-time, data-driven early warning systems in Central Asia.