Hybrid AI/ML Enhanced Seasonal Forecasting for Agricultural Productivity and Insurance

Hybrid AI/ML enhanced downscaling of seasonal forecasts for agricultural productivity and insurance risk management

Objective

• By extending the technical framework to include climate change scenarios through Pseudo-Global Warming (PGW) experiments or dynamical downscaling of Coupled Model Intercomparison Project Phase 6 (CMIP6) global climate models?projections up to the year 2100 can be generated at kilometre-scale resolution

Description

• Seasonal and sub-seasonal climate forecasts are fundamental for agricultural planning, irrigation scheduling, water resource management, and crop yield prediction. Despite these strengths, NWP simulations are computationally intensive and their forecast skill may degrade at extended lead times, especially beyond 10?15 days. To overcome these challenges, this project proposes a hybrid downscaling approach that integrates high-resolution WRF simulations with Artificial Intelligence and Machine Learning (AI/ML) techniques

Impact

• High-resolution (? 3 km) rainfall, temperature and other meteorological forecasts with significantly improved skill over original global model outputs., Seasonal-scale forecasts integrated into crop models for yield estimation., High resolution yield anomaly maps for crop insurance companies., A replicable hybrid downscaling framework for nationwide agricultural forecasting.