Addressing global hotspots of drought-related crop production losses.

Meeting future food demand requires transforming food systems to simultaneously increase production, reduce environmental impacts, and adapt to climate change. As climate variability increasingly affects production stability, understanding where cropping patterns are vulnerable to hydroclimatic stress is a missing but critical step toward improving agricultural production resilience. Here we combine gridded climate data, spatially explicit agricultural statistics, and empirical water-production functions to quantify global patterns of rainfed and irrigated crop-specific Drought Sensitivity-defined as the percent reduction in median yield under extreme hydroclimatic conditions-and drought-associated losses for 17 major crops, representing 75% of global production. This metric identifies locations where crops experience high climate variability and are most susceptible to drought-related losses. We estimate global losses of  - 10.1% and  - 6.8% in median rainfed and irrigated production, respectively, under historically observed extremes-enough calories to feed 2.1 billion people-and identify hotspots in the central US, eastern Brazil, the Mediterranean, and South Asia. Focusing on monsoon cereals (rice, maize, millet, sorghum), we show that sustainable irrigation expansion and targeted crop switching could avoid 62% of rainfed losses while increasing median production by 14%. This scalable framework enables proactive targeting of mitigation actions and investments to stabilize and increase global crop supply.