Climate change poses an existential threat to agricultural communities worldwide. Rising temperatures, shifting rainfall patterns, and increased weather variability will fundamentally alter where and how crops can be grown—with the poorest and most vulnerable communities facing the greatest burden.
This project provides spatially disaggregated estimates of changes in agricultural productivity across multiple time periods. Using agronomic models from the Global Agro-Ecological Zones (GAEZ v5) project, we quantify observed and projected productivity losses under a no-adaptation assumption: farmers are assumed to maintain the crop choices and farming practices observed in the baseline period. See the Key Findings for a summary of our main conclusions.
Our estimates highlight the places where adaptation will be most needed. Adapting to climate change will be challenging everywhere—requiring new technologies, crop switching, and, in some places, shifts in where production happens. But the capacity to make these changes is deeply unequal, leaving many smallholders and low-income regions with far fewer options. The result: food security and rural livelihoods are most at risk where resilience is lowest.
Climate-driven declines in agricultural productivity will be uneven. Current research shows that adaptation is limited even in developed countries—and in many low-income regions it is especially constrained. When yields fall, the impacts ripple outward: food insecurity rises, rural incomes shrink, and communities confront difficult choices about migration and livelihood change.
This webpage is an early warning tool for policymakers, development practitioners, and researchers. By pinpointing areas at risk of severe losses under baseline cropping patterns, it supports:
We report global changes in attainable yields—the biophysical maximum productivity a location can sustain—over 20-year periods at a resolution of roughly 10×10 km. For each grid cell, productivity is computed using a fixed crop mix—held constant across all periods and scenarios—so that only climate conditions vary. All estimates are based on the Global Agro-Ecological Zoning (GAEZ v5) model.
We distinguish between already observed change and future projections. Observed change is measured by comparing attainable yields under the 1981–2000 climate with those under the 2001–2020 climate, using historical climate inputs from a global weather data set tailored for agriculture (AgERA5). Projections are then reported relative to the 2001–2020 baseline, using IPCC-consistent climate scenarios that build on the magnitude of change already observed.
To provide a single, intuitive measure of impact, we aggregate the main crops in each cell by converting attainable production into calorie equivalents. Results are expressed as changes in the annual number of people who could be fed per grid cell, assuming a daily requirement of 2,000 calories per person.