Climate Change Impacts on Global Crop Yield Variability and Food Security

Abstract

Climate change is projected to alter both mean crop yields and year-to-year yield variability, with the latter carrying
particular consequences for food security in regions with limited storage and trade capacity. We combined a
five-member CMIP6 climate model ensemble with an ensemble of four process-based global gridded crop models
to project yield outcomes for maize, wheat, rice, and soybean under SSP2-4.5 and SSP5-8.5 scenarios at
mid-century (2050) and end-of-century (2090), relative to a 1995-2014 baseline, and evaluated two adaptation
scenarios, shifted planting dates with cultivar thermal-time adjustment, and expanded irrigation, against a
no-adaptation reference. Without adaptation, global mean yields declined under SSP5-8.5 end-of-century by 12.4
percent for maize, 8.1 percent for soybean, and 5.6 percent for rice, while wheat showed a small gain of 1.9
percent, driven by higher-latitude gains outweighing lower-latitude losses. Interannual yield variability, measured as
the coefficient of variation across simulated years, increased for all four crops, most sharply for maize, up 34
percent relative to baseline, and least for wheat, up 9 percent. Combined cultivar and planting-date adaptation
recovered roughly half of projected maize and soybean losses; irrigation expansion gave additional but
geographically concentrated benefit, largest in water-limited cropping areas of South Asia and sub-Saharan Africa.
Regional vulnerability, defined jointly by yield decline and variability increase, was greatest in sub-Saharan Africa
and South Asia, where limited adaptive capacity and dependence on rain-fed staple production compound climate
exposure. These results indicate that yield variability, not only mean yield change, is a first-order and
under-addressed dimension of climate risk to food security, and that regionally targeted adaptation can
meaningfully but incompletely offset projected losses.