Abstract Extreme weather events are major sources of crop yield variability and increasingly contribute to severe yield losses, threatening livelihoods, sustainability and food security in Sub-Saharan Africa (SSA). While such events are projected to intensify and become more frequent, gaps remain in understanding which agrometeorological hazards are most strongly associated with yield anomalies and crop failures. Here, we use data from 1981–2016 to assess statistical associations between different hazard indices and yields of five major cereals (maize, millet, sorghum, rice and wheat) across SSA, combining national, subnational and gridded datasets. Drought- and heat-related indicators emerge as the dominant correlates of negative yield anomalies, with vapor pressure deficit (VPD) showing a consistent negative association across crops and scales. Yield anomalies in years classified as crop failures are more strongly linked to these hazards than overall interannual variability, indicating that severe yield losses are especially sensitive to agrometeorological extremes. Using interpretable machine-learning methods for maize and rice, two key staple cereals in SSA, we show that failure-year yield anomalies in maize are more strongly associated with drought-related indicators, whereas rice failures are more sensitive to temperature-related factors alongside drought metrics. Projections using combined Earth system and ensemble output of crop models reveal that under the high-emissions SSP585 scenario, both maize and rice-planting areas will face significant increases in heat and drought hazards by mid-century. Maize-growing regions are projected to experience a 1.8 °C rise in maximum temperature and a 0.14 kPa increase in VPD, while rice-growing areas may see a 1.5 °C rise in maximum temperature, a 15% increase in days above 35 °C, and a 0.13 kPa rise in VPD. These projected changes underscore the need for targeted risk management, improved agronomic practices and robust early warning systems to enhance the resilience of SSA’s agricultural sector in a warming climate.
Wang et al. (Tue,) studied this question.