Asymmetric Warming and Intensified Extremes Dominate Ugandan Crop Yield Variability (1995–2024): The Primacy of Nocturnal Heat over Precipitation

Abstract

This study investigates spatiotemporal trends in temperature and precipitation extremes across Uganda (1995–2024) and their impacts on national crop yields, using ERA5-Land reanalysis and FAOSTAT data. Eight core ETCCDI indices were calculated at 0.25° resolution. Monotonic trends were assessed with the Mann-Kendall test and Sen’s slope. Crop–climate relationships for 29 major crops were evaluated via Spearman’s rank correlation with Benjamini-Hochberg false discovery rate (FDR) correction. Uganda experienced rapid asymmetric warming, with national mean temperature rising by 0.90 °C since 1995 and minimum temperatures rising faster than maximum temperatures, leading to a widespread decline in diurnal temperature range. Tropical nights (TR) increased by more than 40 days across most of the country. Although annual rainfall totals showed no significant long-term change, extremes intensified: heavy rainfall events (Rx5day) strengthened in the southwest and northeast, while consecutive dry days (CDD) lengthened in central and northern Uganda, increasing the frequency of compound hot–dry conditions. After FDR correction, temperature extremes especially nocturnal warmth were the strongest predictors of yield variability, outperforming precipitation indices for most crops. Heat-sensitive staples (cassava, beans, pigeon peas) and coffee showed marked declines under elevated TNn and TR, while seed cotton, cocoa, and pepper showed modest benefits from warmer nights. Intense wet events also reduced yields, especially for groundnuts. Central and southwestern Uganda emerged as hotspots where warming, drying, and heavy rainfall intensification converge. These findings show that climate extremes are already limiting productivity in Uganda’s rainfed agriculture and highlight the need for heat- and drought-tolerant crops, expanded small-scale irrigation, agroforestry-based cooling, and stronger early-warning systems. The study also demonstrates the utility of ERA5-derived ETCCDI indices in data-sparse regions and offers practical evidence to support climate-smart agricultural planning in East Africa.