Evaluating remotely sensed climate products for monitoring ENSO-driven concurrent heat and water-deficit stress in rainfed maize–soybean systems of Southern Africa

Sparse and declining ground-based meteorological networks constrain the monitoring of weather conditions relevant to rainfed maize–soybean systems across Southern Africa. Although gridded climate products offer a scalable alternative, their ability to capture the co-occurrence of heat and water deficit stress remains poorly evaluated. We address this gap by (i) evaluating four daily rainfall and four maximum air temperature (Tmax) products against observations from four Zambian stations (2013/14–2025/26) using conventional and confusion-matrix statistics; (ii) applying the best-performing products to characterise concurrent stress at six major production sites across Malawi, Mozambique and Zambia over 41 seasons (1985/86–2025/26); and (iii) examining associations with ENSO phases and documented droughts. Concurrent stress was defined as hot-dry days (Tmax > 30 °C and rainfall < 1 mm day−1) and hot-dry spells (≥ 3 consecutive hot-dry days). CHIRPS provided the most accurate rainfall estimates (monthly R² up to 0.89; dry-day F1 71.4–83.8 %), while MERRA-2 best identified heat-stress days (F1 up to 82.7 %). The CHIRPS–MERRA-2 combination revealed higher stress frequencies at Zambian sites than at Malawian and Mozambican sites. El Niño seasons were associated with significantly greater concurrent stress (p< 0.001). Stress probability peaked in November (0.37–0.67), highlighting vulnerability during crop establishment. Validated gridded products can therefore support climate-informed monitoring and decision-making in Southern African rainfed systems.