Compound agricultural drought and heat events (CADHEs) pose an escalating threat to global food security, yet current assessments often overlook the distinct roles of stress duration and severity in driving yield loss. To address this gap, we developed a process-based classification framework incorporating antecedent cumulative effects to decouple CADHEs into duration-dominant and severity-dominant types, and to assess their spatiotemporal dynamics and impacts on maize yield in the Huang-Huai-Hai (HHH) Plain. The results showed that drought duration-dominant events ( CADHE dd ) occurred more frequently than heat duration-dominant events ( CADHE hd ) and concurrent events ( CADHE co ), with occurrences concentrated in the V0–V6 stages across central-southern Hebei, most of Henan, and northern Shandong. Furthermore, compared with heat severity-dominant events ( CADHE hs ), drought severity-dominant events ( CADHE ds ) were more frequent, persisted longer, and exhibited higher severity and intensity. Critically, maize yield responses to compound events exhibited strong phenological dependence and clear asymmetry between duration-dominant and severity-dominant stress. Duration-dominant CADHEs, especially CADHE dd , acted as chronic constraints during the vegetative-to-reproductive transition, and were governed by a cumulative deficit mechanism, whereby yield loss was most sensitive to prolonged exposure, as evidenced by a decline rate of 2.63% d -1 ( P <0.01). In contrast, severity-dominant CADHEs, particularly CADHE hs , functioned as acute stressors during the reproductive phase, acting as severity-dependent amplifiers with a significantly steeper marginal sensitivity (-6.39%, P <0.01) than CADHE ds (-4.88%, P <0.01). The R1–R3 stage emerged as the most vulnerable window, where the average yield loss rate exceeded 15% under heat severity-dominant conditions ( P <0.01). These findings indicate that while drought imposes widespread cumulative penalties, heat severity acts as a critical tipping point for yield collapse. Sustainable agricultural risk management should therefore shift from static protocols to driver-specific strategies, prioritizing buffering measures against chronic drought and avoidance strategies against acute heat stress.
Li et al. (Fri,) studied this question.