A process-oriented framework for daily dynamic monitoring of summer maize drought: From meteorological stress to agricultural impact

CONTEXT The progression of crop drought from meteorological deficit to agricultural damage is a continuous, stage-specific process. However, prevailing drought monitoring approaches, including single-type indices (e.g., SPI) and integrated composite indices (e.g., CDI), either fail to capture this continuum or obscure critical phase transitions. Consequently, an operational framework capable of dynamically tracking the complete, phase-resolved drought process at high temporal resolution remains a critical gap. This gap fundamentally hinders the systematic characterization, continuous tracking, and a holistic understanding of drought dynamics from meteorological stress to agricultural impact. OBJECTIVE This study aimed to bridge this gap by developing and validating a process-oriented, phase-decoupled monitoring framework for summer maize drought. The primary objectives were to: (1) formally define the drought process through three nodes (stress initiation, visible damage occurrence, recovery completion) and two phases (latency, development), and (2) construct a corresponding Dynamic Monitoring of Maize Drought (DMMD) framework that strategically delegates monitoring tasks to the optimal index for each phase. METHODS The DMMD framework synergistically couples the 10-day Standardized Precipitation Index (SPI 10 ) to monitor the latency phase (identifying stress initiation), and the Vegetation Water Index (VWI) with dynamic thresholds to monitor the development phase (tracking from visible damage occurrence to recovery completion). The framework was validated using multi-source data (meteorological, remote sensing, yield) at a representative station and across the Huang-Huai-Hai (HHH) region of China. RESULTS AND CONCLUSIONS Results demonstrated that the DMMD framework can explicitly characterize the spatio-temporal evolution of maize drought. Specifically, SPI 10 can detect the drought onset 0–24 days ahead (with an average lead time of nine days) relative to the visible damage occurrence identified by VWI. Moreover, drought severity, quantified by the ratio of agricultural drought days, showed a significant negative correlation with standardized climatic yield. This correlation is particularly strong for severe drought with the highest Spearman coefficient of R = −0.7. SIGNIFICANCE This study moves beyond creating another composite index by introducing a phase-decoupled monitoring framework. By explicitly resolving the drought process into distinct stages and strategically delegating monitoring, DMMD framework provides a robust tool for dynamic, process-oriented drought tracking and early warning, offering a concrete foundation for targeted mitigation strategies to safeguard summer maize production.