Hydrological drivers of maize productivity: A new analytical framework

Songyuan City, a semi-arid region in Northeast China and one of the country's major maize-producing cities. This study developed the Crop-Oriented Hydrological Variables Zoning and Quantification Framework (COHV-ZQ), integrating Multiscale Geographically Weighted Regression (MGWR) and clustering analysis to spatially delineate hydrological-crop zones. This framework identifies key hydrological drivers affecting maize growth by quantifying the spatial relationships between hydrological variables (precipitation, evapotranspiration, groundwater, and soil moisture) and net primary productivity (NPP), while enhancing maize yield through optimized water allocation based on stage-specific requirements. Evapotranspiration emerged as the dominant direct driver of maize NPP, while precipitation exerted only indirect effects via soil water replenishment and groundwater recharge (R² > 0.7, p < 0.05). Increases in monthly evapotranspiration (ET) from May to September resulted in heterogeneous enhancements of NPP across space and time: a 1 mm rise in evapotranspiration contributed an additional 2.1–15 g C m⁻² to NPP, depending on the growth stage. No significant correlation was observed between changes in groundwater storage/soil moisture and an increase in maize yield. These findings emphasize the need for localized, temporally adaptive water management strategies in semi-arid agroecosystems to optimize agricultural productivity under resource constraints. • Novel MGWR-clustering framework enables hydrological-crop zoning. • Quantifies maize growth-hydrology linkages across spatiotemporal scales. • Evapotranspiration drives maize yield (2.1–15 g C/m² per mm).