Quantifying crop water footprint and spatial heterogeneity using data assimilation

Water scarcity has long constrained agricultural sustainability in the Huang-Huai-Hai Plain, a vital grain production base in China. Regional water resources can also be regulated to improve water use efficiency in sustainable agriculture. It is often required to precisely assess agricultural water use efficiency. Crop production water footprint can be expected to measure the sustainability and efficiency of water resource utilization during the entire crop growth cycle. This study selected winter wheat as the research subject. Assimilated variables were utilized as remotely sensed leaf area index (LAI) and soil moisture (SM). A quantitative assessment was also developed for winter wheat water footprint, according to dual-variable assimilation of crop models and remote sensing data. Spatial dependency and clustering of winter wheat water footprint were then determined using spatial autocorrelation analysis. Furthermore, winter wheat yield–total water footprint quadrant classification, blue and green water resource dependency, and groundwater extraction proportion were integrated to clarify regional water source dependence and formulate differentiated water footprint management strategies. The results indicated that: 1) Data assimilation significantly improved the accuracy of the WOFOST model to simulate the winter wheat yield. There was strong consistency between the simulation and the statistical yield after data assimilation, with an R² increased to 0.98 and an RMSE reduced to 67.68 kg/hm². The accuracy significantly also improved after simulation, compared with an R² of 0.42 and an RMSE of 566.78 kg/hm²; 2) The average green, blue, and total water footprint of winter wheat were 0.35, 0.30, and 0.65 m³/kg, respectively, after data assimilation. The green and the total water footprint exhibited a spatial distribution pattern higher in the south and lower in the north, while the blue water footprint showed a pattern higher in the north and lower in the south; 3) Spatial autocorrelation of winter wheat green and blue water footprint was stronger than that of the total water footprint. The blue, green, and total water footprint of winter wheat exhibited significant spatial clustering, primarily characterized by high-high and low-low clustering; 4) The northern region should prioritize stable production, water saving regulation, and reduction of groundwater extraction, whereas the southern region should focus on improving precipitation use efficiency. This finding can provide scientific support and decision-making basis for the refined and differentiated water resource strategies in typical water-scarce agricultural regions, such as the Huang-Huai-Hai Plain. A solid theoretical foundation and technical framework can help allocate agricultural water resources at the regional scale.