Quantification of wheat water footprint based on data assimilation of remote sensing and WOFOST model

Water scarcity in agricultural production has emerged as a significant constraint to food security in China. To improve agricultural output and ensure its stability, it is imperative to assess the agricultural water use efficiency. The crop water footprint (WF) is an effective tool to assess the type, amount, and efficiency of agricultural water usage. However, existing quantitative studies on the crop water footprint within regions exhibit limitations in terms of their low spatial resolution and deficient spatial heterogeneity. In this study, the World Food Studies (WOFOST) model was combined with the Ensemble Kalman filter (EnKF) assimilation algorithm, with the leaf area index (LAI) and soil moisture (SM) obtained by remote sensing as state variables. An improved method for quantifying the WF of wheat based on crop model -remote sensing information data assimilation was proposed. Regarding the results, the average blue water footprint (WFblue), green water footprint (WFgreen), and total water footprint (WF) of wheat were 363 m3/t, 551 m3/t, and 914 m3/t, respectively. In the southern of the study area, WFgreen exhibits higher values, with lower levels in the north. WFblue and WF demonstrated inverse spatial patterns, with higher values observed in the northern areas and lower values in the southern regions. The spatial heterogeneity of WFblue and WF was more significant than that of WFgreen. The WF of wheat quantified by the data assimilation method had a high spatial resolution, and could be effectively used to explore the spatial heterogeneity of WF. This study can provide a reliable reference for the effective use of water resources.