Coupling crop growth with ParFlow for simulating agro-hydrological processes on a regional scale

Accurate representation of the coupled interactions between soil water content (SWC), groundwater dynamics, and crop growth is crucial for assessing the impacts of changing environments at the basin scale. Many basin-scale hydrological models often prioritize hydrological processes while overlooking the influence of crop growth on water dynamics. In this study, the EPIC (Erosion-Productivity Impact Calculator) model was coupled with ParFlow (referred to as ParFlow-AGE, ParFlow with Agricultural Environment) to simulate crop growth along with surface water, soil water, and groundwater at the regional scale. The model was calibrated and validated using three years of field data from a typical research area in the Hetao Irrigation District (HID). All the simulated values showed good agreement with the measured data. The simulated SWC demonstrated R2 ranging from 0.65 to 0.71, while the simulated crop height (CH), leaf area index (LAI), and aboveground biomass showed R2 values ranging from 0.90 to 0.93. Additionally, the simulated groundwater depth (GWD) exhibited relatively high accuracy, with R2 ranging from 0.76 to 0.83. During soil water exchange, in the vertical direction, irrigation and rainfall events caused pronounced fluctuations in surface soil water content, while deeper SWC remained relatively stable. In the lateral direction, quantification of water exchange of experimental field A and B using normalized radial water exchange (NRWE) showed that lateral water movement increased with depth (NRWE0-1 < NRWE1-2.6 < NRWE2.6-5). GWD, evapotranspiration, and yield exhibited high spatial variability across the study area. Shallower groundwater depth was obtained near the canal due to lateral seepage, resulting in a significantly higher actual evapotranspiration than other fields. However, the yields of maize and sunflower were 12% and 9% lower than those in other fields, respectively, because of increased stress to crops from water and aeration. Non-productive use of water, including soil evaporation from agricultural fields and evaporation from canal and built-up land, accounted for 35% of the total ETact (135,401 m3 water across 45.57 hm2). These findings validate the feasibility of the coupled model, providing a more practical and powerful tool for large-scale agro-hydrological studies.