A conceptual agricultural water productivity model considering under field capacity soil water redistribution applicable for arid and semi-arid areas with deep groundwater

Agricultural water productivity (AWP) model is an essential tool for irrigation water management that is highly dependent on soil water processes. Soil hydrological models based on numeric solution to the Richards’ equation are time-consuming and difficult to measure, and models based on soil water balance approach are favored especially for crop water simulation because of the less parameters requirement and higher operational efficiency. In most of the soil water balance models such as Williams-Ritchie water balance model, AquaCrop model and Hydrobal model, the under field-capacity redistribution (the redistribution during the period of soil water content below the field-capacity) is omitted and this treatment does not adequately simulate AWP for arid and semi-arid areas with deep groundwater. In these areas, AWP is the ratio between crop yield achieved and the sum of actual evapotranspiration and deep percolation at field scale. Since no more water supply for crop growth except for low frequency irrigation and tiny amount of precipitation, high evapotranspiration will aggravate an upward flow that can enhance transpiration and thus benefit crop growth while deep percolation not available for crop is sustainably accumulated to a considerable volume in under field-capacity redistribution process. To take into consideration the beneficial effects of upward flow on crop growth and the considerable under field-capacity deep percolation loss, a conceptual soil hydrological model considering under field-capacity redistribution (CSHMUR) is developed and coupled with the EPIC crop growth model. In CSHMUR model, soil water redistribution is characterized by two sequential water flows: downward flows affected by the gradient of gravitational potential and upward flows affected by the gradient of matric potential. These two flows are mainly used to simulate deep percolation occurring in redistribution processes and upward flows resulting from matric potential, respectively. The CSHMUR-EPIC model is calibrated and validated with field data for a typical arid area of northwestern China, and it is then applied for the simulation of seven irrigation scenarios. The study highlights that the upward flows aggravated by drought conditions and the under field-capacity deep percolation are remarkable enough and should not be neglected in the AWP estimation for arid and semi-arid areas with deep groundwater. The developed CSHMUR-EPIC model can effectively simulate upward flows and the under field-capacity deep percolation, and thus soil water content (SWC) both in lower and upper soil profiles, actual evapotranspiration and crop growth, resulting in an precise estimation of AWP. As upward flows and the under field-capacity deep percolation vary with irrigation schedule, the model is also helpful in exploring various irrigation schedule to obtain a sustainable agricultural water resources management.