Modeling soil water-salt dynamics and maize yield responses to groundwater depths and irrigations

Abstract: Reduction of water diversion from the Yellow River will intensify water shortage problems in the Yinbei Irrigation District (YID). Reasonable use of groundwater for irrigation is helpful to maintain the agricultural production. Groundwater exploitation may cause groundwater level declines in local areas. This helps to reduce the salinity accumulation in the root zone but decreases the capillary rise. Thus, it is important to figure out the responses of soil water-salt dynamics and crop yields to groundwater table fluctuations for salinity control and stable yields. In this study, HYDRUS-1D model was modified by coupling with the EPIC (erosion-productivity impact calculator) crop growth module for simulating agro-hydrological processes. The new crop module could simulate crop height, leaf area index (LAI), above-ground biomass and crop yield. The information between HYDRUS-1D and EPIC was exchanged by daily step. Root water uptake under water and salt stress was calculated with HYDRUS-1D and imported to EPIC to limit crop growth. EPIC module estimated crop height, LAI and root depth for HYDRUS-1D to calculate soil water-solute dynamics. HYDRUS-1D assumed that soil evaporation remained at the potential rate unless pressure head of the soil surface decreased to a prescribed value. After then this prescribed value was set as a constant head to renew the top boundary condition. However, it cannot reasonably reflect the decrease stage of soil evaporation when using the constant head boundary. This may overestimate soil evaporation. Therefore, a new soil evaporation module, estimating soil evaporation reduction coefficient using soil water content of the top layer (0-10 cm), was added for better describing the soil evaporation under shallow water tables. With the experimental data collected from the maize field in 2008, the model was calibrated by the data of groundwater irrigation treatment and validated by the data of canal irrigation treatment. Simulated soil water content and solute concentration in the root zone (0-90 cm) showed good agreement with the measured values. Root mean square error (RMSE), mean relative error (MRE) and coefficient of determination for soil moisture were 0.03 cm3/cm3, 3.4% and 0.78, respectively. For solute concentration, RMSE, MRE, coefficient of determination were 1.6 g/L, 1.3% and 0.29, respectively. LAI and above-ground biomass values were fitted well with the observations. MRE values for estimated and measured LAI and above-ground biomass were 5.9% and 10.6%, and R2 were both larger than 0.95 for these two items. The model was then used to assess the impacts of groundwater table and irrigation changes on soil water-salt dynamics and maize yields. Nine groundwater depth (GWD) scenarios (100, 110, 125, 140, 155, 170, 185, 200 and 250 cm) and 6 irrigation treatments (0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 times of the present irrigation) were considered. The results showed that soil water content and salt storage in the root zone declined with the reduction of groundwater level and irrigation amount. Due to the decrease of groundwater contribution and soil moisture, lowering groundwater depth resulted in a gradual increase of the average solute concentration in the root zone. Maize yields increased first and then decreased as the groundwater table declined. Generally, the maximum yields were achieved when GWD was between 140 and 155 cm. The maize yields may decrease with reducing the irrigation amount, therefore water-saving strategies were not recommended for local farmers with low incomes. Finally, the optimum groundwater depth of 140-155 cm was suggested, and three irrigations with an amount of 900 m3/hm2 for each will be applied during maize growing period.