Modeling soil water balance and irrigation strategies in a flood-irrigated wheat-maize rotation system. A case in dry climate, China.

• Hydrus-1D model provides an alternative approach for estimating deep percolation in a flood-irrigated field. • Penman-Monteith formula is an important parameter in coupled Hydrus-1D and CROPWAT. • Hydrus-1D in combination with CROPWAT can optimum amount of water required to meet the crop water demand in the field. • CROPWAT model can improve irrigation scheduling in a flood-irrigated field under flood irrigation conditions in dry climates. Demand for irrigation water has been steadily increasing in arid regions where intensification of crop production is supported by flood-irrigation. An assessment of irrigation performance, water productivity, and irrigation schedules is critical to water conservation in dry climates. In this study, we conducted a field experiment to compare soil water balance in a flood-irrigated wheat-maize rotation system during the growing season of 2015–2016. We then modeled the soil water balance and improved irrigation strategies by coupling Hydrus-1D with the CROPWAT model, and using evapotranspiration calculated from climatic data. The calibrated Hydrus-1D model was used to simulate the temporal and spatial variation of evapotranspiration and deep percolation based on measured soil water distribution of soil profiles in the unsaturated zone. Results showed that using soil hydraulic parameters from inversion modeling can simulate soil water flow in multi-layer soil during a crop growing season. Simulation results indicated that about 36.6 and 40.6% (478.6 and 680.1 mm) of total water input in 2015 and 2016, respectively, was consumed by evapotranspiration. Furthermore, simulated deep percolation amounted for approximately 32.3 and 42.9% (403.9 and 696.6 mm) of the total amount of irrigation water in 2015 and 2016, respectively. These results indicated that only a small proportion of irrigation water was used by crops for transpiration. In addition, irrigation performance indicators such as relative water supply, relative irrigation supply, depleted fraction, and overall consumed ratio values indicated poor performance of irrigation practices in the study area. Particularly, crop yields did not increase with increases in irrigation water in flood-irrigated fields during the last ten years. Results of the CROPWAT model indicated that combinations of a fixed irrigation depth and timing, of 40 mm every 10 days, and 50 mm every 10 days were reasonable for wheat and maize, respectively, given the sandy soil in the study area. The improved irrigation strategies will limit water irrigation loss to 20% without significant effects on crop yields. This study provides an alternative approach for estimating deep percolation and crop water requirements supporting management efforts in water conservation in dry climates. [ABSTRACT FROM AUTHOR]