Modelling soil water dynamics of full and deficit drip irrigated maize cultivated under a rain shelter.

The model HYDRUS-1D was used to simulate soil water dynamics of full and deficit irrigated maize grown under a rainout shelter during two crop seasons. Four irrigation treatments were established based on the amount of water applied to fulfil crop water requirements. Treatment D1 was irrigated to fully satisfy crop water requirements, while treatments D2 (mild deficit), D3 (moderate deficit), and D4 (severe deficit) were for increased controlled water stress conditions. The computation and partitioning of evapotranspiration data into soil evaporation and crop transpiration was carried out with the SIMDualKc model, and then used with HYDRUS-1D. The soil hydraulic properties were determined from numerical inversion of field water content data. The compensated root water uptake mechanism was used to describe water removal by plants. The HYDRUS-1D model successfully simulated the temporal variability of soil water dynamics in treatments irrigated with full and deficit irrigation, producing RMSE values that varied between 0.014 and 0.025 cm3 cm−3 when comparing model simulations with field measurements. Actual transpiration varied between 224 and 483 mm. Potential transpiration reductions varied from 0.4 to 48.8% due to water stress, but plants were able to compensate for the water deficits in the surface layers by removing more water from the deeper, less stressed layers. HYDRUS-1D water balance estimates were also comparable with the corresponding ones determined with the SIMDualKc water balance model. Both modelling approaches should contribute to improve the web-based IRRIGA system, used to support farm irrigation scheduling in Brazil.