Water Balance Estimation Model: Field Test and Sensitivity Analysis

Abstract

AbstractA simplified, capacity‐based simulation model has been developed to predict components of the field water balance. This requires only daily information such as the atmosphere evaporative demand and soil‐wetting events, maximum available soil water storage, and root growth characteristics, as well as crop and soil parameters. Actual evapotranspiration was modeled using Eagleman's parameterization. The model was tested for upland rice grown on a 3500‐m2 field plot, and on an undisturbed lysimeter. The predicted actual evapotranspiration (AET) and drainage (D) at the bottom of the soil profile matched the lysimeter data. Also, the model predicted reasonably well the field soil water content data, which were obtained at 15 sites. In the latter case, the root mean square error was about 16% for the seasonal AET, and 19% for D. Errors were within the limits of experimental uncertainties expected when such measurements are analyzed by Darcy's law. The results of a sensitivity analysis with respect to changes in parameter inputs show that the model is physically realistic. However, the sensitivity depends strongly on the supply of water in the soil. This clearly justifies the need to simulate the root‐zone water balance as a time‐dependent process, especially if there is a practical interest in accurate prediction during drought periods.