Modeling water table contribution to crop evapotranspiration

Abstract

A model was developed to account for the time-dependent contribution of the water table to crop evapotranspiration. The same numerical approximation used to solve the water flow in the unsaturated zone was also modified for saturated conditions. For unsaturated flow, the hydraulic conductivity changes with water content and the specific water capacity has finite values. For saturated flow, hydraulic conductivity is constant, and the specific water capacity is zero. The proposed approach considers saturated flow as a special case of unsaturated flow with a constant saturated water content and very small but not zero specific water capacities. Thus flow can be simulated in either unsaturated or saturated zones. The contribution of upward flow to crop evapotranspiration was evaluated during lysimeter experiments in the greenhouse. Spring wheat was planted on asilty clay loam and a fine sandy loam with either no water table or constant water table depths at 50, 100 or 150 cm. Irrigation was applied whenever soil water was depleted below about 50% plant available water. Model predictions of water content and cumulative upward flux as a function of time, for the different water table depths and soils, agreed closely with measured values. The contribution of the water table to evapotranspiration (ET) was found to be 90, 41 and 7% for 50, 100, and 150 cm water table depths respectively for the silty clay loam. Corresponding computed values were 89, 45 and 6%. For the fine sandy loam measured contribution of the water table to ET was 92, 31, and 9% for 50, 100 and 150 cm water tables respectively. Corresponding computed values were 99, 29, and 11 %. It was not practical to simulate the saturated-unsaturated (moving water table) predictions of the model under greenhouse conditions because of the height of the lysimeters needed. Therefore the model was also used to simulate field irrigation management options under several bottom boundary conditions where the water table contributions were significant to crop water use. Results from a one-year simulation were consistent with data for sugarcance grown under similar conditions in the Cauca Valley of Colombia.