ESTIMATING ROOT-WATER-UPTAKE USING AN INVERSE METHOD

It is extremely challenging, if not impossible, to measure distribution of root-water-uptake or root density directly, either in the laboratory or in the field. To estimate root-water-uptake distribution in soils, the authors of this paper developed an inverse method, whereby measured values of water content in soil profiles with root-water-uptake were used as input information, which can be acquired relatively easily and accurately. Then a soil water flow equation incorporating an unknown sink term was solved inversely for the sink term, which characterizes the root-water-uptake distribution in the root zone. The estimated distribution of root-water-uptake rate was used to calculate the normalized distribution of relative root density, from which the root-water-uptake model was established. Based on the model, soil water flow with root-water-uptake was simulated successfully. Examples of numerical experiments were designed to examine the accuracy and stability of the proposed approach for estimating the average root-water-uptake rate, considering different time intervals between successive measurements of the input distributions of soil water content, spatial intervals of the measurement points along the soil profile, simulating depth, different soils, variation of the saturated hydraulic conductivity, and the stochastic nature of the measured soil water content. Examples were also used to demonstrate applications of the method. The results showed that the inverse method is reliable for estimation of distribution of the root-water-uptake rate, compared with the theoretical values. The proposed method is useful for estimation of the normalized distribution of the relative root density, from which the root-water-uptake model can be determined and the soil water flow with root-water-uptake can be simulated continuously.