A Modified Constrained State Formulation of Stochastic Soil Moisture for Crop Water Allocation

Abstract

In response to uncertainty in crop water allocation, several methodologies have been proposed in the literature, most of them considering rainfall as a stochastic variable affecting soil moisture. A methodology considering uncertainties both in irrigation depth and soil moisture is more realistic for irrigated crops as developed here using an explicit stochastic optimization model. This new work is based on an earlier constrained state formulation which did not consider the irrigation depth as stochastic. In constrained state formulation methods, the first and second moments of state variables are developed considering the uncertainties which are then used as constraints in an optimization model. In contrast to alternative methods that are dynamic programming-based, the proposed optimization method can be solved using standard nonlinear optimization tools. Performance of the proposed model is evaluated for the case of two different crops, winter wheat and barley. Model verification is performed by comparing the results with simulation results. The model is quite acceptable and shows considerable improvement over analogous models.