Water Resources Research
Irrigation strategies which maximize crop yield while maintaining target salt concentration in the root zone and/or prevent salt from leaching to the groundwater are computed using a nonlinear, one dimensional, simulation/optimization management model. The included constraint equations maintain a water volume balance and salt transport in the unsaturated zone. Utilized are implicit finite difference forms of the nonlinear, unsteady, unsaturated water flow equation (Richards's equation), and the diffusion-convection solute transport equation. Other constraints include nonlinear functions describing the hydraulic properties of the medium (hydraulic conductivity as a function of matric potential, volumetric water content as a function of matric potential, and a root extraction term). The model solves approximately 8,000 equations simultaneously in time and space. To reduce computer memory and processing time, relatively large time steps are used. To prevent, or correct the inaccuracy n9rmally caused by coarse discretization, an approach of the predicted-corrector method is adopted. This requires partitioning the model into four modules, A, B, c, and D. Optimization is performed in module D. Modules B and c calculate calibration coefficients for use in module D. As a result, intercell water and mass flux ratio in module D have the same accuracy as a more finely discretized simulation model. The model calculates the optimal irrigation amount for a predefined irrigation frequency, for the selected management scenario. In that process it computes a detailed soil water profile and salt distribution, and assures that spatially variable moisture or concentration constraints are satisfied. The model also computes the trade-offs between ground water quality protection and crop production.
Musharrafieh, G.R., Peralta, R.C., Hanks, R.J. and L.M. Dudley. 1995. Optimizing irrigation management for pollution control and sustainable crop yield. Water Resources Research. 31(4): 1077-1086.