Integrated optimal management of ground-water pollution and withdrawal

Abstract

An integrated ground-water management model is formulated as a multivariable constrained nonlinear optimization problem. To simulate the physical and chemical processes occurring within a leaky confined aquifer system, the finite-difference forms of the flow and transport equations are embedded in the management model. The Hooke-Jeeves method, a nonlinear programming technique, in conjunction with the exterior penalty function method is used to solve this management model. The suitability and capability of this method to solve the management problems for study areas of different sizes and different numbers of management periods are demonstrated. The performance evaluation of this proposed methodology establishes its potential applicability for the solutions of different kinds of ground-water management problems. The developed methodology also demonstrates the suitability of the embedding technique to solve a dimensionally large nonlinear ground-water management problem. The proposed methodology does not require the linking of simulation and optimization models externally. It is shown that global optimality of obtained solutions is dependent on the extensive identification of local optimal solutions and the accuracy in prescribed aquifer characteristics.