The Groundwater and the Groundwater Quality Management Problem: Reliability and Solution Techniques

Abstract

The groundwater and the groundwater quality management models are defined as an engineering decision problem. The solution of this problem is not trivial and can be reached with the use of deterministic or stochastic techniques. The determistic approaches (Primal Method) compute only local minima, or the global minimum (Outer Approximation Methods) only according to strict hypothesis. The stochastic methods (Simulated Annealing, Genetic Algorithms or Neural Networks) find the global minimum, but only in statistical sense. The most recent approaches to the problem treat the physical parameters (transmissivity) as a random field. The uncertainty of the transmissivity makes impossible to guarantee the feasibility of the optimal strategy. A reliability level is defined as the ratio between the number of realizations that mantain feasible the optimal solution out of the total number of realizations. Some of the algorithms proposed to reach a fixed reliability are discussed. The idea of linking the goal of the management problem with the number and location of new field measurements is introduced.