Search

Heuristic search techniques are highly flexible but computationally intensive optimization methods that require hundreds, sometimes thousands, of evaluations of the objective function to reach termination criteria in common water resources optimization applications. One way to make these techniques more tractable when the objective function depends on a time-consuming flow and transport model is to employ an empirical approximation of the model. The current study examines the impact of employing artificial neural networks (ANNs) and linear approximators (LAs) on the quality and quantity of solutions obtained from simulated annealing-driven searches on two different ground-water remediation problems. The quality of results obtained when ANNs served as substitutes for the full model was consistently comparable to that of results obtained when the full model itself was called in the course of the search. The effect on quality of results of substituting an LA for the full model was more variable.

A methodology for selecting sampling frequencies in environmental research that focuses on statistical thinking and brings considerable cost savings is described. The aim is to reduce sampling costs without jeopardizing information needed for decision making. It was applied to a 1992 benchmark volatile organic compound data set from 296 wells at the main site of the Lawrence Livermore National Laboratory, while a variant was used on data from groundwater monitoring wells at the Savannah River site. An overall schedule reduction of 30% was achieved in the required number of samples per year, while annual cost savings amounted to about $1.8 million.

The high costs of environmental cleanup have prodded scientists to develop approaches that offer cost-effective alternatives. In the study, a nonlinear optimization model using artificial neural networks and genetic algorithm were used in creating contaminant transport models that can be applied in groundwater management initiatives. Application of the model in the cleanup of a Superfund site revealed efficient groundwater transport prediction. This suggests that the approach offers wide potentials for reducing cleanup costs.

Results indicate that when one characteristic was present and more severe than the others, predictable placement recommendations were made. In more complex combinations, however, no general decision-rules could be found to explain the recommendations. (Author)