Evaluating Data Worth for Ground-Water Management under Uncertainty

Abstract

A decision framework is presented for assessing the value of ground-water sampling within the context of ground-water management under uncertainty. The framework couples two optimization models—a chance-constrained ground-water management model and an integer-programing sampling network design model—to identify optimal pumping and sampling strategies. The methodology consists of four steps: (1) The optimal ground-water management strategy for the present level of model uncertainty is determined using the chance-constrained management model; (2) for a specified data collection budget, the monitoring network design model identifies, prior to data collection, the sampling strategy that will minimize model uncertainty; (3) the optimal ground-water management strategy is recalculated on the basis of the projected model uncertainty after sampling; and (4) the worth of the monitoring strategy is assessed by comparing the value of the sample information—i.e., the projected reduction in management costs—with the cost of data collection. Steps 2–4 are repeated for a series of data collection budgets, producing a suite of management/monitoring alternatives, from which the best alternative can be selected. A hypothetical example demonstrates the methodology's ability to identify the ground-water sampling strategy with greatest net economic benefit for ground-water management.