A Critical Realization Identification Method Based on Pareto Dominance Analysis

Stochastic optimization has been frequently applied to groundwater management problems that involve uncertainties associated with hydrogeologic parameters. In this approach, a series of realizations are used to identify the optimal solution. However, this approach can be computationally intensive, especially for field-scale problems. To alleviate the computational burden associated with these types of stochastic optimization problems, a quantitative procedure identified as Pareto Dominance based Critical Realization Identification (PDCRI) method is developed to screen out the realizations that are most critical to the optimization results prior to the initiation of the optimization calculations. The criticalness of a hydrogeologic parameter realization on the solution is described by its Criticalness Index (CI) value, which is calculated based on a Pareto Dominance analysis. The PDCRI was applied to a hypothetical example to demonstrate its performance. The results show that the PDCRI approach can significantly improve computational efficiency without loss of reliability in optimal solutions and can also be used in other stochastic optimization problems such as optimal remediation system design or groundwater quality monitoring system design under parameter uncertainties.