You are here:
Spatiotemporal Groundwater Monitoring Network Design Using Ensemble Kalman Filtering, Multiobjective Evolutionary Optimization, and Interactive VisualizationEPA Grant Number: F07A20389
Title: Spatiotemporal Groundwater Monitoring Network Design Using Ensemble Kalman Filtering, Multiobjective Evolutionary Optimization, and Interactive Visualization
Investigators: Kollat, Joshua B.
Institution: Pennsylvania State University
EPA Project Officer: Just, Theodore J.
Project Period: August 1, 2007 through August 1, 2010
RFA: STAR Graduate Fellowships (2007) RFA Text | Recipients Lists
Research Category: Engineering and Environmental Chemistry , Academic Fellowships , Fellowship - Civil/Environmental Engineering
The health risks and cleanup costs of groundwater contamination represent a significant burden to our society. This research will develop an improved framework for designing cost effective long-term groundwater monitoring networks that can be adaptively managed based on statistical assessments of contaminate fate-and-transport and changing site conditions.Approach:
This research will make significant contributions towards long-term groundwater monitoring network design by combining Ensemble Kalman Filtering, multiobjective evolutionary optimization, interactive visualization, and adaptive management strategies. The tractability of optimal monitoring network design will be significantly improved through the application of a parallel evolutionary multiobjective optimization algorithm that is capable of “learning” the structure of the network design problem. Subsequent integration of Ensemble Kalman Filtering will add robustness to the proposed optimization framework by providing uncertainty-based evaluations of site risks and contaminant transport dynamics. The final phase of research will develop interactive visualization tools that seek to simplify and enhance monitoring design and decision making. The visualization tools will allow decision makers to better understand and explore large monitoring data sets, ultimately enhancing their ability to adaptively manage the networks over time in a cost effective manner.Expected Results:
The tools developed in this research will be demonstrated on a suite of simulated and experimental long-term groundwater monitoring network design applications. The experiments include network design scenarios for a real-world contaminant plume simulation and a scaled physical aquifer transport experiment made available by the University of Vermont. It is anticipated that the results of these test cases will test and validate that the proposed framework is capable of enhancing the environmental decision maker’s ability to design cost effective groundwater monitoring networks. Ultimately, the new design framework will provide significant cost savings over prior design methods while at the same time improving risk management.Supplemental Keywords:
long-term groundwater monitoring network design, groundwater, groundwater contamination, observation network design, uncertainty, risk, risk management, geostatistics, environmental decision making, Ensemble Kalman Filtering, multiobjective evolutionary optimization, evolutionary algorithm, scientific visualization, adaptive management, parallel computing, computational tools, environmental modeling, contaminant transport,, Scientific Discipline, Water, Ecosystem Protection/Environmental Exposure & Risk, Restoration, Environmental Monitoring, Aquatic Ecosystem Restoration, decision making, groundwater remediation, integrated assessment