INTEGRATING MODEL ABSTRACTION INTO MONITORING STRATEGIES
Environmental Microbial and Food Safety Laboratory
2012 Annual Report
1a.Objectives (from AD-416):
The objective of this research effort is to develop a strategy for incorporating model abstraction techniques into ground-water monitoring and performance assessment programs. The enhanced abstraction strategies and models shall be tested using site-specific databases for specified contaminant sources introduced and monitored in shallow subsurface environments.
1b.Approach (from AD-416):
Analyze the OPE3 tracer experiment using highly-realistic model(s) and subsets of those model(s) with varying degrees of abstraction to simulate a range of model outcomes for comparison to the detailed monitoring datasets; conduct sensitivity analyses on various parameters which were monitored to determine the impact each has on the overall model; use the spatio-temporal geo-statistics and genetic algorithms to optimize the locations and types of sensors required by different conceptual models; use the models abstraction to develop a screening approach for determining the appropriate model abstraction techniques for modeling a specific site.
Two methods were developed using model abstraction techniques to improve contaminant monitoring network designs. Model abstraction is a methodology for reducing the complexity of a simulation models while maintaining the validity of the simulation results with respect to the question that the simulation is being used to address. Model abstraction can help Nuclear Regulatory Commission (NRC) to determine whether simpler models can be used to predict contaminant transport that are easier to understand, and easily communicated to regulators, stakeholders, and the general public, while at the same time adequately representing their sites. Model abstraction used in conjunction with multiple subsurface monitoring networks can be designed for purposes of aquifer characterization, parameter estimation, background concentration characterization, contaminant detection, compliance assessment, plume characterization, source identification, and model discrimination. The methods we developed have resulted in a better understanding of subsurface structural units and fate and transport of pollutants in those units. Both methods have been successfully tested with the data from tracer field experiments. A draft technical document has been prepared for NRC and currently is under peer-review.