|Van Genuchten, Martinus|
Submitted to: Government Publication/Report
Publication Type: Government Publication
Publication Acceptance Date: 9/5/2005
Publication Date: 12/1/2006
Citation: Pachepsky, Y.A., Guber, A.K., Van Genuchten, M.T., Nicholson, T.J., Cady, R.E., Simunek, J., Schaap, M.G. 2006. Model abstraction techniques for soil water flow and transport. Government Publication/Report-NUREG. http://www.nrc.gov/reading-rm/doc-collections/nuregs/contract/cr6884/. Interpretive Summary: It has been amply demonstrated that the increase in complexity of the model does not necessarily mean the increase in its accuracy. If the complexity is not inexorably linked with the accuracy, there may exist an opportunity to simplify models. Model abstraction is defined as the methodology for reducing the complexity of a simulation model while maintaining the validity of the simulation results with respect to the question that the simulation is being used to address. The need in model abstraction may stem from the need to improve the reliability and reduce uncertainty of simulations, to make the modeling and its results more explicable and transparent to the end users, and to enable more efficient use of available resources in data collection and computations. The purpose of this work was to review the available model abstraction techniques, to select model abstraction techniques that are relevant to the subsurface contaminant hydrology, to develop a systematic model abstraction process, and to demonstrate this process with a field study data. Model abstraction techniques were collected that would simplify subsurface hydrology models by changing their structure and parameter determination. A systematic and objective process of model abstraction was proposed. The case study of water flow in soil with macropores was developed that elucidated the advantages and limitations of model abstraction. In particular, benefits from model abstraction were found when the base model gave inexplicable results. This report serves as the technology transfer document to the NRC staff for the consultative and reference use in reviewing subsurface hydrologic models. The implementation of the model abstraction methodology provides a context for more efficient use of modeling in regulatory process. The model abstraction addresses the uncertainty of model structure and model parameter determination, and facilitates transparency and realism of the site-specific modeling essential features, events and processes.
Technical Abstract: This report describes the methodology of model abstraction in subsurface hydrology. Model abstraction is defined as the methodology for reducing the complexity of a simulation model while maintaining the validity of the simulation results with respect to the question that the simulation is being used to address. The need in model abstraction may stem from the need to improve the reliability and reduce uncertainty of simulations, to make the modeling and its results more explicable and transparent to the end users, and to enable more efficient use of available resources in data collection and computations. A comprehensive review of model simplification techniques developed in subsurface hydrology is included. Abstractions of both model structure and model parameter determination are described. A systematic and objective approach to the model abstraction is outlined. A case study is presented that is designed to understand how model abstraction can affect performance assessment of contaminant migration at a relatively humid site. The model abstraction methodology is generic but designed to be of practical use to NRC licensing staff in their review of the performance assessment of waste disposal sites and recovery facilities. The model abstraction can serve as the device for the NRC licensees to determine whether a simple model can be used that is easy to understand and communicate to regulators, stakeholders, and the general public, and can adequately represent their site.