Model calibration and parameter structure identification in characterization of groundwater systems
Document Type
Article
Publication Date
1-1-2011
Abstract
This chapter summarizes recent progress in model calibration and model uncertainty assessment in groundwater modeling. The main focus is on parameter structure identification of hydraulic conductivity using the Bayesian maximum likelihood approach. In general, parameter structure identification seeks to identify the parameter dimension, parameter pattern, and associated parameter values. We discuss the importance of parameterization, which is used to define parameter structure and to formulate the inverse problem of parameter structure identification. We propose a generalized parameterization (GP) method for parameter structure identification. GP integrates the traditional zonation and interpolation methods (including kriging) and avoids the shortcomings of each individual method. It captures aquifer heterogeneity by a combination of zonal structure and continuous distribution. An indicator generalized parameterization (IGP) is further introduced to address the interpolation point selection problem, which is an important issue in parameterization. We introduce model selection, model discrimination, and statistical information criteria to select the best parameterization method. To resolve the nonuniqueness problem in parameterization due to lack of data and data uncertainty, we employ Bayesian model averaging (BMA) to consider multiple parameterization methods for parameter structure identification. Finally, based on several optimality criteria we offer the experimental design as the ultimate goal of collecting additional informative data to reduce estimation uncertainty.
Publication Source (Journal or Book title)
Groundwater Quantity and Quality Management
First Page
159
Last Page
202
Recommended Citation
Tsai, F., & Yeh, W. (2011). Model calibration and parameter structure identification in characterization of groundwater systems. Groundwater Quantity and Quality Management, 159-202. Retrieved from https://repository.lsu.edu/civil_engineering_pubs/1152