Identifier
etd-07082014-233129
Degree
Doctor of Philosophy (PhD)
Department
Civil and Environmental Engineering
Document Type
Dissertation
Abstract
Groundwater resources are vital for sustainable economic and demographic developments. Reliable prediction of groundwater head and contaminant transport is necessary for sustainable management of the groundwater resources. However, the groundwater simulation models are subjected to uncertainty in their predictions. The goals of this research are to: (1) quantify the uncertainty in the groundwater model predictions and (2) investigate the impact of the quantified uncertainty on the aquifer remediation designs. To pursue the first goal, this study generalizes the Bayesian model averaging (BMA) method and introduces the hierarchical Bayesian model averaging (HBMA) method that segregates and prioritizes sources of uncertainty in a hierarchical structure and conduct BMA for saltwater intrusion prediction. A BMA tree of models is developed to understand the impact of individual sources of uncertainty and uncertainty propagation on model predictions. The uncertainty analysis using HBMA leads to finding the best modeling proposition and to calculating the relative and absolute model weights. To pursue the second goal of the study, the chance-constrained (CC) programming is proposed to deal with the uncertainty in the remediation design. Prior studies of CC programming for the groundwater remediation designs are limited to considering parameter estimation uncertainty. This study combines the CC programming with the BMA and HBMA methods and proposes the BMA-CC framework and the HBMA-CC framework to also include the model structure uncertainty in the CC programming. The results show that the prediction variances from the parameter estimation uncertainty are much smaller than those from the model structure uncertainty. Ignoring the model structure uncertainty in the remediation design may lead to overestimating the design reliability, which can cause design failure.
Date
2014
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Recommended Citation
Chitsazan, Nima, "Bayesian Saltwater Intrusion Prediction and Remediation Design under Uncertainty" (2014). LSU Doctoral Dissertations. 394.
https://repository.lsu.edu/gradschool_dissertations/394
Committee Chair
Tsai, Frank
DOI
10.31390/gradschool_dissertations.394