Date of Award

2013

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Civil Engineering

First Advisor

Chang, Shoou-Yuh

Abstract

Pollution of groundwater can be harmful to the environment. The use of subsurface contaminant transport models, combined with stochastic data assimilation schemes, can give on-target predictions of contaminant transport to enhance the reliability of risk assessment in the area of environmental remediation. Observation data are required to guide the deterministic system model to assimilate the true state of the contaminant. Modeling the behavior of contaminant in groundwater is imperative in predicting the fate of the pollutant, in risk assessment, and as a preceding step of the remission process. In this study a two-dimensional transport model with advection and dispersion is used as the deterministic model of contaminant transport in the subsurface. An Adaptive Extended Kalman filter (AEKF) is constructed as a stochastic data assimilation scheme to meliorate the prediction of the contaminant concentration. Simulation results are shown to compare the performance of the numerical, the Extended Kalman filter and the AEKF. The effectiveness of the AEKF is determined by using a root mean square error (RMSE) of pollutant concentrations in contaminant transport modeling. The results of the models indicate that, at the end of the simulation, the introduction of the Extended Kalman filter improved the deterministic model prediction by reducing the model error from 28 mg/L to 18 mg/L, thus improving the prediction accuracy by approximately 35.7%. The AEKF was successful in reducing the errors in the Extended Kalman filter prediction from 18 mg/L to 11 mg/L hence ameliorating prediction accuracy by approximately 38.9%. In general, the implementation of the AEKF was successful in improving the prediction accuracy of the deterministic model by about 60.7% which shows a substantial improvement in the prediction of the contaminant concentration in the subsurface environment.

Share

COinS