Semester of Graduation
Fall 2023
Degree
Master of Science in Civil Engineering (MSCE)
Department
The Department of Civil and Environmental Engineering
Document Type
Thesis
Abstract
Uncontrolled groundwater exploitation can lead to aquifer depletion, land subsidence, and saltwater intrusion. Effective groundwater management is challenging due to the intricate nature of subsurface hydrogeology and spatiotemporally variable pumping, especially in a multi-aquifer system. To ensure sustainable withdrawal, multi-objective optimization is an effective tool for balancing management goals and drawdown effects. However, running simulation-optimization using detailed groundwater models is computationally expensive, pushing decision-makers to decide based on limited scenarios. In this study, a hydrogeological framework was constructed for the Capital Area, Louisiana, allowing for individual assessment of each unit to better understand each aquifer's condition. Moreover, a surrogate-assisted simulation-optimization model was developed to determine a set of optimum withdrawal schemes for the Baton Rouge Industrial District. The Baton Rouge Industrial District experiences significant groundwater extraction, around 31 million gallons per day, with the concern of saltwater encroachment and land subsidence. Long short-term memory (LSTM) networks were applied to construct an efficient surrogate model from a detailed groundwater model. Integrating the non-dominated sorting genetic algorithm (NSGA-II) with the surrogate model, the simulation-optimization framework maximized the total withdrawal of groundwater from potential wells and minimized the overall energy expenses associated with pumping and groundwater head drawdown at the monitoring wells. The proposed approach successfully produced non-dominated optimum solutions that align with the defined objectives. Using LSTM proved efficient in constructing a surrogate model for the complex groundwater simulation model, and the study introduced a practical decision-making framework for groundwater management.
Date
11-16-2023
Recommended Citation
Mani, Melika, "Surrogate-assisted simulation-optimization framework for groundwater management in a multi-aquifer system" (2023). LSU Master's Theses. 5878.
https://repository.lsu.edu/gradschool_theses/5878
Committee Chair
Dr. Frank Tsai