Multiobjective analysis of a public wellfield using artificial neural networks

Area of Science:

• Environmental science
• Water resource management
• Decision analysis

Background:

• Growing competition for scarce groundwater resources necessitates robust decision-making tools.
• Public water supply wellfields face challenges balancing supply needs with contamination vulnerability.
• Disparate stakeholder interests complicate water resource policy development.

Purpose of the Study:

• To apply decision analysis to a public water supply wellfield to balance water supply volume and contamination vulnerability.
• To identify a defensible policy for managing groundwater resources amidst conflicting stakeholder objectives.
• To develop a compromise pumping strategy considering stakeholder preferences.

Main Methods:

• Developed artificial neural networks (ANNs) using groundwater flow model simulation data.
• Embedded ANN-derived equations into a multiobjective optimization model to identify the Pareto frontier.
• Assigned preference values and power factors to three key stakeholders.
• Utilized distance-based methods to identify a compromise pumping policy.

Main Results:

• The Pareto frontier, illustrating the trade-off between water supply and wellfield vulnerability, was identified.
• A compromise pumping policy was determined that balances conflicting objectives according to stakeholder preferences.
• The study demonstrates a practical application of multiobjective decision analysis in groundwater management.

Conclusions:

• Multiobjective decision analysis provides a rigorous framework for managing complex water resource issues.
• Integrating stakeholder preferences is crucial for developing effective and accepted groundwater management policies.
• The developed approach offers a valuable tool for decision-makers facing similar water supply and contamination challenges.