Improved Parameter Resolution with Markov Chain Monte Carlo Simulation of Different Aquifer Tests

Area of Science:

• Hydrogeology
• Environmental Science
• Geophysics

Background:

• Aquifer characterization at contaminated sites relies on hydraulic testing.
• Short-duration tests (slug, pumping) are common but often analyzed independently.
• Integrating diverse tests can enhance understanding of aquifer properties.

Purpose of the Study:

• To develop and demonstrate an aggregate inverse analysis method for combining hydraulic tests.
• To improve the resolution and reliability of estimated aquifer properties.
• To assess aquifer uniformity using combined test data.

Main Methods:

• Aggregate inverse analysis using an analytical groundwater flow model.
• Genetic algorithms for initial parameter estimation.
• Markov chain Monte Carlo simulation for parameter distribution estimation.
• Analysis of parameter correlations and standard deviations.

Main Results:

• The combined testing and inverse analysis method successfully estimated aquifer properties.
• Improved resolution of aquifer parameters was achieved compared to single-test analyses.
• The methodology allowed for the assessment of aquifer uniformity.
• Parameter correlations and standard deviations provided metrics for parameter resolution.

Conclusions:

• Integrating multiple hydraulic tests via aggregate inverse analysis enhances aquifer characterization.
• The proposed method offers a robust alternative to traditional inverse methods, avoiding convergence issues.
• This approach is effective for both unconfined and leaky aquifer systems.