-
oa Predicting Underground Water Levels from Seismic Exploration Parameter - Vs30 using Machine Learning Methods
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 18th International Conference Monitoring of Geological Processes and Ecological Condition of the Environment, Apr 2025, Volume 2025, p.1 - 5
Abstract
The study researches the possibility of using the average shear-wave velocity in the top 30 meters (Vs30) as a proxy parameter to predict underground water levels through machine-learning methods. By focusing on a test region in Texas (USA), the research leverages open-source datasets for Vs 30, underground water depths, and lithology, which are collocated using a cKDTree algorithm to maintain the integrity of original measurements. Three models are employed – a simple linear regression (as a baseline), a Random Forest regressor, and an Extra Trees regressor. While the linear model assumes a direct relationship between velocity and underground water depth, ensemble methods account for the more complex and spatially variable nature of near-surface geological conditions. The workflow is potentially may be transferable to areas where Vs30 data are available but direct hydrogeological measurements are scarce. Future studies could validate this idea in different geological/geophysical settings, incorporate additional engineering and geological parameters (e.g., porosity, permeability), and explore further machine-learning techniques (e.g., support vector machines or neural networks) to increase robustness and accuracy of underground water level predictions.
