Monitoring of water volume in a porous reservoir using seismic data: Validation of a numerical model with a field experiment

TL;DR

This study tackles estimating water volume in a porous reservoir from seismic data by integrating a high-fidelity 3D Biot poroelastic–viscoelastic model with a discontinuous Galerkin solver and a data-driven neural network. Training uses synthetic seismic data with realistic noise, while field data from a controlled Laukaa sand pool validate the approach, including a deconvolution step to normalize source wavelets. The neural network maps frequency-domain seismic responses to water volume, and SHAP analysis provides insight into receiver contributions, guiding sensor configuration. The results show that water volume can be recovered with good accuracy using the full receiver array, with SHAP-guided receiver selection offering a robust alternative, highlighting potential for improved groundwater monitoring in controlled or simplified geological settings and informing future extensions to more complex aquifers.

Abstract

As global groundwater levels continue to decline rapidly, there is a growing need for advanced techniques to monitor and manage aquifers effectively. This study focuses on validating a numerical model using seismic data from a small-scale experimental setup designed to estimate water volume in a porous reservoir. Expanding on previous work with synthetic data, we analyze seismic data acquired from a controlled experimental site in Laukaa, Finland. By employing neural networks, we directly estimate water volume from seismic responses, bypassing the traditional need for separate determinations, for example, of reservoir water-table level and porosity. The study models wave propagation through a coupled poroviscoelastic-viscoelastic medium using a three-dimensional discontinuous Galerkin method. The proposed methodology is validated against experimental data, aiming to improve precision in mapping current water volumes and contributing to the development of sustainable groundwater management practices.

Figures (12)

• Figure 1: Photo of experimental setup at Laukaa test site, showing the grid of geophones, weight-drop source, and wells to measure the water-table level.
• Figure 2: A schematic of the problem geometry with the top showing cross-section and the bottom an oblique angle 3D view. Light-blue colour refers to air-saturated and dark-blue to water-saturated zone. Light-brown colour denotes the surrounding elastic material. The setup contains a total of 57 receivers, marked with red dots and one green dot, all of which are located on the ground surface. The green dot serves as the reference receiver and is located between the seventh receiver on lines 1 and 2. The source location is marked with a yellow star and placed close to the fourth receiver on line 1. This figure is motivated by the problem geometry presented in khalili2022monitoring.
• Figure 3: Left: The two source time functions used in the modeling, first derivative of a Gaussian (solid blue) and Ricker wavelet (dashed red). Right: Cross section of an example mesh used in the simulations.
• Figure 4: Measured seismic data: horizontal $v_{\rm s}$ component (left), vertical $w_{\rm s}$ component (right). Reference traces are shown in the top right panels.
• Figure 5: Synthetic seismic data: horizontal $v_{\rm s}$ component (left), vertical $w_{\rm s}$ component (right). Reference traces for noisy and noiseless data are shown in the top panels.