An Integrated Geophysical Approach to Subsurface Imaging, 20-9210

Printer Friendly Version

Principal Investigators
David A. Farrell
Chris L. Hackert
Brian J. Zook
Sitakanta Mohanty
Debra L. Hughson

Inclusive Dates: 08/01/00 - Current

Background - Adequately and cost-effectively characterizing the subsurface poses challenging problems at many sites. Traditional approaches to subsurface characterization often rely primarily on borehole measurement networks. Limitations of this approach include possible cross-contamination along poorly constructed borehole networks, prohibitive costs associated with constructing borehole monitoring networks, and incomplete characterization of the subsurface. Cost-effective subsurface characterization can be achieved through an increasing use of geophysical methods and a decreased reliance on borehole networks. This project explores the development of a coupled geophysical inversion strategy that integrates electrical resistivity tomography with seismic traveltime tomography in an attempt to improve subsurface imaging beyond that achievable by either method.

Approach - Combining seismic traveltime tomography with electrical resistivity tomography in an integrated inversion framework requires developing relationships between seismic velocity and electrical resistivity. This work explores such relationships through an examination of bulk seismic velocity and electrical resistivity field data to identify correlation relationships between these parameters. Linkages between these parameters based on a more fundamental understanding of the physics and physical processes that govern seismic wave propagation and electric current flow in the subsurface are also being investigated. These linkages and correlations are exploited to develop an easy-to-use geophysical data analysis tool that combines electrical resistivity tomography and seismic tomography methods in a coupled iterative inversion scheme that can utilize off-the-shelf software and can be expanded to include other geophysical inversion schemes. Accomplishing this objective involves (i) identifying appropriate software for seismic tomographic inversion, (ii) upgrading existing SwRI resistivity tomography software to improve computational efficiency and robustness, (iii) developing an iterative coupled inversion framework, and (iv) developing a suite of verification problems.

Accomplishments - To date, the project has successfully developed a coupled iterative inversion strategy that combines both electrical resistivity tomography and seismic tomography to produce an image of the subsurface that is superior to the independent images produced by either method. This general purpose iterative inversion software, SeisRes, couples electrical resistivity to seismic velocity through either simple correlation relationships developed from field and laboratory measurements, or more fundamental physical processes that influence seismic wave propagation and electrical potential distribution in the subsurface. Preliminary investigations have shown that using the more physically based approach, additional information such as the porosity and saturation of the medium may be determined. The project has also developed a flexible general-purpose electrical resistivity tomography software package, ResTomo6, which is used in SeisRes. ResTomo6 represents a significant improvement over the previous version of the software, ResTomo5, developed in a previous IR&D project. The upgrade replaces the previously used back-projection methodology with a more robust, computationally efficient finite element formulation that solves Laplace's equation describing electrical potential distribution in a heterogeneous subsurface. The approach also utilizes an adjoint state sensitivity approach that takes advantage of the finite element formulation to determine the Jacobian matrix needed for the Levenberg-Marquardt scheme that is used in the inverse process. The project has developed both 2D and 3D versions of ResTomo6.

2002 Program SwRI Home