Nonlinear inversion of borehole-radar tomography data to reconstruct velocity and attenuation distribution in earth materials

Journal of Applied Geophysics
By: , and 

Links

Abstract

A nonlinear tomographic inversion method that uses first-arrival travel-time and amplitude-spectra information from cross-hole radar measurements was developed to simultaneously reconstruct electromagnetic velocity and attenuation distribution in earth materials. Inversion methods were developed to analyze single cross-hole tomography surveys and differential tomography surveys. Assuming the earth behaves as a linear system, the inversion methods do not require estimation of source radiation pattern, receiver coupling, or geometrical spreading. The data analysis and tomographic inversion algorithm were applied to synthetic test data and to cross-hole radar field data provided by the US Geological Survey (USGS). The cross-hole radar field data were acquired at the USGS fractured-rock field research site at Mirror Lake near Thornton, New Hampshire, before and after injection of a saline tracer, to monitor the transport of electrically conductive fluids in the image plane. Results from the synthetic data test demonstrate the algorithm computational efficiency and indicate that the method robustly can reconstruct electromagnetic (EM) wave velocity and attenuation distribution in earth materials. The field test results outline zones of velocity and attenuation anomalies consistent with the finding of previous investigators; however, the tomograms appear to be quite smooth. Further work is needed to effectively find the optimal smoothness criterion in applying the Tikhonov regularization in the nonlinear inversion algorithms for cross-hole radar tomography. 

Publication type Article
Publication Subtype Journal Article
Title Nonlinear inversion of borehole-radar tomography data to reconstruct velocity and attenuation distribution in earth materials
Series title Journal of Applied Geophysics
DOI 10.1016/S0926-9851(01)00071-4
Volume 47
Issue 3-4
Year Published 2001
Language English
Publisher Elsevier
Contributing office(s) Office of Ground Water
Description 14 p.
First page 271
Last page 284
Country United States
State New Hampshire
City Thornton
Other Geospatial Mirror Lake
Google Analytic Metrics Metrics page
Additional publication details