Ray-based amplitude tomography for crosshole georadar data: a numerical assessment

Citation
K. Holliger et al., Ray-based amplitude tomography for crosshole georadar data: a numerical assessment, J APP GEOPH, 47(3-4), 2001, pp. 285-298
Citations number
23
Categorie Soggetti
Earth Sciences
Journal title
JOURNAL OF APPLIED GEOPHYSICS
ISSN journal
09269851 → ACNP
Volume
47
Issue
3-4
Year of publication
2001
Pages
285 - 298
Database
ISI
SICI code
0926-9851(200107)47:3-4<285:RATFCG>2.0.ZU;2-T
Abstract
Analyses of travel times and amplitudes of crosshole georadar data provide estimates of the electromagnetic velocity and attenuation of the probed med ia. Whereas inversions of travel times are well established and robust, ray -based inversions of amplitudes depend critically on the complex directive properties of the georadar antennae. We investigate the variations of radia tion patterns in the presence of water-filled boreholes and/or changes of e lectrical material properties in the vicinity of the transmitters or receiv ers. To assess the implications of such complicating factors for ray-based georadar amplitude tomography, we generate crosshole georadar data for a su ite of canonical models using a finite difference time domain (FDTD) soluti on of Maxwell's equations in cylindrical coordinates. The emitting dipole-t ype antenna is approximated by an infinitesimal vertical electric dipole, w hereas a corresponding receiving antenna is emulated by recording the verti cal component of the transmitted electric field. Inversions of the amplitud es of these synthetic data demonstrate that the presence of water-filled bo reholes as well as changes in the material properties along the boreholes m ay cause substantial artifacts in the estimated attenuation structure. Furt hermore, our results indicate that ray-based amplitude tomography of crossh ole georadar data is unable to constrain absolute values of attenuation. De spite these inherent limitations, the method is surprisingly robust at dete cting and constraining relative changes in attenuation. In particular, we f ind the method to be highly effective for locating conductivity contrasts t hat are not associated with corresponding changes in dielectric permittivit y, and hence, cannot be located by travel time tomography alone. (C) 2001 E lsevier Science B.V. All rights reserved.