Geophysical wave tomography

Date of Completion

January 2000

Keywords

Geophysics

Degree

Ph.D.

Abstract

This study is concerned with geophysical wave tomography techniques that include advanced diffraction tomography, traveltime calculation techniques and simultaneous attenuation and velocity tomography approaches. We propose the source independent approximation, the Modified Quasi-Linear approximation and develop a fast and accurate diffraction tomography algorithm that uses this approximation. Since the Modified Quasi-Linear approximation accounts for the scattering fields within scatterers, this tomography algorithm produces better image quality than conventional Born approximation tomography algorithm does with or without the presence of multiple scatterers and can be used to reconstruct images of high contrast objects. Since iteration is not required, this algorithm is efficient. ^ We improve the finite difference traveltime calculation algorithm proposed by Vidale (1990). The bucket theory is utilized in order to enhance the sorting efficiency, which accounts for about ten percent computing time improvement for large velocity models. Snell's law is employed to solve the causality problem analytically, which enables the modified algorithm to compute traveltimes accurately and rapidly for high velocity contrast media. ^ We also develop two simultaneous attenuation and velocity tomography approaches, which use traveltimes and amplitude spectra of the observed data, and discuss some of their applications. One approach is processing geophysical data that come from one single survey and the other deals with the repeated survey cases. These approaches are nonlinear and therefore more accurate than linear tomography. A linear system for wave propagation and constant-Q media are assumed in order to develop the tomography algorithms. These approaches not only produce attenuation and velocity images at the same time but also can be used to infer the physical rock properties, such as the dielectric permittivity, the electric conductivity, and the porosity. A crosshole radar dataset is processed by both approaches and the results are consistent with previous geological and geophysical studies. ^

Link to Full Text

Share

COinS