Complete anisotropic analysis of three component seismic data related to the marine environment and comparison to nine component land seismic data

by 1972- Gumble, Jason Ethan

Anisotropy and fracture characterization in individual layers is realized through iterative layer stripping corrections of PS-wave reflection data in four, converted-wave (PS-wave) synthetic seismic data sets, generated from azimuthally anisotropic (HTI and TTI) models, and a 4-C data set from the Teal South, Gulf of Mexico. The corrections were applied on a layer by layer basis with the objective to evaluate the efficacy of constant polarization rotation and time shift operators. Equivalent isotropic models are compared to anisotropic models after layer stripping corrections using rms amplitude and shear-wave splitting time difference maps to quantify and identify inherent errors in estimation of seismic polarization parameters. For HTI media it is found that radial and transverse components of PS data that have had layer stripping corrections applied, exhibit incorrect symmetry and orientations. This may adversely affect inversion and/or AVO and AVAZ analysis. Layer stripping corrections applied to fast and slow (PS1 and PS2) components exhibit the correct symmetry and orientation. Time differences between PS1 and PS2 are computed using crosscorrelation. Previous studies have addressed some of the problems associated with layer stripping corrections for the case of vertical fractures (HTI media) and poststack layer stripping analyses. This study includes 1 This chapter published as: Gumble, J. E. and J. E. Gaiser, 2006, Characterization of layered anisotropic media from prestack PS-wave-reflection data, Geophysics, 71, D171-D182. 22 an equivalent model with dipping fractures (TTI media) and extends the scope to encompass the effects of anisotropy on prestack data. The results from an application of the same crosscorrelation are also applied to a limited set of 4-C data from the Teal South project in the Gulf of Mexico. Results are consistent with those of previous studies involving solely poststack 4C rotation analysis in terms of average, or zero offset, time differences and symmetry orientation. Offset and azimuth amplitude/traveltime variations, however, indicate that there is more information contained in prestack seismic data than 4C rotation can comprehend.