The original Marmousi data set consists of 240 shots with 96 traces per shot. The first offset is 200 m; both the shot and the receiver spacings are 25 m. The first shot is at the lateral position 3000 m, and the recording time is 2.9 s. Overall, I used the same survey geometry and settings used by IFP, the builders of the original Marmousi data set. This includes the same source and receiver locations, an identical sampling interval and recording time, and the same minimum offset.
However, I made two adjustments to the original Marmousi acquisition settings:
Figure 5 shows a common-shot gather (the first shot of the survey) from the original Marmousi data set , on the top, and the corresponding common-shot gather for the new VTI Marmousi data, on the bottom. The moveout of most reflections at this shot point appears hyperbolic, overall, a direct indication of the smoothness of the model under this shot point location. The reflection at two seconds might be the only exception. Head waves, which are the first arrivals at moderate to large offsets, are faster for the VTI case than for the isotropic case. The difference of 0.2 s, or 12 percent, is directly attributable to the average value of about 0.15 near the surface in that shot-point area. In addition, all reflections at far offsets arrive much earlier in the anisotropic model than the isotropic one, despite the fact that the two models have the same NMO velocity. We can attribute these differences mainly to the nonhyperbolic moveout, commonly associated with anisotropy.
Another common-shot gather, shown in Figure 6 and caused by a source at 5000 m, demonstrates additional complications. The moveout no longer appears hyperbolic, which directly reflects the complexity of the model under this shot point. Differences between the isotropic (top) and anisotropic (bottom) sections are apparent. The diffraction at 0.7 seconds has large amplitudes in the isotropic section, and low amplitudes in the anisotropic one. We attribute this difference to the lower velocity contrast between the diffractor, which is of a high velocity, and the medium above it in the anisotropic model. Such amplitude differences have large implications in amplitude-versus-offset (AVO) analysis of reflections. Well-log (vertical) velocities do not constrain the AVO behavior in anisotropic media; the parameter is also needed to do so.
Another common-shot gather, shown in Figure 7 and resulting from a source at 7000 m, is recorded at a more isotropic, yet complex, region of the anisotropic model. The overall differences between the isotropic (top) and anisotropic (bottom) sections are smaller than those associated with Figures 5 and 6. Head waves for the isotropic and anisotropic models at far offsets arrive at less than 0.1 second apart in this region. As a result, we can expect isotropic imaging to provide more reasonable results in this region.