Post-stack velocity analysis by separation and imaging of seismic diffractions

Salt detection

gom,gom-dip
Figure 5. Second test example. (a) Near-offset section from a Gulf of Mexico dataset. (b) Local slopes estimated by plane-wave destruction.

gom-pwd,gom-pik
Figure 6. Diffraction separation. (a) Diffraction events separated from data in Figure 5(a). (b) Migration velocity picked from local varimax scans after velocity continuation of diffractions.

gom-slc,gom-slc2
Figure 7. Migrated images. (a) Migrated diffractions from Figure 6(a). (b) Initial data from Figure 5(a) migrated with velocity estimated by diffraction imaging.

Figure 5(a) shows another example, also from the Gulf of Mexico. We used the nearest-offset section for diffraction analysis. Plane-wave destruction estimates dominant slopes of continuous reflection events [Figure 5(b)] and reveals numerous diffractions generated by rough edges of a salt body [Figure 6(a)]. We used shaping regularization (Fomel, 2007b) with the smoothing radius of 40-by-10 samples to constrain the slope-estimation process. Focusing analysis generates a time migration velocity [Figure 6(b)] suitable for collapsing diffractions [Figure 7(a)]. Both sharp edges of the salt body and continuous specular reflections appear in the final image [Figure 7(b)]. Inevitably, prestack depth migration (as opposed to time migration) is required to properly position the salt boundary in depth. Time migration, however, provides a reasonable first-order approximation computed at a small fraction of the cost.

Post-stack velocity analysis by separation and imaging of seismic diffractions