Real data example

I tested the methodology on a real 3-D marine dataset from the North Sea. Figures  and are from this dataset. Previous uses of AMO and common azimuth migration have resulted in noticeable acquisition footprint in the first 1000 meters Biondi (1999); Vaillant and Sava (1999). For the test I used a maximum of 40 iterations, with a maximum reduction in residual of 35%. A large reduction would be preferable but many frequencies did not reduce by even 20% after 40 iterations. Figure  shows a constant offset section after regularization with fitting goals (4). Note the absence of an acquisition footprint. Further, note how we have successfully filled even the large hole visible in the fold map of Figure .

 

const-off
Figure 7 A constant offset section from a real 3-D marine dataset obtained by applying fitting goals (4). Note the absence of an acquisition footprint.

I then applied common azimuth migration to the data. Figure  show three slices from the zero-offset migration cube. Pay particular attention to the depth slice. Note how the acquisition footprint has disappeared.

 

mig
Figure 8 The result of migrating the data show in Figure . Note that virtually no acquisition footprint is visible in the data.