Acknowledgments

I thank Gabriel Alvarez, Guojian Shan and Madhav Vyas for discussions and suggestions, and Antoine Guitton for the program to perform adaptive subtraction. I also thank WesternGeco for providing the data set.

 

or00deg
Figure 5 Multiple-attenuated data -- zero-angle section. Strong migrated multiples make interpretation of primaries impossible.

 

pest00deg
Figure 6 Residual-multiple-attenuated data -- zero-angle section. Residual multiples have been attenuated.

 

zzo
Figure 7 Details of Figures and -- zero-angle sections. (a) and (c): multiple-attenuated data. (b) and (d): residual-multiple-attenuated data.

 

orstk
Figure 8 Multiple-attenuated data -- stacked data. Residual multiples dominate the deeper portion of the data.

 

peststk
Figure 9 Data after residual-multiple attenuation -- stacked data. The final data is much cleaner than the multiple-attenuated data.

 

zstk
Figure 10 Details of Figures and at the same portion as shown in Figure -- stacked data. (a) and (c): multiple-attenuated data. (b) and (d): residual-multiple-attenuated data. Notice persistent multiples in the original data on panels (a) and (c) and how attenuated they are on panels (b) and (d).