Figure shows three subsurface offset gathers taken at different lateral positions. The first SODCIG (panel (a)), taken at about the center of the window in Figure (a) (CMP position 13200 m), shows evidence of multiples at about 4000 m depth at -400 m half-subsurface offset. This multiple is difficult to identify in Figure because it essentially stacks out. It is, however visible in the constant subsurface offset section of panel (b) in Figure (left oval). The second SODCIG (panel (b)), taken below the big salt body in Figure (a) (CMP position 20800 m) shows no clear evidence of multiples below the salt. The strong event below 20000 m and about 5000 m in Figure is actually the base of the salt and the events below it are subsalt primaries (see also the oval in panel (a) of Figure ). The last SODCIG, (panel c), taken where the sediments meet the salt flank (CMP position 24900 m), show evidence of multiples below 4000 m depth. To further illustrate this point, Figure shows a comparison between the image taken at zero subsurface offset, which should be mostly (but not exclusively) primaries, and that obtained at -400 m half-subsurface offset, which should be mostly multiples. Notice in particular the multiple above a depth of 3000 m to the right of the inline distance of 24000 m as indicated by the right oval. Also notice the flattish event at about 4000 m depth and 13000 m inline distance (left oval). This is likely to be a multiple obscuring the weak primaries.
Since I expect to attenuate the multiples in ADCIGs rather than SODCIGs, it is interesting to see how the multiples in this 2D migration are mapping to ADCIGs. This is illustrated in Figure that shows ADCIGs for the same SODCIGs shown in Figure . Again, notice in panel (a) the down curvature of the multiple at about 6000 m at zero aperture angle. There seems to be another multiple at above 7000 m depth. In contrast, no clear evidence of multiple can be seen in the ADCIG in panel (b). In panel (c) there is some evidence of multiples at a depth of about 6300 m depth but at the same depth there seem to be primaries as well. The situation will probably be more clear once the ADCIG is computed after migrating all the data.