Future Directions

The obvious direction in which to move this project is migration. By using migration, rather than NMO, as the imaging operator, the limiting assumptions of NMO (v(z), flat reflectors) can be abandoned. Furthermore, the limitations of operating in the offset domain can be overcome by moving to the more intuitive angle domain. In some cases, multiples provide better angular coverage over a recorded cable length. Systematic integration of this extra information could prove revolutionary in regions of poor illumination. But the fruits of this transition are not without challenges. Because NMO is a vertical mapping, each CMP can be processed independently, making the memory requirements of the current implementation reasonable, and parallelization quite simple. Correctly handling the transformation of amplitudes between primary and pseudo-primary may prove even more of a challenge. Also, regardless of whether NMO or migration is used, the move to 3-D is never a forgiving one from the computational standpoint.