CONCLUSIONS

Residual NMO+DMO applied after an initial prestack migration indicates that there is information about the lateral variation of the velocity of the subsurface contained in the dataset shown. Inversion of the relation between residual slowness and interval slowness gives an update to the interval slowness model used for the initial prestack migration. The higher lateral wavenumbers of this update are probably unreliable, due to non-uniform reflector coverage and near-sea-floor effects. The lower lateral wavenumbers give a physically plausible and intuitive change to the slowness model. I make the usual disclaimer, more investigation is needed.