Vertical velocity gradient

Figure 3 shows the velocity as a function of depth at one surface location in the area where the salt body does not interfere with the fault reflections. Although the velocity gradient of about ${\rm 0.4~s^{-1}}$is relatively mild in absolute terms, it is the strongest within this salt model. The effects of the vertical gradient on the moveouts of the reflections after DMO and AMO are measurable, but not too strong. Figure 4 shows the moveouts of the reflection from the fault at a typical CMP location in the area. As expected, the moveout after simple binning (a) points upward with offset, because of the dip effect. Also not surprisingly, the moveout after DMO (c) points downward with offset. Hale and Artley 1993, and several others authors, showed that constant-velocity DMO overcorrects the moveouts in a horizontally layered medium with velocity increasing with depth. In contrast, the moveout from the fault reflection is approximately flat after AMO (b). The simple explanation of these results is that AMO is a residual operator and consequently moves energy less than DMO does. Therefore, it overcorrects the moveout less than constant velocity DMO. In some respects, it behaves similarly to the ``squeezed DMO'' presented by Hale and Artley 1993.