NMO-velocity conflict

The fault-plane reflections for the CMP shown in Figure  are free from interferences with the salt body. But these events move quickly under the ``shadow'' of the salt body as the CMP location moves towards the middle of the model. In this area, the fault-plane reflections are still propagating in the sediments, but they are recorded at surface locations that are right above the salt, and at traveltimes larger than the two-way vertical traveltime from both the top and the bottom of the salt. This situation creates a conflict between the NMO velocity that is correct for the flatter events (e.g. top and bottom of the salt) and the NMO velocity that is correct for the fault-plane reflections. This kind of NMO-velocity conflict creates problems for constant-velocity DMO, that has the tendency to overcorrect the moveout of the dipping events, as discussed by several authors (, , ).

Figure  shows a time-slice of the root-mean-square (RMS) velocity computed from the interval velocity model. This time slice is cut at approximately the traveltime of the fault-plane reflection for the 2-2.5 km offset range. The fault plane cuts this time slice from the top-left corner (at a cross-line midpoint of about 4 km) to the bottom middle (at a cross-line midpoint of about 7 km). Figure , Figure , and Figure  show the effects of the NMO-velocity conflict on the partial stacking of the fault-plane reflections. The Figures show time-slices of the result of partial stacking the data over the whole 2-2.5 km offset range after the application of: binning (Figure ), AMO (Figure ), and DMO (Figure ). The fault reflections below the salt, (labeled as ``below'' salt in the Figures) are wiped out by partial stacking after DMO. In contrast, they are reasonably well preserved after either binning or AMO. However, the salt-flanks reflections that are not affected by the NMO-velocity conflict (labeled as ``above'' salt in the Figures) are strongly attenuated after binning, and well preserved after either DMO or AMO. In summary, AMO well preserves all the dipping events, irrespective whether they are ``below'' or ``above'' the salt. In contrast, both binning and DMO attenuate one of the two types of dipping events. The explanation of these results is similar to the explanation of the results obtained in the horizontally layered part of the model. AMO applies a smaller correction to the data than DMO does, and thus AMO does not overcorrect data that have been moved out with a too fast NMO velocity. But AMO also properly corrects the data that are not in the shadow of the salt body and have been moved out with the appropriate NMO velocity.

Similar results are obtained when the same experiment is carried out for the mid-offset range (1.2-2.0 km). Figure - show the results of partial stacking over the whole offset range after the application of: binning (Figure ), AMO (Figure ), and DMO (Figure ). The fault reflections and some of the slat-flanks reflections that are ``below'' the salt are severely affected by partial stacking after DMO. In contrast, they are reasonably well preserved after either binning or AMO. Notice, however, that the steeply dipping salt-flank reflections (at an in-line midpoint location of about 7 km) are better preserved after AMO than after simple binning.

 

Vel-rms Figure 5 Time slice through the RMS velocity function, taken at the same traveltime of the time slices shown in Figure , , and .

 

Ts-Bin-stack-box Figure 6 Time slice through the results of partial stacking the far-offset range (2-2.5 km) after binning. The ``above''-salt reflections are strongly attenuated.

 

Ts-Amo-df-stack-box Figure 7 Time slice through the results of partial stacking the far-offset range (2-2.5 km) after AMO. Reflections from ``below'' and ``above'' the salt are both well preserved.

 

Ts-Dmo-df-stack-box Figure 8 Time slice through the results of partial stacking the far-offset range (2-2.5 km) after DMO. The ``below''-salt reflections are strongly attenuated.

 

Ts-Bin-2 Figure 9 Time slice through the results of partial stacking the mid-offset range (1.2-2 km) after binning.

 

Ts-Amo-df-2 Figure 10 Time slice through the results of partial stacking the mid-offset range (1.2-2 km) after AMO.

 

Ts-Dmo-df-2 Figure 11 Time slice through the results of partial stacking the mid-offset range (1.2-2 km) after DMO.