The real data used in this study were provided by WesternGeco. The synthetic model used in the previous section was modeled after this real dataset. Thus many of the imaging problems faced in the synthetic data presented themselves in the real data. Since we had a rough idea of the subsurface geology, the velocity analysis was relatively easy. The actual geology and synthetic model are likely very different, but the size and shape of the salt body is fairly uniform between the two, and the velocity gradients are comparable. Figure 11 shows extracted image gathers from both outside and under the salt. We can see that the events are relatively flat, so our AVA algorithm should work well. Figure 12 shows the stack of the angle gathers. There are a few notable bright spots around the right side of the salt and closer to the bottom of the salt that may correspond to hydrocarbons.
The A*B image of the WesternGeco data is shown in Figure 13, and the image after the AVA muting algorithm is show in Figure 14. Overall, the AVA algorithm seems to have worked well, with the A*B image confirming that the bright spots in the stacked image could very well be reflections from a low impedance gas or oil sand. Similar to the synthetic image, the algorithm helps to illuminate reflectors under the salt. However, the real data image seems to have even less energy under the salt. This is expected since the wavefield in the real data undergoes a lot more scattering than in the synthetic model.