Figure shows the image extracted at zero offset, which is equivalent to the ``stacked image" for Kirchhoff-like migration methods. The vertical edge is well imaged at zero offset, but when we analyze the image as a function of offset at the fixed surface location corresponding to the vertical salt edge (right panel in Figure ), we immediately notice that, at the depth interval corresponding to the salt edge, there is no focusing along the offset axis. In contrast, the focusing along offset is obvious when we analyze the image at the surface location corresponding to mild reflector dips (left panel in Figure ). As expected, the lack of focusing in the HOCIGs carries over to the image after transformation to angle domain by slant stacking (Figure ). In the next section we will explain the degradation of the horizontal-offset CIGs by a simple geometric analysis.
image-cig-new
Figure 1 Image of the North Sea data set. The black lines superimposed onto the image indicate the positions of the HOCIGs shown in Figure . |
Cig-all-vz
Figure 2 HOCIGs extracted from the prestack image cube. Notice the blurring in the right panel at the depth of the salt edge. |
Ang-Cig-all-vz
Figure 3 Angle-domain CIGs corresponding to the HOCIGs shown in Figure . Notice the blurring in the right panel at the depth of the salt edge. |
Figure shows a vertical-offset CIG (VOCIG) for the same data set. Since the offset is vertical, the image cube is cut at a constant depth, not at constant surface location. The depth of this CIG corresponds to the black line superimposed onto the image in Figure . Now the reflections from the salt edge are focused around zero offset, while the reflections from the flattish reflectors are unfocused. Similarly, after transformation to angle domain (Figure ) the reflections from the salt edge show a slight moveout that could be used to update the migration velocity function. This task would be impossible if we had to rely solely on the information contained in the ADCIG obtained from HOCIG (Figure ). However, while the moveouts from the salt edge are clearly interpretable in (Figure ), the moveouts from flat reflectors are not.
In summary, neither set of CIGs has readily available the information that is needed for velocity updating. In the next section we present a simple method to merge the HOCIG with the VOCIG, and produce a single set of CIGs that satisfies our requirements.
image-cig-hz-new
Figure 4 Image of the North Sea data set. The black line superimposed onto the image indicates the depth of the VOCIGs shown in Figure |
Cig-1.8-vz-hz
Figure 5 VOCIG extracted from the prestack image cube. Notice the good focus at the horizontal location of the salt edge. |
Ang-Cig-1.8-vz-hz
Figure 6 Angle-domain CIG corresponding to the VOCIGs shown in Figure . Notice the slight moveout of the the salt edge reflection. |