Methods of extracting FEAVO effects in the angle gathers

In order to invert, one must extract the anomalies. Since they are expressed both in the midpoint-angle domain and in the angle-domain common image gathers, their separation must proceed in a synergistic fashion. Figure 6 shows they bear a striking morphological resemblance to channels, so pattern-recognition techniques applied to a coherence volume Marfurt et al. (1998) of the prestack cube may yield good results. Another way to discriminate them may be building a dip volume Fomel (2000) using plane-wave destructors Claerbout (1991). This may be successful because it is actually biomimicry: the human eye discriminates the anomalies based on their dip, or in the case of the data-domain hyperbola sags, on the discontinuities in the dip (dip derivative). These morphological approaches may be corroborated with spectral discrimination. In some cases, such as for the Grand Isle dataset, the FEAVO-causing velocity anomalies are small enough that different wavelengths may be affected differently (Figure 8).

 

kfdp
Figure 8 Top: CMP gather at 4.7 km, filtered with 4-10 Hz, 20-25 Hz, 60-70 Hz. Bottom: section with offset = 1.291km, filtered with 4-180 Hz, 4-10 Hz, 20-25 Hz, 60-70 Hz. Different frequency bands result in different extents of the FEAVO anomalies.