Fault prediction method based on multi-azimuth coherence attribute

Seismic coherence attribute is the key technology for identifying large and mesoscale structures and stratigraphic anomalies in seismic interpretation, while the current conventional attributes are based on post-stack seismic data. The development of wide-azimuth seismic acquisition and interpretation technology provides new ideas for improving the seismic identification accuracy of underground structures, but it is still challenging to use informative wide-azimuth seismic data to carry out seismic geometric attributes research. Therefore, this paper makes full use of the difference of seismic anomalies caused by geological discontinuities in different observation azimuths, and the azimuthal seismic data perpendicular to strikes of fault have a better effect on identifying the fault, we develop a method to extract multi-azimuth coherence attribute based on azimuthal seismic data. First of all, the seismic traces in the spatial analysis windows of the azimuthal seismic data are interpolated, and azimuthal analysis windows are constructed along different directions. Secondly, the traces at different positions in azimuthal analysis window are weighted by inverse distance weighting algorithm when calculating the azimuthal coherence. Finally, we fuse the information in different azimuthal coherence by principal component analysis, and the normalized result is defined as the multi-azimuth coherence. The effectiveness and stability of the method proposed in this paper have been verified in practical applications. Compared with the conventional methods, the proposed method contains the characteristics of the geological structure in different azimuth, highlights the geological edge in all directions, and provides local and global discontinuities, in addition, the signal-to-noise ratio of multi-azimuth coherent images has been improved effectively.