Combining Spectral Decomposition and Coherence: mapping lateral discontinuities at a vertical scale

ADVERTISEMENT
From SEG Wiki
Jump to: navigation, search

Spectral decomposition and Dip of Maximum Similarity (Coherence) are two attributes found in Rock Solid Attributes. The use of these two Seismic attributes together can be very beneficial when trying to delineate and map features of varying size throughout seismic. It is possible to break the data down into different spectral scales and also look at the lateral similarity of the wavelets at these different spectral scales. By breaking down the spectral scale, parts of channels or faults that may not be easily seen in full spectrum data are clearer. Smaller features appear better in higher frequencies and larger features appear better in low frequencies. Therefore, the dissimilarities in the data at all different scales can be seen easier. The question becomes the order of the work flow and which one is the most beneficial.

Spectral Decomposition

Spectral decomposition extracts spectral amplitudes from different frequency bands in the data. Spectral Decomposition in Kingdom is broken down into Envelope Sub-band and Trace Sub-band. SD Envelope Sub-band is defined in Kingdom as the "output of the amplitude envelope of the frequency/wavenumber bands from spectral decomposition".2 SD Trace Sub-band is defined in Kingdom as the "output of normal traces containing frequencies or frequency/wavenumber bands from spectral decomposition".2

SDEnvelope.jpg SDTrace.jpg


Dip of Maximum Similarity (Coherence)

Coherence measures the similarity between lateral traces. For Kingdom RSA, Coherence is calculated using the semblance method. The Semblance is computed based off the "measure of coherent power existing between a number of traces versus the total power of all traces".2 This means that the coherency is calculated by looking at lateral traces in a range of dips over a specified window.

CoherenceEDE.jpg


Workflow

EDEworkflowvs.jpg

Spectral Decomposition on Coherence

The images below show Spectral Decomposition run on the Coherence Attribute.

SDE on Coherence.jpg SDT on Coherence.jpg


Coherence on Spectral Decomposition

The images below show Coherence run on 5 different frequency bands from Spectral Decomposition.

Coherence on SDE.jpg Coherence on SDT.jpg

Conclusion

There are large benefits to the combined use of Spectral Decomposition and Coherence. Some of the main uses for this attribute combination can be to map faults and channels. A very helpful tool is to RGB co-blend the different spectral bands to view the complete feature or structure. Being able to map complete channels or faults can help better identify prospects as well as confirm trapping or sealing for a prospect without ambiguity. As seen above, the Coherence run on SD Trace Sub-band give the best image for this attribute combination. This makes sense considering SD Trace Sub-band outputs trace information at the different frequencies. Coherence then uses the trace information directly to calculate similarity trace by trace. The worst image is the SD Trace Sub-band on Coherence because the coherency outputs a measure of similarity with very little trace information remaining. Therefore, when you run SD Trace on the coherency sub-volume, there is no trace information to output.

References

1Fangyu Li and Wenkai Lu (2014). ”Coherence attribute at different spectral scales.” Interpretation, 2(1), SA99-SA106.

2IHS Kingdom. “Kingdom Help.” 2016.

External Links