To boldly go into the next dimension: 3D raypath interferometry issues

David C. Henley

ABSTRACT

The technique known as raypath interferometry was developed to correct seismic reflection data for difficult near-surface conditions by generalizing and relaxing the assumptions used by conventional surface-correction algorithms (residual statics). We have demonstrated the success of the technique on model data as well as on several sets of 2D field data, both PP and PS. The method improves not only the alignment and coherence of reflection events, but also their waveform consistency.

We have begun to extend raypath interferometry to 3D. In previous work, we introduced the source-receiver azimuth as a third dimension for grouping 3D traces for analysis. Using common-azimuth bins, we need only a 2D trace transform (initially the RT, or radial trace transform) to move the 3D data to a common-raypath domain for applying interferometry. We demonstrated the success of this approach by applying it to the vertical component of the Blackfoot 3D-3C field data, and showing improved event coherence at all stages of the process except the CMP stack, which was not computed due to limitations in our inverse RT Transform.

Here, we apply raypath interferometry to the radial horizontal (PS) component of the same Blackfoot data set. As with the PP component, we demonstrate improved event coherence, but do not compute CCP stack traces, due, once again to our limited RT Transform preventing a proper inversion. Hence, we explored the alternative of replacing our RT Transform with an invertible Tau-P Transform. The Tau-P Transform, however, is not compact, requiring at least an order of magnitude more storage than its input X-T gather, when performed with sufficient resolution to allow high-fidelity inversion. In future work, we will implement data storage strategies that will enable full comparisons of CMP and CCP stack traces of corrected and uncorrected 3D-3C data.

View full article as PDF (7.67 Mb)