TTI anisotropic RTM produces superior subsalt imaging

CGGVeritas has developed a Tilted Transverse Isotropy (TTI) anisotropic reverse time migration (RTM) algorithm. This, combined with the added azimuthal information available from wide-azimuth data, has been used extensively in the production processing of many surveys from around the world to provide considerable uplift in the imaging of structures beneath complex salt bodies.

In recent years RTM has become the preferred solution for depth imaging in complex areas, such as beneath an irregular salt canopy. RTM is based on a full two-way solution to the wave equation which correctly handles complex velocity fields to construct images of all possible arrivals, including turning waves and prism waves, while imposing no dip limitations on the image. In many cases this ability to make use of the complex wave modes allows imaging of parts of the subsurface that otherwise have no direct illumination.

Until recently, the imaging of structures lying beneath dipping anisotropic overburdens has been limited to ray-based TTI migration algorithms. Unfortunately, these do not perform well in comparison with wave-equation algorithms for imaging beneath complex overburdens. Although upgrading ray-based algorithms for TTI is relatively straightforward, the same is not true for wave-based algorithms. In these the implementation is complex, often unstable and computationally intensive. Early RTM applications were limited to isotropic and vertical transverse isotropy implementations due to the difficulty of incorporating TTI. In areas of dipping sedimentary sequences such as around salt and in mini-basins, the images were not very clear.

TTI imaging in the Gulf of Mexico (GoM) has been impeded further by the lack of appropriate check shots and vertical seismic profiles available to constrain the anisotropic parameters. However, the increase in wide-azimuth (WAZ) acquisition has meant that more reliable TTI parameters can now be derived. The greater range of azimuths available has shown TTI effects that were not apparent in narrow-azimuth data, and this provided the impetus for CGGVeritas to develop a TTI RTM algorithm (Zhang and Zhang, 2008).

To study the effect of TTI anisotropy, two orthogonal surveys over a salt-withdrawal mini-basin were processed through several iterations of dual-azimuth tomography, and common image gathers (CIGs) at the same output locations were compared. In the steeply dipping mini-basin one set of gathers was found to be over-corrected and the other under-corrected. A VTI velocity update failed to resolve this problem. This velocity variation with azimuth is attributable to TTI anisotropy, caused by dipping sediments in the mini-basin. The long offsets and orthogonal azimuths provided sufficient information to incorporate TTI into the velocity model for a further update, which flattened both sets of CIGs. Only TTI imaging could flatten all the CIGs, and the resulting velocity model also produced more geologically viable subsalt structures in the study area. Unsurprisingly, further comparison studies of wide-azimuth data have shown that TTI RTM produces more coherent and focused images than isotropic or VTI RTM, particularly along steeply dipping salt flanks, in mini-basins, and in subsalt regions.

The difficulties of performing TTI RTM migration in general production have now been overcome. With the increasing availability of WAZ data, the derivation of TTI parameters is possible. CGGVeritas has been able to combine TTI and RTM in the processing of many WAZ datasets from the deepwater GoM and from other places around the world to produce superior subsalt images. These show more accurate focusing, positioning, and amplitudes in complex areas, with better depth control, better terminations of sediments against the salt flanks, and enhanced continuity of deeper reflections.

TTI RTM and related topics remain a focus for research in CGGVeritas. In January Zhang and Sun published a paper in First Break describing a true amplitude version of RTM that provides the correct amplitude vs. angle relationship and showing how noise attenuation could be handled naturally in the common reflection angle domain. Further research is continuing, especially into the improvement of the derivation of TTI parameters from WAZ data and extending full wavefield inversion for TTI media.

For more information, visit www.cggveritas.com/RTM.