Introduction To December Focus: A glimpse into the future of seismic imaging?

Slowly, inexorably, prestack depth migration (PSDM) is replacing prestack time migration (PSTM) as our seismic imaging workhorse, just as prestack time migration previously replaced poststack time migration. We saw this first in the Canadian Foothills, where anisotropic PSDM is required to correctly position the images of our hydrocarbon-bearing reservoirs beneath the thick and, in places, steeply-dipping anisotropic overburden (e.g. Vestrum, 2003). PSDM then moved offshore, and is now being routinely applied to improve our confidence in the imaging of the deep section before new exploration wells, which can cost hundreds of millions of dollars, are drilled. Most recently, we heard from WPX Energy during questions at this year’s URTeC conference that PSDM is required for geosteering in unconventional shale plays because it “gives much better imaging of small faults than PSTM” (Southcott & Harper, 2014).

As we look to the future, we may expect further imaging improvements as we reduce the number of explicit (or implicit) approximations that we make during seismic processing in areas such as attenuation compensation, noise attenuation, residual statics, the modeling of the velocity and anisotropy fields, and migration. In our first focus article, Juergen Fruehn demonstrates how improving the resolution of the PSDM velocity model by relaxing some of the constraints in their tomographic migration velocity model building workflow leads to higher resolution, more geological velocity models, and improves the final imaging. Presumably, these higher resolution velocities also yield better pore pressure predictions.

The next three focus articles, which were independently submitted, show the value of integrating compressional wave (PP) and converted wave (PS) data into a common imaging solution. This is somewhat analogous to the uplift in data quality that may be obtained by integrating PP and PS data during AVO inversion (e.g. Barnola & Ibram, 2012). Terence Krishnasamy’s article describes how PP and PS data extracted from four component ocean bottom data can be simultaneously processed during V_p and V_s tomographic velocity model building. This naturally couples the resulting PP and PS depth images, making them ideally conditioned for joint AVO inversion. Zaiming Jiang presents a framework for reverse-time depth migration of PP and PS data, and uses data from the synthetic Marmousie-2 model to illustrate the advances that he made in its implementation during his Ph.D. studies at the University of Calgary. Aaron Stanton, presently working towards his Ph.D. at the University of Alberta, shows us with synthetic data how least squares migration can be used to damp down sampling, illumination and cross-talk artifacts during one-way wave equation depth migration of PP and PS data, thereby improving the resolution of the ensuing PP and PS images.

In the final focus article, David Cho shows us how we can minimize misties between our seismic and well data, using a local Q correction during seismic stretching and squeezing. Such workflows will be required long into the future, even as PSDM becomes the standard imaging engine, as smoothness constraints imposed during velocity model building, velocity/anisotropy ambiguity, noise contamination, and limited seismic offsets will continue to stymie our attempts to build imaging velocity models that perfectly match the velocity data in our wells.