AMO inversion to a common azimuth dataset: Field data results

Robert G. Clapp

bob@sep.stanford.eda

ABSTRACT

I cast 3-D data regularization as a least-squares inversion problem. I form a linear operator that maps from irregular dataspace 5-D data space to a regular 4-D common azimuth volume using a cascade of linear interpolation and Azimuth Move-out (AMO) binning. I regularize the inversion by adding minimizes the difference between various ($t, {\rm cmp}_x, {\rm cmp}_y$) cubes by applying a filter that acts along offset. AMO is used to transform the cubes to the same hx before applying the filter. Further efficiency is gained by inverting each frequency independently. I apply the methodology on marine dataset and compare the results with two more conventional approaches.

The irregularity of seismic data, particularly 3-D data, in both the model domain (in terms of subsurface position and reflection angle) and the data domain (in terms of midpoint, offset, and time) causes imaging problems. Migration methods desire a greater level of regularity than is often present in seismic surveys. There are two general approaches to deal with this problem. One approach is to treat the imaging problem as an inverse problem. The migration operator can be thought of as a linear transform from the recorded data to image space. Duquet and Marfurt (1999); Prucha et al. (2000); Ronen and Liner (2000) use the migration operator in a linear inverse problem to overcome irregular and limited data coverage. A regularization that encourages consistency over reflection angle is used to stabilize the inverse. These approaches have shown promise but are generally prohibitively expensive, and rely on an accurate subsurface velocity model.

Another approach is to try to regularize the data. AMO provides an effective regularization tool Biondi et al. (1998) and is generally applied as an adjoint to create a more regularized volume. These regularized volumes still often contain an `acquisition footprint' or more subtle amplitude effects. Chemingui (1999) used a log-stretch transform to make the AMO operator stationary in time. He then cast the regularization problem as a frequency-by-frequency inversion problem using a Kirchoff-style AMO operator. He showed that the acquisition footprint could be significantly reduced. The downside of this approach is the relatively high cost of Kirchoff implementation of AMO.

Biondi and Vlad (2001) built on the work of Fomel (2001) and set up an inverse problem relating the irregular input data to a regular model space. They regularized the problem by enforcing consistency between the various ($t, {\rm cmp}_x, {\rm cmp}_y$) cubes. The consistency took two forms. In the first, a simple difference between two adjacent in-line offset cubes was minimized. In the second, the difference was taken after transforming the cubes to the same offset AMO. Clapp (2005b) set up the data regularization with AMO as an inverse problem creating a full volume ($t,{\rm cmp}_x,{\rm cmp}_y,hx,hy$). Clapp (2005a) modified this approach to form a Common Azimuth dataset. The full volume of Clapp (2005b) is mapped into a Common Azimuth volume using AMO as part of the inversion.

In this paper, I build on the work of Clapp (2005a). I modify the data mapping operator to allow more mingling between hx volumes in the model space. In addition, each log-stretched frequency is done independently, allowing many more iterations to be applied economically. The methodology is applied on a marine dataset and compared to both a flex-binning and more standard AMO based approach.