Gabor depth imaging using a new adaptive partitioning algorithm

Yongwang Ma, Gary F. Margrave

Wavefield extrapolation by spatially variable phase shift is currently a very competitive depth migration technique. In this paper, we present prestack migration results of the Marmousi synthetic dataset using a new adaptive Gabor wavefield extrapolation.

Gabor depth imaging algorithm can be used to approximate the generalized phase-shift-plus-interpolation or the non-stationary phase shift, which are two extreme cases in Gabor imaging schemes. Therefore, there are many ways to explore wavefield extrapolations with Gabor imaging method. The key to an efficient Gabor imaging algorithm is to develop an adaptive partitioning scheme that only localizes the wavefield as required by the lateral velocity variation. We have tested three methods of adaptive partitioning, however, those two presented previously have their limits and they are relatively more difficult to implement in 2D (for 3D imaging) than the one we will describe. We present the details of the new adaptive partitioning method in 1D (for 2D imaging). The extension of the partitioning method to 2D will be described in another paper in this volume. This method creates adaptive partitions using a controlled lateral position error. Software has been developed using the new adaptive partitioning algorithm, reducing substantially computation burden in depth imaging when compared to the full generalized phase-shift-plus-interpolation integral. The performance of Gabor depth imaging using this adaptive partitioning algorithm is illustrated with images from prestack depth migration of the Marmousi dataset.