COMMON-AZIMUTH VELOCITY ESTIMATION

Another attractive application of common-azimuth processing of 3-D marine data is the possibility to perform velocity estimation based on the focusing of wavefield at depth. ``Focusing analysis'' has been successfully applied to estimate depth-migration velocities for 2-D data MacKay and Abma (1992); Yilmaz and Chambers (1984). It has some inherent advantages with respect to velocity estimation based on coherency of migrated events across offsets, that is the criterion routinely used in conjunction with Kirchhoff prestack migration. Residual moveout applied to migrated constant-offset data is only a partial residual focusing operator, since it does not refocus the data along the midpoint axes. By contrast, focusing analysis based on downward-continuation operators does include a midpoint component. Because of these inherent advantages, velocity-estimation methods based on wavefield focusing, have been revived lately and proposed for 3-D velocity estimation as Common Focal Point (CFP) methods Kabir and Verschuur (1996); Morton and Thorbecke (1996). However, the straightforward application of CFP methods to 3-D data with arbitrary geometry could be extremely computationally intensive. Whereas, the application of CFP methods to marine data using the common-azimuth imaging tools described in this paper would be computationally efficient, and thus they could become practical methods to estimate velocity for 3-D prestack depth imaging of a wide class of marine data sets.