Stacking is an important process to the seismic exploration industry. It is an effective way to reduce the size of data sets and to enhance reflections while attenuating noise. However, the validity of stacking multiple-coverage data is questionable in the case of PS converted wave data because, even for a horizontal reflection in a constant velocity media, raypaths in a CMP gather strike different reflection points.

Prestack partial migration operators are useful tools in reducing the size of
seismic data. Dip moveout (DMO) is the most common prestack
partial migration operator. Rosales (2002)
comments on a series of DMO operators for PS data. The operators
differ in numerical approximations of the moveout equation, processing
domain and implementation domain. He also introduces a more
accurate PS-DMO operator in the log-stretch *f-k* domain that
gives an appropriate amplitude distribution.

Biondi et al. (1998) introduce a more general prestack partial migration operator called Azimuth Moveout (AMO). AMO has the advantage of transforming prestack data into equivalent data with arbitrary offset and azimuth, moving events across midpoints according to their dip. Several advantages have been described for the AMO operator. Among them are: 1) partial stacking of prestack data, in order to create regularly sampled common offset-azimuth cubes Biondi (2000); Chemingui and Biondi (1997); Chemingui (1999) and 2) data regularization of irregular sampled data which preserves amplitudes Biondi and Vlad (2001).

This work presents the equivalent of the PP-AMO operator for converted wave data. We explain the geometrical interpretation of our PS-AMO operator, in which the concept of CCP transformation is important since it is the base for event movement according to its dip. Our PS-AMO operator is a cascade operation of PS-DMO and inverse PS-DMO. We exploit the knowledge of the fast and accurate PP-AMO in the log-stretch frequency-wavenumber domain Vlad and Biondi (2001) by selecting the PS-DMO operator in the log-stretch frequency-wavenumber domain introduced by Xu et al. (2001), reformulated and improved by Rosales (2002).

The PS-AMO operator has a significant future application, the regularization of ocean bottom seismic (OBS) data. The presence of already existing platforms produces holes in the data. This information can be safely regularized with an appropriate operator, in this case a PS-AMO operator.

6/7/2002