CONCLUSION

The method we have developed here has proven to be effective at removing multiple energy in the case where a simple 1-D velocity function can be assumed. The fundamental parameterization of the beam stacking operators hold promise of this method being applied to more complex multiple reverberation problems, where moveouts in the data are non-hyperbolic and lateral velocity variations exist. We hope to extend the method to 3-D and take full advantage of the information obtained from multi-streamer acquisition.