Two-parameters residual-moveout analysis for wave-equation migration velocity analysis

Discussion and conclusions

The introduction of a second term to the conventional RMO function for angle-domain CIGs improves the flatness of the corrected gathers for both CIGs I used in my testing and for both choices of two-parameter RMO function I proposed.

The answer to the question of whether using a two-parameter RMO function yields more reliable gradients when applied in automatic MVA methods is more ambiguous. The correlation analysis I presented indicates that the ``Taylor'' RMO function yields more robust gradients than the simple one-parameter RMO function for both CIGs I used as representative of situations when either strong lateral velocity variations or anisotropy occur. The impact of these improvements in real situation is difficult to predict. More testing and analysis are needed to determine whether the additional computation and code complexity introduced by the addition of a second term to the RMO function are worthy.

The ``Orthogonal'' RMO function may yields better gradients, but it is also more sensitive with respect to the thickness of the depth-averaging window for the power spectra. This fragility is caused by the location of the extremum of the second term of the ``Orthogonal'' RMO function in the middle of the angular range. Although the ``Orthogonal'' function has some theoretical advantages, its lack of robustness make it a less desirable choice.

Two-parameters residual-moveout analysis for wave-equation migration velocity analysis