3D velocity-independent elliptically-anisotropic moveout correction |

Measuring local slopes is a thorough and azimuthally flexible way to characterize traveltime surface geometry, which, in the 3D case, provides useful information about azimuthal variations in moveout velocity. We have demonstrated an application for this feature in performing an elliptically anisotropic moveout correction in 3D. No velocities are picked in order to perform this moveout correction, and since we use plane-wave destruction filters to measure local slopes, the entire process is automated. Local moveout velocities can be calculated as a function of the local slopes, and the azimuthal angle of anisotropy can be estimated locally if one measures the first mixed-derivative of the traveltime surfaces at each point. By recording the traveltime shifts applied by our automated method, we formulate a highly overdetermined linear system to solve for moveout parameters as a function of time. This inversion scheme was shown to be very accurate on a synthetic data example, but remains to be tested on field data. The only practical limitation in the synthetic example comes from steeply dipping events which introduce aliased slope measurements. Although this type of aliasing can be mitigated to some extent with additional processing, further testing will be required on typical field geometries with relatively coarse crossline spacing.

Even in multi-layer cases, where conflicting azimuthal anisotropies are present, the proposed moveout correction itself can be performed accurately and automatically without velocity or parameter estimation. In these cases, the effective moveout parameters can be estimated from our method. Extensions of this method following an iterative scheme analogous to a layer-stripping or Dix-type inversion strategy may provide a powerful option to automatically recover interval parameters as well.

3D velocity-independent elliptically-anisotropic moveout correction |

2013-07-26