Normal moveout (NMO) correction is an important data processing step for VSP CDP transform and other applications such as estimating the seismic velocity below borehole receivers and regularizing/interpolating source points. Most of the previously used VSP NMO correction methods are valid for isotropic media, but they can cause significant errors in multiple layered and/or anisotropic (VTI) velocity media. We present here a new methodology to accurately and efficiently perform VSP anisotropic NMO correction for walkaway VSP (WVSP) or 3D VSP data. We adapt the Fomel and Stovas' NMO correction formulation, a method originally developed for surface seismic data, to NMO correction for WVSP or 3D VSP data with large source offsets in VTI media. We employ a global optimization algorithm called simulated annealing (SA) to automatically search for the four parameters in the Fomel and Stovas' formulation. In the SA algorithm, we use semblance coherence analysis in the VSP common receiver gather domain as an object cost function. That is, after a right NMO correction, a reflection event is aligned and its semblance coherence coefficient is maximized. To compress/attenuate the coherent and incoherent noise in the upgoing reflection wavefield of VSP data, we use a three-step processing flow consisting of NMO correction, median filtering, and reverse NMO correction. We demonstrate the application of our new methodology with a field WVSP dataset. The field data example shows a significant improvement in the image quality of the VSP data after attenuating noise by this new methodology.
Presentation Date: Tuesday, September 26, 2017
Start Time: 11:25 AM
Location: 371A
Presentation Type: ORAL