Application of cross-correlation of angle stacks for velocity quality control in seismic imaging

Willis, Allan (Spectrum Geo Inc.) | Busono, Pudjo (Spectrum Geo Inc.)


Contretras et al. (2005) briefly mentioned time alignment of partial-angle stacks for correction of residual NMO, and Araman et al. (2014) talked about evaluating the overall quality of the pre-stack gather flatness via estimates of gross time misalignment between angle stacks. Angle stack similarity QC Evaluation of pre-stack data consistency, as proposed by Paternoster et al. (2009), can be performed by crosscorrelating pairs of sub-stacks at a finer sampling rate across a long enough time interval. In addition, Coleous et al. (2013) and Araman et al. (2014) both pointed out that correlation between angle stacks over an objective-defined window (e.g. at the key horizon or reservoir level) is a commonly used and quantitative QC method for AVA/AVO or seismic inversion studies. When near-and far-angle stacks are cross-correlated, areas where seismic events line up nearly perfectly with minimum or no shifts between the angle stacks would have zero or nearly zero lag times on the cross-correlation calculation. High cross-correlation peak lag times would be indicative of misfit in seismic events between the angle stacks, and this would represent non-flatness of seismic gathers and hence a non-optimal velocity model. Thus, this is the basis for using the cross-correlation as a global QC of optimal velocity model or image-focusing quality.