Generalized semblance and multi-CMP traveltime parameter estimation using Singular Value Decomposition

This work introduces a novel approach using a multiparametric equation specifically applied within the multi-CMP domain. Our method leverages SVD semblance, capturing the most coherent signal in the first eigenimage. This enables the determination of 2D stacking parameters for a set of CMP families, including velocity and dip of events between lateral CMPs, through analysis akin to conventional velocity analysis. Unlike traditional methods employed in CRS and CDS, such as global and local optimization strategies, our approach does not require an initial velocity model or additional parameters. CRS methods typically involve three parameters in 2D, while CDS methods may require more. In contrast, our method focuses on key parameters like velocity and dip, reducing complexity while maintaining accuracy. The use of the power method results in a reduced computational cost without sacrificing precision in stacking parameter estimation. The method was applied to both synthetic data with varying noise levels and real data from the Jequitinhonha and Camamu Basins, located along the Brazilian passive margin. The results were successful in enhancing the signal-to-noise ratio and improving the continuity of reflectors, demonstrating the robustness and practical utility of our approach. • A new multi-CMP stacking method based on SVD semblance is introduced. • The method estimates stacking parameters without global optimization. • SVD semblance improves velocity and dip resolution over classical semblance. • The approach was validated with synthetic and real seismic data. • Results show higher coherence values and better signal-to-noise ratio.