Dimensionality-reduced estimation of primaries by sparse inversion

Wave-equation based methods, such as the estimation of primaries by sparse inversion, have been successful in the mitigation of the adverse effects of surface-related multiples on seismic imaging and migration-velocity analysis. However, the reliance of these methods on multidimensional convolutions with fully sampled data exposes the ‘curse of dimensionality’, which leads to disproportional growth in computational and storage demands when moving to realistic 3D field data. To remove this fundamental impediment, we propose a dimensionality-reduction technique where the ‘data matrix’ is approximated adaptively by a randomized low-rank factorization. Compared to conventional methods, which need for each iteration passage through all data possibly requiring on-the-fly interpolation, our randomized approach has the advantage that the total number of passes is reduced to only one to three. In addition, the low-rank matrix factorization leads to considerable reductions in storage and computational costs of the matrix multiplies required by the sparse inversion. Application of the proposed method to two-dimensional synthetic and real data shows that significant performance improvements in speed and memory use are achievable at a low computational up-front cost required by the low-rank factorization.

https://slim.gatech.edu/Publications/Public/Journals/GeophysicalProspect...