Time-jittered marine acquisition–-a rank-minimization approach for 5D source separation

Simultaneous source marine acquisition has been recognized as an economic way of improving spatial sampling and speedup acquisition time, where a single- (or multiple-) source vessel fires at jittered source locations and time instances. Consequently, the acquired simultaneous data volume is processed to separate the overlapping shot records resulting in densely sampled data volume. It has been shown in the past that the simultaneous source acquisition design and source separation process can be setup as a compressed sensing problem, where conventional seismic data is reconstructed from simultaneous data via a sparsity-promoting optimization formulation. While the recovery quality of separated data is reasonably well, the recovery process can be computationally expensive due to transform-domain redundancy. In this paper, we present a computationally tractable rank-minimization algorithm to separate simultaneous data volumes. The proposed algorithm is suitable for large-scale seismic data, since it avoids singular-value decompositions and uses a low-rank based factorized formulation instead. Results are illustrated for simulations of simultaneous time-jittered continuous recording for a 3D ocean-bottom cable survey.

https://slim.gatech.edu/Publications/Public/Conferences/SEG/2016/kumar20...