Multisource least-squares migration of marine streamer data with frequency-division encoding

The computational cost of conventional migration can be reduced by multisource migration of supergathers, and the accompanying crosstalk noise can be reduced through least-squares inversion. This approach is not directly applicable to marine streamer data, however, due to the mismatch between the limited number of live traces/shot recorded in the field and the pervasive number of traces generated by computer simulation. This leads to a strong mismatch in the misfit function, and results in strong artifacts (crosstalk) in the migration image. To eliminate this noise, we present a frequency division multiplexing (FDM) strategy with iterative least-squares migration (ILSM) of supergathers. The key idea is, at each ILSM iteration, to assign a unique frequency band to each shot gather. This assignment ensures that there is no spectral overlap among shots, and therefore their crosstalk is zero. Our results in applying this method to 2D marine data for the SEG/EAGE salt model show better resolved images than standard migration computed at about 1/10 the cost. We conclude that multisource migration for a marine geometry can be successfully achieved by a frequency-division encoding strategy, as long as crosstalk-prone sources are segregated in their spectral content. This is both the strength and the potential limitation of this method.