An Efficient 4D Processing Flow for Variable-depth Streamer Data

We present a new processing algorithm to co-datum two or more streamer datasets acquired with different cable shapes to a common depth. This method is applied early on in a time-lapse processing sequence so that standard 4D co-processing, such as 4D QCs and 4D binning, can easily be performed. We also show how we can match the spectra of a legacy baseline with flat tow and a monitor data acquired with variable streamer depth, such that they both have minimal ghost content. Applying this technique to a 2013 North Sea dataset acquired several times without time lag, we find that the 4D difference on maximally de-ghosted data is better than when the broadband data is downgraded to the legacy data. The deghosting technique, which we call Ghost Wavefield Elimination (GWE), can be used to remove as much of the ghost in the legacy data as is achievable, dependent on the signal-to-noise ratio of the legacy data in and around the ghost notches. In our case study we achieve excellent 4D repeatability with around 10% background noise (NRMS).