Rough sea surface implications on receiver deghosting

Efficient removal of ghost events is the key to achieving broadband seismic data. On the receiver-side, accurate ghost removal can be performed when both pressure and vertical particle velocity information is available. However, deghosting pressure-only measurements by spectral division with a flat sea surface or a statistical ghost function is still commonly applied. In this paper, we evaluate the performance of wavefield separation and deghosting with a flatsea surface (or statistical) ghost function for a rough weather marine seismic data acquisition. The behaviour of the pressure ghost function under rough sea conditions is analysed in comparison to the flat and the statistical pressure ghost functions. The three pressure ghost functions (i.e., true, flat, and statistical) converge to the same amplitude and phase values for the lower frequencies, but they diverge from each other for higher frequencies. The error created by wavefield separation and deghosting by spectral division is quantified, using synthetic and real seismic data. Wavefield separation provides deterministic and full receiver-side deghosted up-going wavefields, while both flat and statistical deghosting methods result in significant errors with implications in pre- and post-stack evaluation.