oa Adaptive Surface Waves Attenuation from Wide-Azimuth Land Data

Attenuation of Surface Waves is the first show stopper that is faced in land data processing. However,<br>the characteristics of Surface Waves, Groundroll and Guided Waves, can be extracted from the input<br>data themselves to feed a cascaded adaptive filtering in order to remove these troublesome noises.<br>This approach is done in the FX domain to model the signal and noise. The part of the model<br>corresponding to the noise is then subtracted from the data using a least squares approach. The<br>benefit of such an approach is to wipe-out the Surface waves recorded on raw surface seismic data.<br>This constitutes a technical breakthrough for many difficult areas especially for modern surveys with<br>Wide-Azimuth acquisition.<br>The first issues that should be addressed in land processing are the attenuation of Groundroll (GR) and<br>Guided Waves (GW). Both are Surface Waves recorded on a vertical geophone and are the result of<br>interfering P and SV waves that travel along or near the ground surface. GR is characterized by low<br>velocity, low frequency and high amplitude and can be strongly dispersive and aliased. GW are visible<br>on records as repeated linear arrivals on the longer offsets due to multiple refractions and/or converted<br>refractions. GW are generated by some specific sub-surface geological conditions such as fast-slow-fast<br>velocity interfaces at depth and these pollute the long offsets of data with high amplitudes, fast<br>velocity and higher frequency linear noises than GR but with weaker dispersion. A cascaded application<br>of an adaptive noise attenuation taking into account the GR and GW characteristics can fully benefit<br>Wide-Azimuth acquisition and will have a major impact on the S/N ratio and stack response. More of<br>the acquired data can be utilized to create the stack (e.g. incidence angle might be increased from<br>25/30° to 40/45°) and this has many positive benefits such as increased accuracy of velocity<br>interpretation, multiple attenuation, increase in S/N ratio and improved stack response. This paper<br>focuses on a data driven approach that performs 3D cascaded adaptive filtering of aliased and<br>dispersive Surface Waves at their true spatial coordinates (AGORA GR & GW) and will be described<br>using both Narrow and Wide-Azimuth land data examples. <br>Acknowledgements<br>We would like to thank Petroleum Development of Oman and CGGVeritas for their permission to<br>publish this paper.