Separation of P–P and P–SV wavefields by high resolution parabolic Radon transform

Abstract

In three-component seismic reflection data, P–P and P–SV wavefield separation is an important issue which has great influence on interpretation and imaging. Under the assumption that P–P and P–SV wave reflections can approximately be regarded as parabolas when the offset is less than the depth, a P–P and P–SV wavefield separation method based on high resolution parabolic Radon transform is presented in this paper. By mapping three-component seismic data in the Radon model, P–P and P–SV waves separate from each other, after which separated wavefields are reconstructed by muting P–P and P–SV waves respectively in the transform domain. As the separation process relies on the focusing properties of both wave events in the Radon domain, the increased focusing power of the high resolution parabolic Radon transform enables separating P–P and P–SV wavefields robustly. Whereas the smearing problem of the conventional least squares parabolic Radon transform leads to poor separation suffering from residual energy, and the separating results by τ–p transform exhibits apparent amplitude distortion caused by event overlapping in the τ–p domain.