Suppressing Ground Roll, P-S Conversions and Out-of-Plane Scattering in Near Surface Seismic Data Using 3C Polarization

Multiple approaches have been employed in the past to filter ground roll in near surface seismic data. Although many seismic surveys nowadays are currently recorded with three-component (3C) receivers, older filters are usually designed for one-component (1C) data. Invoking two additional spatial dimensions with orthogonal components of particle motion should enhance the separation and thus suppression of ground roll from reflection energy. The aim of this study is to analyse and eliminate not only source induced ground roll but also P-S conversions and out-of-plane scattered wavefield energy based on their respective 3C polarisation by means of two approaches. The method is tested on a field dataset collected on the campus of Utrecht University, the Netherlands, for the purpose of investigating the subsurface for shallow geothermal heat storage. In the noisy field data the polarization analysis is moderately effective and could resolve most individual arrivals and remove ground roll, P-S conversions and out-of-plane scattered wavefield energy. Besides this the linearity method can provide valuable information about the nature of individual P- and S-wave events if the body waves are polarized along a specific spatial axis.