Windowing of dispersive noise in two domains

Abstract

During the collection of seismic data on land, various types of near-surface horizontally propagating waves are generated by the seismic source. These waves are more or less dispersed depending on the elastic properties of the surface layers and can be of sufficient amplitude to cause a difficult coherent noise problem. A process for suppressing noise arising from dispersive direct waves in presented. It is most applicable to data for which such noise severely masks the signal and where conventional velocity filtering alone gives insufficient signal-to-noise (S/N) enhancement. The method involves signal compression and windowing and has been applied successfully to flexural ice-waves and to dispersive ground roll.