The correlation of seismic events on multicomponent data in the presence of coherent noise

Abstract

SUMMARY A numerical analysis has been undertaken to show the potential benefits and relevance of eigen-decomposition of covariance matrices formed using multicomponent seismic data from two stations. It is well known that a single-station covariance matrix provides polarization estimates of the particle motion at a given point in space. A two-station (binocular) covariance matrix is shown to provide similar information at two points in space, as well as indicating the correlation between the two stations. By choosing a suitable reference wavelet, correlating events at differing stations may be identified. A major advantage of the procedure is that it can successfully identify overlapping arrivals (in the time domain) when a single-station analysis breaks down completely. The binocular analysis can be sequentially applied to a multistation seismic gather to identify a correlating event. The technique is applied to synthetic gathers with differing signal-to-noise ratios and the picked events compared with a single-station analysis. It is then successfully applied to the data from a physical scale-model experiment.