Computational and Applied Math Proseminar

Tuesday, November 29, 2005, 12:15 p.m. GWC 604

Wolfgang Stefan

Dept. Math. & Stats.

Signal Restoration through Deconvolution applied to Deep Mantle Seismic Probes

Abstract In this study we present a method of signal restoration to improve the signal to noise ratio, sharpen seismic arrival onset, and act as an empirical source deconvolution of specific seismic arrivals. Observed time series g_i are modeled as a convolution of a simpler time series f_i, and an invariant point spread function (PSF) h that accounts for the blurring effect of the wave path. The method is used on the shear wave time window containing SKS and S, whereby using a Gaussian PSF produces more impulsive, narrower, signals in the wave train. The resulting time series facilitates more accurate travel time estimation of the individual seismic arrival onsets. We demonstrate the success of the reconstruction method on synthetic seismograms, where sharper onsets of arrivals facilitate more accurate and objective determination of travel times of individual phases.

The method is also used on real data where clean and sharp reconstructions are obtained, even for signals with relatively high noise content. Some of the data highlight how this approach can be employed to reveal details of the signal that are not readily apparent in raw waveforms. In particular, we show that two phases, that traverse the lowermost mantle and arrive nearly coincident in time near 88-92 degrees in epicentral distance (S_{ab} and S_{cd}), can be identified in the deconvolved traces by a systematically broadened pulse shape.

For further information please contact: mittelmann@asu.edu