Global detection and localization of seismic sources by using beamforming of multiple body wave phases with USArray

Abstract:

Detection methods are usually developed to observe earthquakes, and are not relevant to observe long event with emergent signals (e. g. event with long source duration). We present a new method to detect and localize seismic events without prior information about their source. This method explores the consistency and characteristic behavior of teleseismic body waves recorded by a large-scale seismic network. We show that the use of a seismic network as an antenna is a powerful tool to analyze sources without the need to pick phases arrivals. This allows the characterization of low amplitude events that compose the noise.

The procedure consists of three steps. First, for every tested source location we perform a time-slowness analysis and compute the Tau-p transform from the dataset. For waves emitted by teleseismic sources, the amplitude of this transform has a very characteristic behavior with maxima corresponding to different seismic phases arrivals. Relative location of these maxima on the time-slowness plane strongly depends on the distance to the earthquake. In a second step, we convolve the Tau-P amplitude with a time-slowness filter whose maxima are computed based on prediction of global travel-time calculator (Buland and Chapman, 1983) in order to explore this dependence. As a third step we gather the results obtained with different sources to get a space/time likelihood function for the occurrence of a seismic event. This process is performed at different frequency bands to observe possible variations in time.

We apply this method to continuous vertical-component seismograms of USArray. We highlight non earthquake events that occurred during 2010. We then compare our results with datasets of stations closer to the events and a numerical model for ocean low frequency noise. We identify several low frequency microseisms occurring all along the year.