Real-time seismic event detection and processing is of interest of seismologist as no
human interference is needed in processing large amounts of incoming data that cannot
be reviewed manually by an analyst. This paper focuses on event detection using Fisher
statistics, i.e. working with a seismic array and using coherency between the signals recorded
at the different stations. After detection, further processing using all 3 components of the
seismic signal are used in order to discriminate between different types of polarization. The
back-azimuth and inclination as well as a indication of the ﬁrst S-arrival are resolved by
solving the eigenproblem of the covariance matrix. The software is tested on data from
explosive shots and 2 regional small scale earthquakes from the Northern-Netherlands. The
Fisher-statistics are able to detect incoming signals, even if the signal-to noise ratio becomes
very low. The success of analyzing the polarization depends on the signal-to-noise ratio.
The results for the explosive shots are extremely accurate, whereas the performance for the
real earthquakes is dependent on the magnitude. A quake with M = 3.5 is well resolved,
but an event with a magnitude of 2.1 leads to less favorable results.
GJ van den Hazel. Detection and processing 3C signal from a small scale seismic array
published, MSc thesis Utrecht University, 2008