Kouvola earthquake swarm - using a cross-correlator to find very small events and cluster them

<p><span>Kouvola area, a part of the Vyborg rapakivi batholith in southeastern Finland, has been experiencing an intraplate earthquake swarm since December 2011</span><span>. T</span><span>he events hav</span><span>e</span><span> magnitudes ranging from M</span>_{<span>L</span>}<span> -1.2 to 2.8</span><span> and they happen in the uppermost two kilometers of the crust. </span><span>The </span><span>Vyborg </span><span>batholith has a long history of earthquake swarms</span><span> with macroseismic data from 1751 onwards and the first </span><span>instrumentally </span><span>recorded swarm </span><span>in </span><span>2003-2004. </span></p><p><span>Inspired by the ongoing activity</span><span>,</span><span> Institute of Seismology</span><span> of</span><span> University of Helsinki</span><span> (ISUH)</span><span> has </span><span>installed</span> <span>temporary</span> <span>seismic station</span><span>s</span><span> in </span><span>the </span><span>area </span><span>to </span><span>complement</span> <span>seismic station</span><span>s</span><span> of the Finnish National Seismic network</span><span> (FNSN)</span><span>. </span><span>The detection t</span><span>h</span><span>reshold of FNSN is </span><span>M</span>_{<span>L</span>}<span>1.0, not sufficiently low to catch the small</span><span>est </span><span>earthquakes of the swarm.</span></p><p><span>Several ta</span><span>i</span><span>lored cross-correlator</span><span>s</span><span> have been developed at the ISUH to </span><span>lower the event detection threshold</span><span>. These can</span> <span>be used to </span><span>detect even very small seismic events</span><span> well below the current FNSN detection t</span><span>h</span><span>reshold</span><span>. The method is especially well suited </span><span>to s</span><span>warm events, which </span><span>generate</span> <span>nearly identical </span><span>signal</span><span>s</span><span> due to their common origin. </span></p><p><span>O</span><span>nly the </span><span>largest</span><span> events </span><span>of the swarm </span><span>can be used to calculate focal mechanisms</span> <span>or other event parameters reliably. </span><span>One approach to </span><span>use all data is w</span><span>aveform clustering</span><span>.</span> <span>E</span><span>vent groups with identical signal</span><span> can be formed</span><span>, allowing e.g. calculation of</span> <span>composite focal mechanisms for each </span><span>event </span><span>cluster</span><span>.</span></p>