The use of waveform cross correlation for creation of an accurate catalogue of mining explosions within the Russian platform using joint capabilities of seismic array Miknevo and IMS arrays

Thursday, 18 December 2014: 5:45 PM
Ivan Kitov1, Mikhail Rozhkov2 and Irina Sanina1, (1)Institute of Geosphere Dynamics RAS, Moscow, Russia, (2)CTBTO Preparatory Commission for the Comprehensive Nuclear Test-Ban Organization, Vienna, Austria
For seismic monitoring, the task of finding and indentifying the sources of various seismic events is getting more and more difficult when the size (magnitude, yield, energy) of these events decreases. Firstly, the number of seismic events dramatically increases with falling magnitude - approximately by an order of magnitude per unit of seismic magnitude. Secondly, mining explosions become detectable and represent one of the biggest challenges for monitoring for magnitudes below 3.5 to 4.0. In the current study of mining activity within the Russian platform, we use the advantages of location and historical bulletins/catalogues of mining explosions recorded by small-aperture seismic array Mikhnevo (MHVAR) and extensive data from several IMS arrays at regional and far regional distances from the studied area. The Institute of Geosphere Dynamics (IDG) of the Russian Academy of Sciences runs seismic array MHVAR (54.950 N; 37.767 E) since 2004. Approximately 50 areas with different levels of mining activity have been identified by MHVAR and reported in the IDG catalogue as mining events. Signals from select mining events detected by MHVAR are sought at IMS arrays. Continuous data from MHVAR and IMS arrays (e.g. AKASG) are processed jointly using waveform cross correlation technique. This technique allows reducing the detection threshold of repeated events by an order of magnitude as well as accurately locating and identifying mining explosions. To achieve the highest performance of cross correlation, we have selected the best sets of waveform templates recorded from a carefully tested set of master events for each of the studied mines. We also test the possibility to use the Principal and Independent Component Analysis to produce sets of synthetic templates, which best fit the whole set of master events for a given mine.