Performance of the southern Korean Peninsula seismograph network and refinement of hypocentral parameters of local earthquakes, 2004–2008

• Published in: Geosciences journal (Seoul, Korea) Vol. 15; no. 1; pp. 83 - 93
• Main Authors: Park, Eun Hee, Lee, Jung Mo, Kim, Woohan
• Format: Journal Article
• Language: English
• Published: Heidelberg The Geological Society of Korea 01-03-2011; Springer Nature B.V; 한국지질과학협의회
• Subjects: Algorithms; Earth and Environmental Science; Earth Sciences; Earthquakes; Geophysics; Mathematical models; Mineral resources; Networks; Peninsulas; Seismic activity; Seismic engineering; Seismic phenomena; Seismography; Seismology; Standard deviation; Travel time; Velocity; 지질학; hypocentral parameters; information; seismograph network; Gutenberg-Richter recurrence relation
• ISSN: 1226-4806; 1598-7477
• DOI: 10.1007/s12303-011-0001-4

This study evaluated the performance of the seismograph network in the southern Korean Peninsula. The average distance between stations in this network is about 25 km. According to the Gutenberg-Richter recurrence relation, the seismograph network appears to have recorded all earthquakes larger than
M
≈ 2.1−2.2 that occurred in the 9-year interval from 2001 to 2009. Values of
a
and
b
were estimated as 4.383 and 0.8544 by linear regression, with standard deviations of 0.06 and 0.018, and as 4.251 and 0.8293 by the maximum likelihood method, with standard deviations of 0.08 and 0.049, respectively. The results agree within 1 standard deviation. The network performance was evaluated with respect to hypocenter estimation accuracy by two hypocenter determination algorithms: the recently refined GA-MHYPO algorithm and the LocSAT algorithm used by the Korea Institute of Geoscience and Mineral Resources (KIGAM). The algorithms were evaluated with synthetic traveltime data generated using the local velocity structure and the present network geometry. In view of the velocity structure of the southern Korean Peninsula and the
a priori
velocity model used by LocSAT, GA-MHYPO determines hypocentral parameters more accurately than does LocSAT by a factor of five. GA-MHYPO is able to determine epicenters of events with large azimuthal gaps, whereas LocSAT yields large errors. GA-MHYPO cannot determine the focal depths of events shallower than 10 km with the present seismograph network, but if the network had ideal 10-km spacing, GA-MHYPO could accurately determine the focal depths of 5-km-deep events. The 122 earthquakes with magnitudes larger than or equal to 2.0 that occurred from 2004 to 2008 were relocated using GA-MHYPO. The determined hypocentral parameters determined were compared with those reported by the Korea Meteorological Administration (KMA) and KIGAM. Epicenter locations differed by as much as 10 km for inland events and as much as 30 km for offshore events. This may be attributed to large azimuthal gaps in offshore events and the large differences between the
a priori
velocity models used by the KMA and KIGAM and the true offshore velocity structure. The focal depths determined here were quite different from those reported by KIGAM, and there were many events deeper than 15 km. No relationship with tectonic structure or geology was evident. The KMA and KIGAM could improve the accuracies of hypocentral parameters in their final earthquake reports by using a better
a priori
velocity model in their hypocentral determination packages.