[S01-1] Open session I

[S01-1-01] Automatic hypocenter determination for the Seismological Bulletin of Japan using Bayesian estimation and its applications

^○Koji Tamaribuchi (Meteorological Research Institute, Tsukuba, Japan)

A quick estimation of spatial and temporal hypocenter distributions in swarms and aftershocks is essential for estimating the source geometry and possible damaged area. The automatic hypocenter determination method is important to understand seismic activities in real time. We developed a new method using Bayesian estimation to identify multiple concurrent earthquakes to be used for the Seismological Bulletin of Japan. This method makes integrated use of P- and S-wave arrival times and maximum amplitude. We applied this method to earthquakes in the vicinity of Japan. The ratio of detected earthquakes compared with those in the manually checked event list is almost 100% (M>=1.0) at inland and shallow areas. At offshore or deep areas, it is about 80% (M>=1.0). We also applied this method to the aftershock activity of the northern Nagano Prefecture on November 22, 2014. In this case, this method could detect more than 1,700 events in 24 hours, although the previous automatic determination system could detect only 250 events at that time. This indicates that our method is useful to understand seismic activities in real time.
The new seismic monitoring system using this method have started its operation at the Japan Meteorological Agency in April 2017. In the 2017 Kumamoto earthquake, which began on April 14, the ability of the method was fully demonstrated. By the end of May, number of hypocenter determinations reached about 70,000, which was far more than the level possible by manual processing, and the system played an important role to grasp the spatio-temporal characteristics of the activity.