1:45 PM - 2:00 PM
[J04-3-02] Real-time prediction of ground shaking without source information: Data assimilation and simulation of seismic wave propagation for Earthquake Early Warning
Many of the present Earthquake Early Warning (EEW) systems quickly determine the hypocenter and magnitude, and then predict strength of ground motions. The Mw 9.0 Tohoku earthquake, however, revealed some technical issues with such methods: under-prediction at large distances due to the large extent of the fault rupture, and over-prediction because the system was confused by multiple aftershocks that occurred simultaneously. To address these issues, we have proposed a new concept for EEW (Hoshiba and Aoki, 2015, BSSA), in which the present wavefield is estimated precisely in real time by applying data assimilation, and then the future wavefield is predicted by simulation of seismic wave propagation. Information on the hypocenter and magnitude is not required. We call it “numerical shake prediction" by analogy to “numerical weather prediction" in meteorology.
In many methods of the present EEW systems, the strength of ground motions (PGA, PGV, seismic intensity) are predicted using the hypocentral distance and magnitude based on a ground motion prediction equation, which usually leads the prediction of concentric distribution. However, actual ground shaking is not always concentric, even when site amplification is corrected. The strength of shaking may be much different among earthquakes even when their hypocentral distances and magnitudes are almost the same. For some cases, PGA differs more than 10 times, which leads to imprecise prediction in EEW.
In numerical shake prediction, because future is predicted from the present condition, it is possible to address the issue of the non-concentric distribution. Once the heterogeneous distribution is actually observed in ongoing wavefield, future distribution is predicted accordingly to be non-concentric. We will indicate examples of M 6 crustal earthquakes occurred at central Japan, and the 2016 Kumamoto earthquake (Mw 7.0), during which M 6 class earthquake was remotely triggered apart from 70 km from the epicenter.
In many methods of the present EEW systems, the strength of ground motions (PGA, PGV, seismic intensity) are predicted using the hypocentral distance and magnitude based on a ground motion prediction equation, which usually leads the prediction of concentric distribution. However, actual ground shaking is not always concentric, even when site amplification is corrected. The strength of shaking may be much different among earthquakes even when their hypocentral distances and magnitudes are almost the same. For some cases, PGA differs more than 10 times, which leads to imprecise prediction in EEW.
In numerical shake prediction, because future is predicted from the present condition, it is possible to address the issue of the non-concentric distribution. Once the heterogeneous distribution is actually observed in ongoing wavefield, future distribution is predicted accordingly to be non-concentric. We will indicate examples of M 6 crustal earthquakes occurred at central Japan, and the 2016 Kumamoto earthquake (Mw 7.0), during which M 6 class earthquake was remotely triggered apart from 70 km from the epicenter.