9:45 AM - 10:00 AM
[SSS10-04] Fault structure of intense normal-fault earthquake activity in the upper plate off Fukushima and Ibaraki, Japan
Accurate hypocenter distribution of small earthquakes offers essential information about the structure and the stress state of faults. However, detailed hypocenter distribution with good depth accuracy in offshore areas is usually unavailable because seismic stations are basically distributed only on the land. Along the Japan Trench, the focal depths of offshore earthquakes can be estimated with good accuracy by using P and S arrival times at offshore S-net stations since 2016. However, the accurate depths of earthquakes in earlier periods remain challenging to be obtained.
Since the 2011 Mw 9.0 Tohoku-Oki earthquake, normal-faulting earthquakes have intensely occurred off the east coast of northeastern Japan. According to the JMA hypocenters, many of these events appear to occur within a zone that tilts to the east, but this feature may be due to systematic errors in the depth of the hypocenters. In fact, according to the MT catalog by NIED F-net, many of the normal faulting events are concentrated near the surface and does not show the distinct tilting feature. To understand actual fault structures and stress state in this area, it is important to know accurate hypocenter distribution.
The arrival time data of the depth phase can constrain the focal depths of earthquakes far from the observation network. For offshore earthquakes in northeastern Japan, sP depth-phases are frequently observed in seismographs with an epicentral distance of of 50 km to 200 km (Umino et al., 1995). We here use the sP depth-phase to systematically estimate the focal depths of shallow offshore earthquakes. We pay special attention to the area off Fukushima-Ibaraki, where normal-faulting earthquakes have occurred most frequently after the Tohoku-Oki earthquake. While previous studies in this area used the arrival times of depth-phase obtained from manual picking, we use a method based on waveform modeling to utilize the depth-phase in an objective and efficient manner.
First, we visually examined the observed waveforms of earthquakes in the offshore region at depths shallower than 40km. Many of the seismograms showed distinct later phase arrivals between the P- and the S-phase arrivals. The amplitudes of the later phases were dominated in the vertical component, and the delay times from the direct P-phases were about the same at stations located at different epicentral distances. We calculated a synthetic waveform at each seismic station using the code of Zhu & Rivera (2002) and the focal mechanism from the F-net MT solution. The synthetic waves showed that the sP phase arrives at a time close to the observed later phase, suggesting that the observed phase represents the sP phase.
We used these sP phases to constrain the depths of the 1,297 offshore events of 3.2 ≦ Mw ≦ 5, which occuerred from Mar. 2003 to Mar. 2022. Only waveforms obtained at land stations were used as they were continuously available from before the Tohoku-Oki Earthquake.
We obtained the focal depth at which the synthetic waveforms best explained the timings of observed sP phases based on the cross-correlation coefficient. By combining this method with the double-difference method (Waldhauser & Ellsworth, 2000), we relocated the hypocenters of earthquakes whose depths could not be stably determined only by the method using sP phase.
The obtained focal depths tended to be about 10 km shallower than those in the JMA unified catalogue and were systematically deeper than those by the F-net solution. On the other hand, our depth estimates agree with the focal depths determined using S-net data since 2016. Our results show that normal-faulting earthquakes were mainly distributed along an eastward dipping zone. These results suggest that this east dipping structure seen in the JMA hypocenters is not an artifact due to insufficient accuracy of the hypocenter location but is probably a significant feature related to regional stress, strength, fluid or temperature distribution.
Since the 2011 Mw 9.0 Tohoku-Oki earthquake, normal-faulting earthquakes have intensely occurred off the east coast of northeastern Japan. According to the JMA hypocenters, many of these events appear to occur within a zone that tilts to the east, but this feature may be due to systematic errors in the depth of the hypocenters. In fact, according to the MT catalog by NIED F-net, many of the normal faulting events are concentrated near the surface and does not show the distinct tilting feature. To understand actual fault structures and stress state in this area, it is important to know accurate hypocenter distribution.
The arrival time data of the depth phase can constrain the focal depths of earthquakes far from the observation network. For offshore earthquakes in northeastern Japan, sP depth-phases are frequently observed in seismographs with an epicentral distance of of 50 km to 200 km (Umino et al., 1995). We here use the sP depth-phase to systematically estimate the focal depths of shallow offshore earthquakes. We pay special attention to the area off Fukushima-Ibaraki, where normal-faulting earthquakes have occurred most frequently after the Tohoku-Oki earthquake. While previous studies in this area used the arrival times of depth-phase obtained from manual picking, we use a method based on waveform modeling to utilize the depth-phase in an objective and efficient manner.
First, we visually examined the observed waveforms of earthquakes in the offshore region at depths shallower than 40km. Many of the seismograms showed distinct later phase arrivals between the P- and the S-phase arrivals. The amplitudes of the later phases were dominated in the vertical component, and the delay times from the direct P-phases were about the same at stations located at different epicentral distances. We calculated a synthetic waveform at each seismic station using the code of Zhu & Rivera (2002) and the focal mechanism from the F-net MT solution. The synthetic waves showed that the sP phase arrives at a time close to the observed later phase, suggesting that the observed phase represents the sP phase.
We used these sP phases to constrain the depths of the 1,297 offshore events of 3.2 ≦ Mw ≦ 5, which occuerred from Mar. 2003 to Mar. 2022. Only waveforms obtained at land stations were used as they were continuously available from before the Tohoku-Oki Earthquake.
We obtained the focal depth at which the synthetic waveforms best explained the timings of observed sP phases based on the cross-correlation coefficient. By combining this method with the double-difference method (Waldhauser & Ellsworth, 2000), we relocated the hypocenters of earthquakes whose depths could not be stably determined only by the method using sP phase.
The obtained focal depths tended to be about 10 km shallower than those in the JMA unified catalogue and were systematically deeper than those by the F-net solution. On the other hand, our depth estimates agree with the focal depths determined using S-net data since 2016. Our results show that normal-faulting earthquakes were mainly distributed along an eastward dipping zone. These results suggest that this east dipping structure seen in the JMA hypocenters is not an artifact due to insufficient accuracy of the hypocenter location but is probably a significant feature related to regional stress, strength, fluid or temperature distribution.