14:45 〜 15:00
[SSS07-29] The 2021 Mw7.0 and Mw6.7 Miyagi-Oki earthquakes, northeastern Japan, nucleated at a seismic/aseismic transition zone in the postseismic period of the 2011 M9 Tohoku earthquake
キーワード：seismic aseismic transition、2011 M9 Tohoku earthquake、Miyagi-Oki earthquakes、Rupture process、Repeating earthquakes
After the M9 Tohoku earthquake, the interplate seismicity drastically increased in the downdip extension while disappearing within the main rupture area. The 2011 M9 Tohoku earthquake provides a rare opportunity to investigate how the earthquake cycle of a great (M9) earthquake impacts the generation of smaller (M7-8) earthquakes on the same fault. March Mw7.0 and May Mw6.7 earthquakes off Miyagi in 2021 both occurred on the downdip side of the rupture zone of the Tohoku earthquake. We investigated the regional seismicity and the source processes of these two events.
We first relocated earthquake hypocenters around the source regions using the Double-Difference method (Waldhauser & Ellsworth, 2000). We obtained the relocated hypocenters of 2736 interplate events. Only a tiny number of interplate earthquakes occurred before the Tohoku earthquake in and around the source region of the March earthquake, but the number sharply increased after the Tohoku earthquake. The hypocenter of the March Mw7.0 earthquake is located very close to the hypocenters of four M5-6 earthquakes, which started to occur after the Tohoku earthquake. The inter-event distances of these four earthquakes were much shorter than the source sizes of the four earthquakes and the mainshock, even when considering the estimation error; they probably ruptured the same seismic patch. Inside these M5-6 repeating earthquake source regions, M2-3 repeating earthquakes also occurred, forming a hierarchical structure. Almost all of these repeating earthquakes occurred after the Tohoku earthquake, suggesting that the March mainshock was initiated in a conditionally stable region where the repeating earthquake sequence emerged after the Tohoku earthquake. The initiation of an M7 earthquake in a deep conditionally-stable region can be a transient feature only possible in the early postseismic period.
We then estimated the spatiotemporal slip distributions of the two mainshocks in the same manner, following Hartzel & Heaton (1983). We used the acceleration waveform data from onland stations of NIED KiK-net and offshore stations of S-net. The results show that the March Mw7.0 mainshock had two large-slip regions: several kilometers ESE (first rupture; t=4–7 s) and ~20 km south of the hypocenter (second rupture; t=8–12 s). The rupture area of the March mainshock showed a complementary relationship with the aftershock area located between the two large slip areas. In the May mainshock, the maximum slip occurred a few kilometers southeast of the hypocenter, with the total slip area being longer in the north direction as a moderate slip propagated to the north.
The two M~7 mainshocks ruptured the westernmost part of seismic patches of the 1978 Mw7.5 Miyagi-Oki earthquake and loaded the eastern shallow seismic patches for the sequence, including the patch of the 2005 Mw7.1 Miyagi-Oki earthquake. The further updip area is a part of the main rupture area of the Tohoku earthquake and hosts almost no interplate seismicity after this earthquake. Assuming that the spatial pattern of interplate earthquakes along the Japan trench will be restored to a situation similar to that before the Tohoku earthquake in the future, the downdip seismically active area should expand to the updip area. The initial ruptures propagated in the updip direction for the 2021 M~7 earthquakes, similarly to the smaller interplate earthquakes that occurred after the Tohoku earthquake (Yoshida et al., 2022). Continued monitoring of interplate seismicity is essential to examine how creep and plate-locking evolve during the M9 earthquake cycle because previous earthquake simulation studies and observations suggest the occurrence of shallower interplate slip before the Tohoku earthquake.
We first relocated earthquake hypocenters around the source regions using the Double-Difference method (Waldhauser & Ellsworth, 2000). We obtained the relocated hypocenters of 2736 interplate events. Only a tiny number of interplate earthquakes occurred before the Tohoku earthquake in and around the source region of the March earthquake, but the number sharply increased after the Tohoku earthquake. The hypocenter of the March Mw7.0 earthquake is located very close to the hypocenters of four M5-6 earthquakes, which started to occur after the Tohoku earthquake. The inter-event distances of these four earthquakes were much shorter than the source sizes of the four earthquakes and the mainshock, even when considering the estimation error; they probably ruptured the same seismic patch. Inside these M5-6 repeating earthquake source regions, M2-3 repeating earthquakes also occurred, forming a hierarchical structure. Almost all of these repeating earthquakes occurred after the Tohoku earthquake, suggesting that the March mainshock was initiated in a conditionally stable region where the repeating earthquake sequence emerged after the Tohoku earthquake. The initiation of an M7 earthquake in a deep conditionally-stable region can be a transient feature only possible in the early postseismic period.
We then estimated the spatiotemporal slip distributions of the two mainshocks in the same manner, following Hartzel & Heaton (1983). We used the acceleration waveform data from onland stations of NIED KiK-net and offshore stations of S-net. The results show that the March Mw7.0 mainshock had two large-slip regions: several kilometers ESE (first rupture; t=4–7 s) and ~20 km south of the hypocenter (second rupture; t=8–12 s). The rupture area of the March mainshock showed a complementary relationship with the aftershock area located between the two large slip areas. In the May mainshock, the maximum slip occurred a few kilometers southeast of the hypocenter, with the total slip area being longer in the north direction as a moderate slip propagated to the north.
The two M~7 mainshocks ruptured the westernmost part of seismic patches of the 1978 Mw7.5 Miyagi-Oki earthquake and loaded the eastern shallow seismic patches for the sequence, including the patch of the 2005 Mw7.1 Miyagi-Oki earthquake. The further updip area is a part of the main rupture area of the Tohoku earthquake and hosts almost no interplate seismicity after this earthquake. Assuming that the spatial pattern of interplate earthquakes along the Japan trench will be restored to a situation similar to that before the Tohoku earthquake in the future, the downdip seismically active area should expand to the updip area. The initial ruptures propagated in the updip direction for the 2021 M~7 earthquakes, similarly to the smaller interplate earthquakes that occurred after the Tohoku earthquake (Yoshida et al., 2022). Continued monitoring of interplate seismicity is essential to examine how creep and plate-locking evolve during the M9 earthquake cycle because previous earthquake simulation studies and observations suggest the occurrence of shallower interplate slip before the Tohoku earthquake.