Subduction off Miyagi and Fukushima prefectures, northern Honshu, has been recognized as having a triple-planed structure of seismicity at the deep thrust zone in the 40-60 km depth range. This triple seismic zone is composed of thrust-type earthquakes, down-dip compressional and down-dip tensional earthquakes from top to bottom. At the time of the 2011 Tohoku-oki earthquake, peculiar phenomena such as radiation of short-period seismic waves and pre- and after- slow slips within the asperities of M7 class earthquakes were observed over this thrust zone west of the 2011 main rupture zone. Further to the south, where the Philippine Sea plate is subducting beneath Kanto, a triple seismic zone has also been recognized particularly under southwest Ibaraki prefecture. The thrust-type earthquakes at the top of the triple seismic zone off northern Honshu and beneath Kanto have been believed to be interplate events representing the relative motion between the overriding and subducting plates. I conclude that the thrust-type earthquakes beneath southwest Ibaraki prefecture are in fact within the crust of the subducting Philippine Sea slab, not at the surface because their slip vectors are different from the relative motion between the subducting and overriding plates. Therefore, there would be an aseismic plate boundary above the seismicity. I also show that the dip angles of the westerly dipping fault planes of the thrust-type earthquakes off Miyagi prefecture are smaller by 6° in average than the dip of the slab surface in this region, except for the six years prior to the Tohoku-oki earthquake, i.e. prior to 2006. Furthermore, the slip vectors coincide with the relative motion between the overriding and subducting plates only during this period. I infer that the topmost earthquakes of the triple seismic zone off Miyagi prefecture prior to 2006 are thus likely to be within the crust of the subducting plate. The slow slips before and after the Tohoku-oki earthquake would have occurred not within the asperities but along the aseismic plate interface, and the short-period seismic waves would have been radiated due to fractures within the crust associated with the overshooting rupture at the time of the Tohoku-oki earthquake. Many of the so-called repeating earthquakes at the topmost surface of the subducting plate would be in fact intra-crustal events within the slab. M9 earthquakes would interact with the triple seismic zone, not only mechanically, but also through fluid migration, because earthquakes in the triple seismic zone involve dehydration reaction. The irregularity of the occurrence of M9 earthquakes might be due to the inhomogeneous distribution of hydrated minerals in the incoming plate. The subduction zones having M9 earthquakes or under Kanto have a collisional character. I propose to term subduction having both a collisional character and a triple seismic zone as “super-subduction”. The relative motion between the plates is accommodated by the deformations of the crust of the subducting slab as a “plate boundary zone”. The viewpoint of “super-subduction” is necessary to understand earthquakes in subduction zones with a collisional character and dehydration reactions in the slab.

On the basis of statistical (soft) clustering of a horizontal displacement field from Global Navigation Satellite System, we conduct tectonic classification around the Izu Peninsula, a collisional zone in Japan. We determine the best number of the clusters by a statistical index, and classify 44 observation points into “Fuji cluster”, “Southern Izu cluster”, “Sagami cluster”, or “Oshima cluster”. The boundaries between the clusters almost correspond to locations of recent large earthquakes. Although several points differ from a recent concept of the Izu microplate, we obtain the following perspectives: (1) Northern part of the Izu Peninsula is integrated with Honshu, the main island of Japan. (2) Multiple cluster boundaries merge around Izu Tobu Volcano Group. (3) Fault zones of landward extension of subduction trenches (the Suruga Trough and the Sagami Trough) do not correspond to the cluster boundaries.

Broadband ocean bottom seismometers (BBOBSs) have been developed since 1990s and put into practice to explore the structure of the Earth’s interior beneath oceanic regions, e.g., mid-oceanic ridges, subduction zones, hotspots, and oceanic lithosphere-asthenosphere boundary. The BBOBSs have been used to investigate earthquake source process including slow earthquakes and oceanographical phenomena. The best way for the broadband seismic observation in the oceanic region seems that the borehole seismic observatory connected with the ocean floor cable, which is realized in several networks deployed close to the coast but it is rare even now. For scientific targets far away from the coast, we still need to develop and operate the autonomous system of BBOBS with better S/N for a temporary observation for focused scientific targets. In this review article, focusing mainly on achievements made in Japan, we describe a history of the development of BBOBS and some related instruments, scientific results based on observations with the BBOBSs, and future direction of broadband ocean bottom seismology.

The Chubu and the Kinki regions in central Japan were widely and seriously damaged by the 1586 Tensho earthquake. The detail of this event is not well understood because reliable descriptions in historical documents for Tensho era are few. Previous researches have provided wide variety of arguments about location of source fault and magnitude of this event, which yield several differing interpretations. In this study, we focus on ground liquefaction which has been not intensively investigated yet for this event. The four areas under consideration involve liquefaction sites discovered in the ruins or inferred from the reliable descriptions. These are the Tonami Plain, Toyama Plain, Nobi Plain and north-east coast of Lake Biwa. We attempt to estimate source fault of this event by evaluating the possibility of liquefaction to seismic ground motion assuming hypothetical source faults. Hypothetical source faults are along the Shokawa fault group, southern part of the Atera fault group and the Yoro-Kuwana-Yokkaichi faults. The possibility of liquefaction is evaluated by P_L value which is calculated using predicted seismic intensity and borehole data (including N-value and soil data). Predicted seismic intensity is calculated using Estimation Tools for Earthquake Ground Motion by empirical attenuation relations of J-SHIS. Borehole data from 296 boring sites in public database is used. As a result, P_L value is large in all these areas in case of occurrence a single earthquake of M_j 7.9 on the Shokawa fault group. It means that the event likely cause liquefaction and can explain liquefaction in all these areas.

The M_J 6.1 Northern Osaka earthquake of 18 June, 2018, is a crustal earthquake that occurred near the Osaka metropolitan area. While moment tensor solutions and several previous studies indicate that this earthquake involved the rupture of two fault planes, one having a predominantly dip-slip mechanism and the other having a predominantly strike-slip mechanism, contribution of each rupture to observed strong motions has not been sufficiently studied. In this study, we aimed at investigating the generation process of the strong ground motions during the earthquake and modeling the process with the pseudo point-source model. In the pseudo point-source model, the Fourier amplitude of strong ground motions is expressed as the multiplication of source, path, and site effects, where source spectra following the omega square model and empirically evaluated site amplification factors are used. In addition, Fourier phase characteristics of weak motion records are used. The model has shown great applicability to megathrust and intra-slab earthquakes despite its simplified source description, however, its applicability to crustal earthquakes has not been fully studied. Therefore, investigating its applicability to a crustal earthquake was among our objectives. First, the rupture process of the earthquake was estimated with the inversion of strong motion records using empirical Green’s functions. Reverse and strike-slip faults were used following a moment tensor solution. We found that slip was concentrated around the hypocenter on the reverse fault and slightly above the hypocenter on the strike-slip fault. Seismic moment was distributed almost evenly to both faults. These results indicated that up-dip rupture propagation caused the forward directivity effect around the epicenter and that the rupture on both faults contributed to strong motions. Then two types of pseudo point-source models were constructed. One involved only one subevent and the other involved two subevents. Either the reverse or the right-lateral fault was used for the one-subevent models and both the faults were used in the two-subevent model. The one-subevent model well captured the characteristics of observed strong motions in terms of acceleration Fourier spectra and velocity waveforms, indicating excellent applicability of the pseudo point-source model to this crustal earthquake. The two-subevent model did not lead to evident improvement of the results because the result for the one-subevent model was already satisfactory. This result does not necessarily mean that only one fault plane contributed to the generation of strong ground motions during the earthquake. Instead, this indicates that the contributions from two faults were not separated from each other on the time axis and the ground motion could be well explained with one subevent by the pseudo point-source model. The effect of forward directivity was generally small for this earthquake, however, at OSK002 just above the fault, the Fourier spectrum was underestimated at high frequencies between 2-6 Hz, which can be attributed to the up-dip rupture propagation of the strike-slip fault. Our future works include introducing a variable corner frequency model accounting for directivity effect to the pseudo point-source model and to see the model can improve the results.

This paper describes that Iwami-Tatamigaura was uplifted during the 1872 Hamada earthquake based on the analysis of the height and ages of emerged sessile assemblage, and topographic data. In two sea caves located in the northern part of the survey area, we observed the emerged sessile assemblage, which consists mainly of calcareous tubeworm, Pomatoleios kraussii, and collected samples for ¹⁴C dating and measured their altitude referring against the Tokyo Peil (T.P.). Two levels of assemblage distributed at 1.29 to 1.78 m (T.P.) in the upper and 0.90 to 1.28 m (T.P.) in the lower can be recognized in the East Cave. Analyzing the calibrated ¹⁴C age data of the lower assemblage, their height relative to the vertical living range of the present assemblage, and adjacent tide gauge data, it is inferred that the northern part of Iwami-Tatamigaura has 0.8-1.1 m uplifted during the 1872 Hamada earthquake. Assuming a 0.8 m coseismic uplift in the whole of survey area, most of the central and the southern parts of wave-cut-bench would be submerged below the low tide before the earthquake, based on the height distribution represented by Digital Surface Model created from the Unmanned Aerial Vehicle survey. This result shows a discrepancy with the fact that the most part of the wave-cut-bench had already emerged in the historical picture “Tokonoura-Ezu” painted in 1817. To explain the landscape of the picture, the amount of coseismic uplift in the central and the southern part must be 0.3-0.4 m or less. These data suggest that the regional difference in the amount of uplift occurred between the northern part and the central to southern part. Because no recent surface rupture can be identified between the two areas, such crustal deformation was probably caused by tilting movements toward the south or the southwest. We also propose the other emergence event before the 1872 Hamada earthquake from the upper assemblage in the East Cave. Calibrated ¹⁴C age data and height of them suggest a relative sea-level fall of at least 0.4 m around the 14th century or later. This result seems that coseismic uplift event has occurred with an interval of about 500 years, but there are the other several candidates for the cause of this phenomenon, such as intermittent small uplifts with short intervals, aseismic uplift, and eustatic sea-level fall.

The characteristics of long-period strong ground motions observed in the Tokyo Bay area from the 2009 Suruga Bay earthquake (M_j 6.5) and the 2016 southeast off Mie Prefecture earthquake (M_j 6.5) were studied. The epicenters of two events are in the same direction but the epicentral distance of the 2016 event from Tokyo bay area is about three times as large as that of the 2009 event. The high-frequency ground motion of the 2009 event is significantly larger than that of the 2016 event. However, the spectral amplitudes at frequencies below 0.1 Hz were almost equal level. The multiple-filter analysis and frequency-wavenumber spectral analysis were performed on the low-frequency ground motions around the Tokyo Bay area. The records of the 2016 event were more dispersive than that of the 2009 event and the duration of the ground motion was extended due to the dispersion of surface waves. The surface waves propagated from almost the same direction at frequencies below 0.15 Hz for both event and the coming direction changed from the direction of epicenter to West at higher frequencies. In addition, Love waves propagating from the direction perpendicular to the epicenter direction were detected only in the records of the 2009 event. From these results, it is considered that the surface waves are predominant in both records, but the incident wave fields are different. In order to clarify the characteristics of the seismic wave incident to the Kanto Plain, the records of both earthquakes in the Kanto Mountains and Izu Peninsula were investigated. The record of the 2009 event was short in duration and body waves were predominant. On the other hand, the record of the 2016 event had a long duration and it was thought that surface waves were predominant. The spectral amplitudes at frequencies below 0.1 Hz were almost equal and the spectral ratios of two events were almost the same in the western mountain area and the Tokyo Bay area. The characteristics of the incident waves to the Kanto Plain are reflected in the seismic motions of the Tokyo Bay area. It is presumed that the Nankai Trough accretionary prism amplified the seismic motions of the southeast off Mie Prefecture event, and the amplitude became same level to those of the Suruga Bay event.

A large earthquake (M7.1) occurred during the 1914 eruption of Sakurajima volcano in Kagoshima prefecture, Japan. I estimated seismic intensities in the present Japan Meteorological Agency (JMA) scale for the 1914 Sakurajima earthquake in Kagoshima city. Previous studies on seismic intensities in Kagoshima city are used their own unique methods to estimated seismic intensities from data of damaged houses and stonewalls. I used a method of estimating seismic intensities from data of damaged houses, which has been commonly used. Previous studies using this method have assumed that each household owns one house, but the numbers of damaged houses in some towns are significantly larger than those of households in the case of the 1914 Sakurajima earthquake. I, therefore, made two assumptions A and B. The assumption A is that each household owns one house. The assumption B is that the ratio of houses to households in each town equal to the maximum ratio of total damaged houses to households. As a result, the maximum seismic intensities for the assumptions A and B are 6 upper and 6 lower, respectively. The areas with higher seismic intensity than 6 lower are consistent with the previous studies on the 1914 Sakurajima earthquake, but the areas with lower seismic intensities than 5 upper are inconsistent because I did not take stonewall damages into consideration, which the previous studies did. I compared our results with distributions of seismic intensities predicted by an attenuation relation between intensity and distance with two different sets of source parameters. One set of source parameters showed seismic intensities of 6 lower and 6 upper in Kagoshima city, and the other set showed those of 5 lower and 5 upper. The observed seismic intensities are on average lower than the predicted ones, probably suggesting that large coseismic slip area is distributed further than epicenter, that Kagoshima city is located in the opposite direction of the rupture propagation and its directivity effect make the amplitudes smaller, or that amplitudes at frequencies which affect seismic intensities are smaller than those expected from the magnitude. The observed seismic intensity distributions are also rougher than the predicted ones and could be affected by local soil conditions.