Search

Abstract: We describe two transient slow slip events that occurred before the 2011 Tohoku-Oki earthquake. The first transient crustal deformation, which occurred over a period of a week in November 2008, was recorded simultaneously using ocean-bottom pressure gauges and an on-shore volumetric strainmeter; this deformation has been interpreted as being an M6.8 episodic slow slip event. The second had a duration exceeding 1month and was observed in February 2011, just before the 2011 Tohoku-Oki earthquake; the moment magnitude of this event reached 7.0. The two events preceded interplate earthquakes of magnitudes M6.1 (December 2008) and M7.3 (March 9, 2011), respectively; the latter is the largest foreshock of the 2011 Tohoku-Oki earthquake. Our findings indicate that these slow slip events induced increases in shear stress, which in turn triggered the interplate earthquakes. The slow slip event source area on the fault is also located within the downdip portion of the huge-coseismic-slip area of the 2011 earthquake. This demonstrates episodic slow slip and seismic behavior occurring on the same portions of the megathrust fault, suggesting that the faults undergo slip in slow slip events can also rupture seismically. [Copyright &y& Elsevier]

Abstract: Recently, the occurrence of slow earthquakes such as low-frequency earthquakes and very low-frequency earthquakes have been recognized at depths of about 30 km in southwest Japan and Cascadia. These slow earthquakes occur sometimes in isolation and sometimes break into chain-reaction, producing tremor that migrates at a speed of about 5–15 km/day and suggesting a strong interaction among nearby small asperities. In this study, we formulate a 3-D subduction plate boundary model with two types of small asperities chained along the trench at the depth of 30 km. Our simulation succeeds in representing various types of slow earthquakes including low-frequency earthquakes and rapid slip velocity in the same asperity, and indicates that interaction between asperities may cause the very low-frequency earthquakes. Our simulation also shows chain reaction along trench with propagation speed that can be made consistent with observations by adjusting model parameters, which suggests that the interactions also explain the observed migration of slow earthquakes. [Copyright &y& Elsevier]

To study the physical properties along the subducting plate boundary at the Japan Trench, a seismic study was carried out using ocean bottom seismometers (OBSs) and artificial sources was carried out. The 250 km long survey line with an almost N–S strike crosses the two major focal zones of the 1968 Tokachi-Oki earthquake and the 1994 Sanriku-Haruka-Oki earthquake. Ray tracing, and non-linear inversion were used to analyze the data obtained by the OBSs. P-wave velocity structure down to the island-arc-Moho was obtained.The result shows distinct heterogeneities along the plate boundary at 40°10′N, which are located exactly at the southern boundary of major moment release regions of the above two earthquakes. P-wave velocities for the crust north of 40°10′N are approximately 7% slower than those for the south. The depths of the island-arc-Moho vary from 15 km in the south to 21 km in the north below the sea surface. Three explanations of the velocity heterogeneities are presented in terms of the migration of water. [Copyright &y& Elsevier]

In the region off the Boso Peninsula, Japan, the Pacific plate is subducting westward beneath both the Honshu island arc and Philippine Sea plate, while the Philippine Sea plate is subducting northwestward beneath the Honshu island arc. These complex tectonic interactions have caused numerous seismic events occurred in the past. To better understand these seismic events, it is important to determine the geometry of the plate boundary, in particular the upper surface of the Philippine Sea plate. We conducted an active-source seismic refraction survey in July and August 2009 from which we obtained a 2-D P-wave velocity structure model along a 216-km profile. We used the velocity model and previously published data that indicate a P-wave velocity of 5.0 km/s for the upper surface of the subducting Philippine Sea plate to delineate its boundary with the overriding Honshu island arc. Our isodepth contours of the upper surface of the Philippine Sea plate show that its dip is shallow at depths of 10 to 15 km, far off the Boso Peninsula. This shallow dip may be a result of interference from the Pacific plate slab, which is subducting westward under the Philippine Sea plate. Within our survey data, we recognized numerous seismic reflections of variable intensity, some of which came from the upper surface of the Philippine Sea plate. An area of high seismic reflection intensity corresponds with the main slip area of the Boso slow slip events. Our modeling indicates that those reflections can be explained by an inhomogeneous layer close to the upper surface of the Philippine Sea plate. [ABSTRACT FROM AUTHOR]

Abstract: We obtained ambient seismic noise interferograms as seismic reflection images using autocorrelation functions (ACFs) in the northeastern Japan subduction zone. The ACFs with a time window length of 120s were calculated from the continuous seismic records obtained at each seismic station over an analysis period of 300days. These ACFs show some distinct signals with relatively large amplitude without any significant temporal variations during the analysis period. The signals, which are stable, appear at both small lag times of less than 10s and large lag times of 20–50s during the analysis period. The lag time of 10s corresponds to the travel time of the PP reflection arrival from the continental Mohorovičić discontinuity. The signals with the large lag times between 30 and 50s correspond to the back-scattered signals from the mantle wedge or the plate boundary; these signals are identified clearly at the stations located on the back-arc side. In the ACFs calculated from the records obtained from the fore-arc side stations, weak signals (interpreted as the reflection from the plate boundary) with a lag time range of 20 to 30s are observed. We constructed depth-migrated images using the ACFs to obtain the reflectivity profile by assuming that the ACFs represent Green''s functions composed of a random wavefield excited by stochastic sources or scatterers distributed in the vertical or near-vertical direction from the stations. Further, we assumed that the ACFs can be treated as zero-offset seismic traces recorded at each of the stations. The depth-migrated images show a relatively seismically transparent structure within the subducting Pacific slab and a reflective structure within the mantle wedge; this reflective structure is characterized by low-velocity zones corresponding to the wedge flow imaged by 3-D seismic velocity tomography. [Copyright &y& Elsevier]

Abstract: The 2007 Niigata-ken Chuetsu-oki earthquake occurred on July 16, 2007 with a magnitude of 6.8. Immediately after the mainshock, a dense network of ocean-bottom seismometers and temporary land seismic stations were deployed to obtain the accurate aftershock location. A seismic survey using ocean-bottom seismometers, land stations and controlled sources at sea and on land was also conducted along three profiles to estimate the detailed velocity structure of the source region. A thick sedimentary layer covers the crust, and this layer is thickest near the coast. The upper crust has a P-wave velocity of 6km/s and a large lateral heterogeneity with respect to thickness and velocity. The lower crust, with P-wave velocity of 7km/s, has a thickness of approximately 10km. The thickness of the crust is estimated to be approximately 24km. The precise aftershock distribution was obtained by using the high-resolution velocity structure. The aftershocks in the upper crust form a plane dipping to the southeast. Most of the aftershocks are located in the upper crust; in addition, a small number of aftershocks in the lower crust seem to be positioned on the same plane formed by the aftershocks in the upper crust. The mainshock fault estimated from the aftershock distribution is positioned in the high-velocity body of the upper crust. A large deformation of the sediments above the epicentral region is interpreted to be due to the repeated occurrence of large earthquakes. The lower crust immediately below the fault has low velocity. The source region of the 2007 event in the upper crust is sandwiched by a weakening structure. We suggest that the ductile deformation of both the sediments and the lower crust causes stress accumulation in the source region. [Copyright &y& Elsevier]

Abstract: We have performed a two-dimensional numerical simulation to elucidate the physical processes governing earthquake behavior when significant stress perturbations are produced by interaction between fault segments. Our model involves two seismogenic segments separated down-dip on a subduction zone plate boundary and incorporates a rate- and state-dependent friction law. Based on repeated simulations under different scenarios, we find that the rate- and state-dependent frictional parameters (b–a) and d c in the deeper seismogenic segment are required to be smaller than those in the shallower segment in order that more deep earthquakes occur than shallow ones. Our simulations show that slip amounts in either seismogenic segment increase in each of the co-, pre- and post-seismic stages when an earthquake occurs shortly after another earthquake in the other seismogenic segment. Conversely, when earthquakes occur in a single seismogenic segment several times in succession while the other segment remains locked, all three pre-, co-, and post-seismic slip amounts become smaller. These results imply that precursory changes do not necessarily occur at the same level on every occasion. In cases of multiple rupturing, the co-seismic slip of the later earthquake in a pair is approximately characteristic when frictional stability in the aseismic segment between the two seismogenic fault segments is strong enough to produce different rates of seismicity on each segment. Our simulation also shows that the degree to which rupture initiation points vary between different earthquakes is higher in low seismic coupling regions. This result may prove useful in estimating seismic coupling coefficients from observations of interplate hypocenters. [Copyright &y& Elsevier]

Abstract: Large earthquakes have repeatedly occurred in the area off southeastern Hokkaido Island, Japan, as the Pacific Plate subducts beneath the island, which is on the North American Plate. The most recent large earthquake in this area, the 2003 Tokachi-oki earthquake (Mw=8.0), occurred on September 26, 2003. In order to investigate aftershock activity in the rupture area, 47 Ocean Bottom Seismometers (OBSs) were quickly deployed after the main shock. In the present study, we simultaneously estimate the hypocenters and 3-D seismic velocity models from the P- and S-wave arrivals of the aftershocks recorded by OBSs. The subducting plate is clearly imaged as a northwest dipping zone in which Vp is greater than 7 km/s, and the relocated hypocenters also show the subducting Pacific Plate. The aftershock distribution reveals that the dip angle of the plate boundary increases abruptly around 90 km from the Kuril Trench. The bending of the subducting plate corresponds to the southeastern edge of the rupture area. The island arc crust on the overriding plate has P-wave velocities of 6–7 km/s and a Vp/Vs of 1.73. A region of Vp/Vs greater than 1.88 was found north of the epicenter of the main shock. The depth of the high Vp/Vs region extends about 10 km upward from the plate interface. The plate boundary just below the high Vp/Vs region has the largest slip at the main rupture. A high Vp anomaly (~7.5 km/s) is found in the island arc crust in northeast part of the study area, which we interpret as a structural boundary related to the arc–arc collisional tectonics of the Hokkaido region, as the rupture of the main shock terminated at this high Vp region. We suggest that the plate interface geometry and the trench-parallel velocity heterogeneity in the landward plate are principal factors in controlling the rupture area of the main shock. [Copyright &y& Elsevier]

Abstract: We determined the three-dimensional Vp and Vs structures beneath Japan by applying seismic tomography to a large number of arrival times recorded at temporary stations in the Japan Sea and the Pacific Ocean, as well as those at permanent stations on the Japan Islands. As a result, we obtained more precise seismic images than previous studies. In the crust and the uppermost mantle, southwestern Honshu exhibited weaker heterogeneity than the other areas in Japan, corresponding to the distribution of active volcanoes. Stripe-like heterogeneities exist in the subducting Pacific slab. Relatively low-velocity zones correspond to low-seismicity areas in the Pacific slab, suggesting that the slab is possibly torn or thin around the areas. The fact that nonvolcanic deep tremors associated with the subducting Philippine Sea slab beneath Shikoku, Kii, and Tokai do not occur in zones of high Vp, high Vs, and low Vp/Vs ratio may reflect the existence of fluids generated by the dehydration processes of the slab. Prominent and wide low Vp and Vs zones exist beneath central Honshu at the depth range of 30–60km, where the volcanic front related to the subducting Pacific plate is located and seismicity around the Philippine Sea plate is very low. This condition may exist because magma genesis processes related to the subducting Pacific plate activate the same processes around the Philippine Sea plate. [Copyright &y& Elsevier]

Abstract: The Tsushima Basin is located in the southwestern Japan Sea, which is a back-arc basin in the northwestern Pacific. Although some geophysical surveys had been conducted to investigate the formation process of the Tsushima Basin, it remains unclear. In 2000, to clarify the formation process of the Tsushima Basin, the seismic velocity structure survey with ocean bottom seismometers and airguns was carried out at the southeastern Tsushima Basin and its margin, which are presumed to be the transition zone of the crustal structure of the southwestern Japan Island Arc. The crustal thickness under the southeastern Tsushima Basin is about 17 km including a 5 km thick sedimentary layer, and 20 km including a 1.5 km thick sedimentary layer under its margin. The whole crustal thickness and thickness of the upper part of the crust increase towards the southwestern Japan Island Arc. On the other hand, thickness of the lower part of the crust seems more uniform than that of the upper part. The crust in the southeastern Tsushima Basin has about 6 km/s layer with the large velocity gradient. Shallow structures of the continental bank show that the accumulation of the sediments started from lower Miocene in the southeastern Tsushima Basin. The crustal structure in southeastern Tsushima Basin is not the oceanic crust, which is formed ocean floor spreading or affected by mantle plume, but the rifted/extended island arc crust because magnitudes of the whole crustal and the upper part of the crustal thickening are larger than that of the lower part of the crustal thickening towards the southwestern Japan Island Arc. In the margin of the southeastern Tsushima Basin, high velocity material does not exist in the lowermost crust. For that reason, the margin is inferred to be a non-volcanic rifted margin. The asymmetric structure in the both margins of the southeastern and Korean Peninsula of the Tsushima Basin indicates that the formation process of the Tsushima Basin may be simple shear style rather than pure shear style. [Copyright &y& Elsevier]

Abstract: The Woodlark Basin, located south of the Solomon Islands arc region, is a young (∼5 Ma) oceanic basin that subducts beneath the New Britain Trench. This region is one of only a few subduction zones in the world where it is possible to study a young plate subduction of several Ma. To obtain the image of the subducting slab at the western side of the Woodlark Basin, a 40-day Ocean Bottom Seismometer (OBS) survey was conducted in 1998 to detect the micro-seismic activity. It was the first time such a survey had been performed in this location and over 600 hypocenters were located. The seismic activity is concentrated at the 10–60 km depth range along the plate boundary. The upper limit just about coincides with the leading edge of the accretionary wedge. The upper limit boundary was identified as the up-dip limit of the seismogenic zone, whereas the down-dip limit of the seismogenic zone was difficult to define. The dip angle of the plate at the high seismicity zone was found to average about 30°. Using the Cascadia subduction zone for comparison, which is a typical example of a young plate subduction, suggests that the subduction of the Woodlark Basin was differentiated by a high dip angle and rather landward location of the seismic front from the trench axis (30 km landward from the trench axis). Furthermore, as pointed out by previous researchers, the convergent margin of the Solomon Islands region is imposed with a high stress state, probably due to the collision of the Ontong Java Plateau and a rather rapid convergence rate (∼10 cm/year). The results of the high angle plate subduction and inner crust earthquakes beneath the Shortland Basin strongly support the high stress state. The collision of the Ontong Java Plateau, the relatively rapid convergence rate, and moderately cold slab as evidenced by low heat flow, rather than the plate age, may be dominantly responsible for the geometry of the seismogenic zone in the western part of the Woodlark Basin subduction zone. [Copyright &y& Elsevier]

The Izu-Bonin Trench (IBT) has characteristics different from other trenches in many aspects. A chain of serpentine seamounts is exposed at the forearc slope 40–50 km west of the Izu-Bonin Trench axis. In this area, few large earthquakes (>M7.0) have occurred at shallow depths (0–100 km) and many large earthquakes have occurred at greater depths (>400 km). To reveal the mechanism of serpentine diapiring and the relation between heterogeneity in earthquake occurrence and plate subduction, seismic refraction/reflection studies at the forearc slope, with two lines perpendicular and parallel to the Izu-Bonin Trench axis, were carried out in 1998.Detailed seismic velocities were obtained for both lines by seismic tomography using refracted first arrivals and reflected arrivals. On the E–W line, the subducting plate interface and the Moho of the subducting oceanic plate were recognized. The velocity of the mantle wedge is lower than that of the average velocity of oceanic mantle. Along the subducting plate, a layer between the two plates (Plate Boundary Layer (PBL)) with a low-velocity appears just beneath the serpentine diapir and forms a sill shape, which becomes thicker with increasing velocity from the diapir toward the west, and connects to the mantle wedge. The low-velocity materials in the PBL are interpreted as chrysotile, which is a low-temperature phase of serpentine produced by the hydration of peridotite by water supplied by the subducting slab. The chrysotile in the PBL might act as a lubricant and decrease seismic activity along the subduction zone, and this can explain the characteristics of seismicity in the Izu-Bonin subduction zone (IBSZ). In contrast to the E–W line, the crustal structure along the N–S line is rather homogeneous. [Copyright &y& Elsevier]