Search

In this study, we performed two-dimensional numerical simulations of temperature distribution associated with subduction of the Philippine Sea (PHS) plate, and estimated the dehydration process of hydrous mid-ocean-ridge basalt (MORB) in the oceanic crust in the Tokai district, central Japan. We discuss the relationship among temperature, dehydration, and a seismogenic zone of an expected megathrust Tokai earthquake, and the occurrence of a slow slip event (SSE) and deep low-frequency earthquakes (LFEs). The depth range of the seismogenic zone for the megathrust earthquake was estimated to be 8–22 km, narrowing toward the east. The most suitable value of the pore pressure ratio on the plate interface was estimated to be 0.97, indicating minimal frictional heating there. The temperatures of the upper surface of the PHS plate, where the Tokai SSE occurred from 2000 to 2005, were estimated to be 350–450 °C. Therefore, the Tokai SSE is considered to have occurred at the transition zone between unstable and stable sliding. In addition, hydrous MORB was transformed from blueschist into greenschist near the region where the Tokai SSE occurred. The temperatures of the upper surface of the PHS plate, where LFEs occur, were estimated to be 450–500 °C. Therefore, LFEs are considered to occur near the down-dip limit of the transition zone. The amount of dehydration from the oceanic crust of the subducting PHS plate near the region where LFEs are distributed in a belt-like form decreases toward the east with deepening of the hypocenters of the LFEs. Thus, the mechanisms of generation of LFEs in the Tokai district might differ from those in southwest and central Japan where LFEs generally occur at depths between 27 and 35 km. In summary, we consider that the occurrence of LFEs is related to both temperature conditions and dehydration process. However, there is an exceptional area in the eastern part of the Tokai district where the occurrence of LFEs is restricted by temperature conditions rather than dehydration from the oceanic crust because the hypocenter depth is greater than that in southwest and central Japan.

Recent high-pressure and high-temperature experiments indicate that metastable olivine might persist in the cold core of a slab due to the low reaction rate of the olivine–wadsleyite phase transformation. Recent seismological observations detected a metastable olivine wedge that survives to a depth of 630 km in the Mariana slab. To consider the problem of non-equilibrium phase transformation, we developed a two-dimensional (2-D) Cartesian numerical code that incorporates the effects of kinetics into a thermal convection model. We consider the kinetics of the 410-km olivine–wadsleyite and the 660-km ringwoodite–Pv + Mw phase transformations, including the effects of water content at the 410-km phase boundary. The latent heat release of the 410-km non-equilibrium phase transformations inside the slab is also considered. The results show positive correlations between some of the controlling parameters and the length of the metastable olivine wedge: the faster the subducting velocity, and the lower the water content, the deeper is the metastable olivine wedge. With increasing depth of phase transformation, the effect of latent heat release is enhanced: heating of, at most, 100 °C occurs if olivine transforms into wadsleyite at a depth of approximately 570 km in our model setting. Temperature increase due to the latent heat released stimulates further phase transformation, resulting in further temperature increase, acting as a positive feedback effect. We also attempt to explain the seismological observations by calculating the temperature and phase structures in the Mariana slab. If we assume that the age of the Mariana slab is 150 Myr, the subduction velocity is 9.5 cm/yr, phase transformation occurs from the grain boundary of the parental phase, and the water content is 250 wt. ppm for a grain size of 1 mm, 300 wt. ppm for one of 5 mm, and 100 wt. ppm for intracrystalline transformation, then the metastable olivine wedge survives to a depth of 630 km, which is in good agreement with the seismological observations. This suggests that the deeper portion of the Mariana slab is relatively dry. Assuming that the depression of the 660-km discontinuity by ∼20–30 km within the Mariana slab, as is indicated by seismological observation, is explained by a combination of the depression caused by a negative Clapeyron slope in the cold slab and that due to the kinetics of the 660-km phase transformation, we obtain a gentle Clapeyron slope of −0.9 MPa/K for the phase transformation from ringwoodite to Pv + Mw.

The depth of the post-spinel transformation is investigated for the Tonga slab, by using data from short period seismic networks at western United States and Japan for about 100 deep and intermediate-depth earthquakes within a small (∼200 km by 200 km) area near 20°S. Later phases in a time window ∼3 to 20 s after direct P waves are analyzed to search for S-to-P converted waves at the 660 km discontinuity, which represents the post-spinel transformation. We find that immediately beneath the foci of the deepest earthquakes the discontinuity is depressed down to the depths of 685 ± 5 km on average, and that it dips towards WNW by 10 ± 3 km within 70 km laterally. We constrain the thermal structure near the S to P conversion points based on a plausible assumption that the deepest earthquakes occur around the coldest core of the Tonga slab. The distribution of the hypocenters relocated in this study as well as previously published tomographic images of the same region suggest that the Tonga slab bends upward when approaching the 660 km discontinuity and transiently stagnates around the discontinuity. With these observations as the constraints, we numerically model the thermal structure of the Tonga slab, and estimate the temperature around the conversion points as 1200 ± 100 °C, which is 300 ± 100 K colder than the surrounding mantle. As the average depression of the discontinuity (down to 685 ± 5 km) corresponds to an pressure excess over the global average (660 km) by 1.0 ± 0.2 GPa, the assumption of equilibrium post-spinel transformation results in an estimate of the Clapeyron slope (C1) of - 3.3 - 2.7 + 1.3 MPa/K. We also obtain an independent estimate of the Clapeyron slope (C2) of −2.0 ± 1.0 MPa/K, based on the observation of the dip of the discontinuity and the computed temperature variation (by about 200 K). The discrepancy between C1 and C2 is marginally significant and can be diminished by considering that the slab materials at the conversion points are currently descending across the phase boundary fast enough and thus the depth of the post-spinel transformation is controlled by nucleation kinetics as well as by the temperature.

Global seismic tomography has recently revealed horizontally lying slabs near the upper and lower mantle boundary beneath the Northwestern Pacific region. Although physical mechanisms that could produce such slab stagnation have been proposed based on numerical simulations, there has been little research into what occurs after slab stagnation. We proposed trench advance and trench jumps as effective mechanisms related to the fall of stagnant slabs into the lower mantle, and our numerical simulations of temperature and fluid flow associated with slab subduction in a 2-D box model confirmed these mechanisms. Our results indicate that a supply of slab material associated with further slab subduction after slab stagnation plays an important role in differentiating further slab stagnation from the falling of slabs into the lower mantle. A shortage of material supply would produce extended slab stagnation near the 660-km boundary for ringwoodite to perovskite + magnesiowustite phase transformation, whereas downward force due to further slab subduction on a stagnant slab would enhance its fall into the lower mantle. The behaviors of falling stagnant slabs were not affected by Clapeyron slope values associated with phase equilibrium transformation within the range from −3.0 to 0.0 MPa/K. Compared with models of normal mantle viscosity, a high-viscosity lower mantle played a role in hindering the fall of slabs into the lower mantle, resulting in complicated shapes and slow falling velocities. Lower mantle viscosity structure also affected slab behavior. Slabs tended to stagnate when a low-viscosity zone (LVZ) existed just below a depth of 660 km because friction between the slab and the LVZ was weak there. Slab stagnation around a depth of 660 km also occurred when a high-viscosity zone existed below a depth of 1200 km and acted as a resistive force against a slab, even if the slab existed in the lower part of the upper mantle.

A large-scale slow slip event with a magnitude equivalent to Mw = 7.4 has been observed at seven continuous GPS stations located in the Guerrero–Oaxaca region, southern Mexico, for several months from late 2001 to middle 2002 [Geophys. Res. Lett. (2003) 30: doi: 10.1029/2003GL017219]. We re-analyzed the GPS time series data to estimate the displacements associated with the slow slip event and also the displacement rates during the interseismic period prior to this event, by applying an inversion technique using a Bayesian information criterion (ABIC). We then carried out GPS data inversion again with ABIC and some detailed forward modeling to obtain spatial distributions of the back-slip rates and the slow slip on a 3-D curved plate interface between the subducting Cocos and the overriding North America plates. The results show that the average direction of the back-slip rates is N31.3 ◦ E ± 5.6 ◦ , which is congruent with that of plate convergence in this region. Interplate coupling is found to be very strong down to a depth of about 45 km, indicating a coupling ratio of 0.83–0.86 except for a couple of segments on the model region, and decreases dramatically down below. The average direction of the slow slip is oriented about 20 ◦ –35 ◦ counterclockwise from the opposite direction of the plate convergence. The slow slip detected on the mid-depth segment to the upper rim of the model region reaches about 9–18 cm in the landward region of the Guerrero seismic gap from inversion and forward modeling. From these results, we conclude it is possible that the slow slip may have invaded a deeper part of the strongly coupled, seismogenic zone at least up to a depth of about 25 km. We also estimate stress changes due to the slow slip event. Some decrease in shear stress on the plate interface at these depths suggests that if such a slow slip event occurred episodically, it is possible that the time of occurrence of a forthcoming large earthquake in the Guerrero gap would be delayed to some extent. © 2004 Elsevier B.V. All rights reserved.

Recently, the Geographical Survey Institute of Japan completed the installation of a GPS continuous observation network in Japan, which has enabled us to investigate real-time crustal movements. In this study, we attempt to obtain spatial distribution of interplate coupling and relative plate motion between subducting and overriding plates in southwest Japan, using horizontal and vertical deformation rates, which were observed at 247 GPS observation stations during the period from April 6, 1996 to March 20, 1998. For this purpose, we carried out an inversion analysis of geodetic data, incorporating Akaike's Bayesian Information Criterion (ABIC). As a result, strong interplate coupling was found off Shikoku and Kumanonada regions, which corresponds well with the fault regions of the 1946 Nankai (M 8.1) and the 1944 Tonankai (M 8.0) earthquakes, respectively. We also found that interplate coupling becomes weak at depths deeper than about 30 to 40 km beneath the Shikoku and Kii peninsula. The recurrence time of great trench-type earthquakes was roughly estimated as 107 years, which is consistent with previous research. The direction of relative plate motion is oriented N53°W, which is close to the direction predicted from the plate motion model. On the other hand, a large forward slip was found in the Hyuganada region off southeast of Kyushu. Since the coseismic displacements associated with the two 1996 Hyuganada earthquakes (M 6.6, M 6.6) are removed from the GPS data, this suggests that after-slip occurred near the source region and/or that Kyushu moves southeastward stationarily due to other tectonic forces.

We calculated velocity anomalies by constructing thermal structural models of horizontally lying slabs beneath Kuriles, Central Japan, Northern Izu–Bonin and Central Izu–Bonin arcs. The values of V p and V s anomalies become maximum in the depth range of 100 to 200 km within the slabs, reaching 5 to 6.4% for V p anomalies and 7 to 8.5% for V s anomalies. The calculated V s anomalies are larger than the V p anomalies, and the ratio of the former to the latter is 1.2 to 1.5 within the slab. This suggests that V s anomalies are more sensitive to temperature variation than V p anomalies, indicating that the existence of the horizontally lying slabs is detectable in more detail by using S-wave. We also calculated spatial distributions of physical properties in the horizontally lying slabs beneath the four regions. In the calculation, we assumed that the material of the slabs consists of olivine (Mg 0.89 Fe 0.11 ) 2 SiO 4 . The estimated values of physical properties within the slabs are larger than those in the surrounding upper mantle, except for Poisson's ratio, while the values within the metastable wedge of olivine become smaller. This is most remarkable within the slab beneath the Central Izu–Bonin arc due to the lowest temperature among the four regions. The horizontally lying slab above the 660 km seismic discontinuity beneath the Central Izu–Bonin arc is denser up to 60 kg/m 3 and the slab within the metastable wedge of olivine is lighter up to −120 kg/m 3 . The values of the physical properties, except for Poisson's ratio, decrease with increasing the horizontal distances of the horizontally lying portions of the slabs. This is due to thermal assimilation, which is largely dependent on moving velocity of the slabs.

We attempted to detect reflected waves from deep-focus earthquakes at the lower boundary of a descending slab. For this purpose, we calculated theoretical waveforms in the realistic heterogeneous velocity and Q structures beneath the Japanese islands, and compared them with seismic waveforms observed in the Kanto-Tokai districts of central Japan. Assuming that earthquakes occur within the deeper portion of the slab, we investigated the effects of variation in focal mechanisms, focal depths, velocity contrasts near the lower plate boundary, and the Q structures on theoretical seismograms on the earth's surface. The results show that the amplitudes of the P to P and S to S reflected waves gradually increase toward the subduction zone, and take maximum values near the trench axis. However, travel time difference between the direct wave and the following reflected wave decreases there, making it difficult to detect the reflected wave. Based on these theoretical calculations, we investigated digital three-component seismograms observed in the Kanto-Tokai districts by the National Research Institute for Earth Science and Disaster Prevention, Science and Technology Agency (NIED). The seismograms observed were obtained from deep-focus earthquakes, from 1990 through 1996, with magnitude of 5.0 and greater. After carefully checking seismograms from 15 deep-focus events, we could not detect the reflected waves from the lower plate boundary systematically at expected arrival times from the theoretical calculation. This indicates that the reflected waves are either masked by coda waves followed by the direct P and S waves or below the noise level that is detectable by seismometers. There is also a possibility that the seismic velocity contrast between the slab and the underlying upper mantle may not be so sharp. However, the possibility to detect the S to S reflected waves increases if we check observed waveforms in the horizontal range between 200 to 300 km landward from trench axis for a moderate size deep-focus earthquake with down-dip compression, which occurs near the upper plate boundary at depths of 300 to 400 km.

Inside subducting slabs the interaction of different metastable phase transformations (α-olivine → β-spinel, β-spinel → γ-spinel, γ-spinel → perovskite + magnesiowustite) associated with the latent heat release and absorption for these transitions can result in a complex thermal slab structure. Thermokinetic coupling processes cause thermal anomalies and buoyancy contrasts between slab and mantle that strongly influence the stress field within subducting plates. We have described the thermal field within the downgoing slab in a Lagrangian framework by a self-consistent, high-resolution, two-dimensional thermokinetic coupling model, which also incorporates the effects of nucleation site saturation. Using an adaptive method, the threshold for the nucleation site saturation is found to be less than 0.1% transformation degree in the cold slab interior. This can significantly shift the metastable wedge of α-olivine deeper, as suggested from previous thermokinetic slab models. However, this effect is compensated by the latent heat release. During the α-olivine → β-spinel and β-spinel → γ-spinel phase transitions, latent heat release produce very sharp phase boundaries in the slab. For fast slabs the phase boundaries reveal significant metastable perturbations from the equilibrium state in the cold interior. However, the strong thermal interaction between both ongoing phase transitions results in sharp overlapping phase boundaries with dramatic consequences for the slab stress field. We have investigated the detailed physical properties and the state of stress in the deeper portion of a subducting plate by using up-to-date temperature field and physical properties determined from recent high P-T experiments. We find that the slab is denser up to 100 kg m−3 than the surrounding mantle. Remarkable denser portions up to 260 kg m−3 are located just below the uplifted α → β and β → γ equilibrium phase transitions, while lighter portions up to −210 kg m−3 can be found just above the depressed γ-spinel to perovskite + magnesiowustite transition and for the metastable wedge of α-olivine. These density differences effectively act as a body force and produce a significant stress field, which we calculated with a two-dimensional finite-element code. The results show that the buoyancy-induced forces produce maximum shear stress up to 23 MPa along the metastable wedge and deeper portion just above the depressed last phase transition involving γ-spinel. For the latter, the dominant state of stress is down-dip compression. The calculated P-T dependent state of stress is very similar to the depth distribution of deep-focus

Post-seismic behavior induced by a deep-focus earthquake in subduction zone is closely investigated numerically on realistic upper-mantle models, taking into account the pressure and temperature conditions. For this purpose, the post-seismic stress changes expected from the deep event are calculated by a two-dimensional finite element method (2D FEM), assuming newtonian and power-law creep behaviors as justified by high P – T experiments. Several cases are investigated in detail for various fault depths and thermal structures. Two cases are considered for the location of the fault: one when an earthquake occurs within oceanic crust overlying a subducting plate, and the other when an earthquake is located within the plate. We also examine the case when the temperature distribution within the plate is lower by about 200°C on average than that in the upper mantle at the same depth. In addition, the post-seismic behavior due to a deep-focus event in a petrological thermal model of a mantle wedge is investigated. The results show that the post-seismic behavior associated with the deep event occurring within the oceanic crust indicates a pattern of relaxation of the coseismic stress changes. Lower temperatures inside the plate would suppress the post-seismic displacements and stress changes effectively. The petrological model also has some influence on the post-seismic flow pattern in the mantle wedge. We also study a possible correlation between deep-focus earthquakes and a shallow large event which has been detected in the northwestern Pacific region. However, effective mechanisms to explain the phenomenon could not be identified from the viewpoint of post-seismic behavior caused by power-law creep.

The interplate coupling and stress accumulation process associated with the subduction of the Philippine Sea plate beneath the Eurasian plate along the Nankai trough, where large earthquakes have occurred periodically, have been investigated in detail, taking into account the three-dimensional complicated structure of the Philippine Sea plate and its motion relative to the Eurasian plate. For this purpose, we mainly used extensive geodetic data in southwest Japan, such as levelling, tide gauge and trilateration, which presumably are effective means of revealing coupling properties, and applied a three-dimensional finite element method. Here, two seismic periods associated with the 1944 Tonankai and 1946 Nankaido earthquakes were considered—interseismic and pre-seismic stages. Comparing the coupled regions obtained for the two cases, a possible explanation of its temporal change was proposed: it might gradually spread from a shallower to deeper portion, with transition from the interseismic to pre-seismic stage. A comparison of the coupling properties between southwest Japan and northeast Japan reveals the following properties. In southwest Japan, the coupling is generally strong, and its region is limited in a relatively shallow portion. The pre-seismic coupled regions coincide with the co-seismic and post-seismic fault zone, suggesting that the accumulated stress is released at the time of a large earthquake. In northeast Japan, on the other hand, the coupling is weak and its extent appears to reach a depth of about 100 km, and the co-seismic and post-seismic regions are limited to the upper half of the coupled region. The weakly coupled region in southwest Japan seems to correspond well to the region with high seismic activity, a high dip angle subducting slab and the possible existence of seamounts, in contrast to the strongly coupled region correlated with low seismicity, low dip angle subduction and no seamounts. On the western coupled region off Shikoku, an anomalously strongly coupled region, corresponding to a convergence rate of 7–8 cm year −1 , was identified. This may be caused by geomorphological properties, such as the low dip angle of the subducting plate and smooth plate boundary, resulting in a large amount of stress concentration there. The stress field associated with the interplate coupling was also calculated. There is fair agreement between the observations and calculations near the plate boundary beneath the southern Kii peninsula and Shikoku. Although the effects of the interplate coupling may have some influence on strike-slip faulting with the N-S compressive axes beneath Shikoku, other earthquake generating stress, such as a two-layered double-buckling, would be dominant there.

The Tonankai earthquake occurred in 1944 along the Nankai trough between the continental Eurasian plate and the subducting Philippine Sea plate. We investigated spatial and temporal correlations between the occurrence of this event and intraplate seismic activities in the inland region of southwest Japan. In order to elucidate the possible relationship and dynamic coupling between the 1944 Tonankai and the 1946 Nankaido earthquakes, we calculated stress changes due to the Tonankai earthquake by applying a three-dimensional finite element method and compared them with post-seismic activity. Agreement between them appears quite well at a depth of ∼ 10 km, indicating that stress release in an E-W direction associated with the Tonankai earthquake may have reduced the seismic activity in the inland region. Particularly, the observed drastic decrease of seismicity in the Wakayama region just after this event can be explained by significant decrease of compressive stresses in the E-W direction. Extensive seismicity quiescence preceding the 1946 Nankaido earthquake, which occurred adjacent to the source region of the Tonankai event, may be related to the enhancement of N-S oriented compressive stress due to accelerating northward movement of the Philippine Sea plate. Therefore, in addition to the decrease of stresses in the E-W direction, the increase of the N-S oriented compressive stresses might play an important role to yield low seismic activity just after the Tonankai event. The calculated results show that the post-seismic shear stress increased and the normal stress decreased near the hypocentral area of the 1946 Nankaido earthquakes, while the shear stress decreased and normal stress increased in its surrounding area on the fault. This indicates that the Tonankai earthquake made it easier for the rupture of the Nankaido earthquake to originate at the observed hypocenter. This triggering mechanism may explain the synchronous pattern of historical great interplate earthquakes along the Nankai trough; a great earthquake occurs in the western region within a few years after the first breakage in the eastern block.

We investigate the changes of tectonic stress in the crust and upper mantle beneath southwest Japan associated with the occurrence of the 1944 Tonankai and the 1946 Nankaido earthquakes, and their possible relation to impending earthquakes by using a three-dimensional finite element method. We construct a three-dimensional model incorporating the assumed fault planes of these earthquakes. The calculated results indicate that the temporal changes in the direction of the compressive stress from N-S to E-W beneath the Kii channel and the eastern part of Shikoku observed after the two earthquakes may be attributed only to the seaward rebound of the hanging-wall side of the assumed faults for these earthquakes. The temporal change appears to have returned to the original stage after 20 yr, indicating that the coseismic disturbance of the stress field could persist for several years. The two earthquakes may have caused a local stress drop of