After the 1995 Hyogoken-Nanbu earthquake, National Research Institute for Earth Science and Disaster Prevention (NIED) has constructed new nation-wide high sensitivity seismograph network. We call this network “NIED Hi-net” or “Hi-net”. Hi-net stations uniformly cover the Japanese Islands in addition to old high sensitivity seismographs. To reduce the seismic noise near the ground surface, a three-component velocity-type seismograph is installed at the bottom of a borehole. Recorded signal at each station is continuously transmitted to NIED data management center (DMC) at Tsukuba using IP-VPN technology. Hi-net waveform data are sent to the JMA and other related institutes for seismic activity monitoring and seismological research. NIED archives all waveform data generated by the Hi-net and other seismic networks. These waveform data and JMA hypocenter catalog are open to the public via the Hi-net website. Although less than 10 years has passed since the Hi-net data became available, this network has brought us a lot of contributions to geoscience. Stable and long-term observation by this seismograph network may play important roles to mitigate earthquake disasters and to predict earthquakes.

抄録全体を表示

After the disastrous 1995 Kobe earthquake, a new national project has started to drastically improve seismic observation systems in Japan. This project was requested to construct a broadband seismograph network with a spacing of 100 km covering the whole of Japan utilizing existing broadband stations as much as possible. National Research Institute for Earth Science and Disaster Prevention (NIED) has constructed a new broadband seismograph network consisting of 73 stations called NIED F-net. The broadband seismograph (STS-1/2) and strong motion velocity-type seismograph (VSE-355G2/G3) are installed in 30 to 50 m long vaults at each F-net station. All the data obtained from the F-net are archived at NIED and provided to researchers and general public through the Internet. The NIED F-net has been routinely analyzing moment tensor solutions for earthquakes of magnitude greater than 3.5 based on the unified hypocenter catalog maintained by the Japan Meteorological Agency. High quality data provided from the dense F-net are utilized to address a lot of interesting research issues: seismogenic stress field, the source process of large earthquakes, structure of the Earth’s interior, Earth’s background free oscillations, and very low frequency earthquakes.

抄録全体を表示

The history and developments of strong-motion observation in Japan are reviewed. A special attention is paid to the recent developments of instruments and nation-wide observation networks. After the 1995 Hyogoken Nanbu earthquake, thousands of strong-motion seismographs and seismic intensity meters were installed all over Japan. Summary of those networks and the latest strong-motion seismographs are described. The future developments for desirable strong-motion observation are also discussed.

抄録全体を表示

The observation of the crustal deformation has contributed to the researches and studies in seismology and geodesy. Recent advance of space geodetic techniques including GPS (Global Positioning System) and InSAR (Synthetic Aperture Radar Interferometry) greatly improved the accuracy and spatio-temporal resolution to observe the crustal deformation on land. Networks of borehole tiltmeters and strainmeters enabled us to detect very small signals related with slow slip events on the subducting plate boundary. Some of recent discoveries of new phenomena and advanced insights for crustal activities including large earthquakes are owing to these developments and organized networks for the monitoring of crustal deformation. Because the history of these observations is much shorter than the interval of large earthquakes, the continuation and development of the monitoring are important for further researches and understanding. In order to understand the crustal activities, it is also important to discriminate between tectonic and non-tectonic deformation induced by underground water and surface loads such as ocean and atmosphere in the observed signals.

抄録全体を表示

Technologies, results and problems to be solved on the borehole measurements are briefly reviewed with particular attention to the measurement of crustal deformation. Detection of tilt and strain change associated with very low frequency tremors at depth in the southwestern Japan should be noted as one of the best topics. Reduction of noises due to meteorological and hydrological origins has been discussed for the past thirty years and remains still incomplete. In order to reduce such noises, deep borehole can be one of the solutions. Two major questions to the conventional hydraulic fracturing method have been raised for the past twenty years. Several kinds of stiff pressurizing system can be solutions for the measurement of crustal stress at depth shallower than 2 or 3 km. For the case deeper than 3 km, usual stress measurement system cannot be applied because of borehole breakouts or drilling induced fractures. In such cases, further development of measurement method is necessary for direct measurement of stress.

抄録全体を表示

This paper reviews on recent progress in seafloor earthquake observations and instruments in Japan. Ocean bottom seismometers (OBS) have improved into a 24-bits digital OBS. Great strides were made in the 2000’s. The research of the earth’s interior from the ocean has been progressed rapidly by utilizing the data from the ocean borehole observatory and the broadband OBS array. Large scale array observations with the use of the long term OBS have imaged earthquake activities in a space-time domain in and around large asperities along Kuril Trench, Japan Trench, and Nankai Trough. A seafloor earthquake monitoring also has been evolving into a new generation for a two-dimensional dense network.

抄録全体を表示

Seafloor geodesy is important because most of seismogenic zones of subduction earthquakes are under the seafloor. Observation of seafloor crustal movements has been initiated and proceeded these ten years with methods different from geodetic observation on land, because electromagnetic waves are unavailable underwater. GPS/Acoustic seafloor positioning, a novel method for observation of a regional crustal movement, has yielded promising results. Ocean bottom pressure is the most realistic observable for the monitoring of vertical movement, and can be continuously observed in real time with a cabled system.

抄録全体を表示

Methods for computing synthetic seismograms for aspherical global earth models (i.e., methods for the spherical coordinates) are reviewed. The methods for the spherical coordinates should be applicable to the problems in the Cartesian coordinates. However, formulations in the spherical coordinates are, in general, tedious, and their detailed methodology appears to be unfamiliar with researchers in other studying fields. The purpose of this paper is to review the methods for the spherical coordinates without using tedious mathematics. Due to the limitation of the space, I focus on methods which have been applied to global tomographies. The approximation methods, merits and limitations are discussed.

抄録全体を表示

The development of high-performance parallel computers such as the cluster of PCs and the Earth Simulator supercomputers made it possible for realistic 3-D simulation of seismic wave propagation and strong ground motions in heterogeneous structure. The seismic wavefield in heterogeneous structure is often evaluated numerically by solving equation of motions based on the finite-difference method (FDM). In the present paper, the FDM simulation of seismic wave propagation in heterogeneous structure is reviewed briefly with recently developed techniques such as for accurate representation of seismic phenomena in heterogeneous structure. Efficient techniques to reduce computational cost of the FDM simulation are also presented. The effectiveness of the vector-computing hardware for the FDM simulation and a parallel simulation technique are demonstrated with linearly increasing computational performance using large number of processors of the Earth Simulator.

抄録全体を表示

This paper deals with the stochastic scattering theory for short-period seismic waves in random media. The stochastic approaches to simulate the seismic wave propagation in the heterogeneous lithosphere was employed by K. Aki in seismology in 1970s. His pioneering studies proved that stochastic approaches are very useful for understanding short-period seismic wavefield which is usually too complicated to simulate by deterministic approaches. Numerous researches have since employed stochastic methods to study short-period seismograms. H. Sato and his colleagues have greatly contributed to the development of the scattering theory. Analyses of seismograms based on the theory have revealed that the random inhomogeneity in the lithosphere is well characterized by a power-law spectrum, with the regional variation related to seismotectonic conditions. Recent observations and advanced data analyses, however, are requiring more development of the scattering theory to simulate seismic wave propagation not only in isotropic random media but also in anisotropic random media. This paper covers multiple-forward scattering theories for velocity shift and seismogram envelopes. The achievement of the scattering theory in isotropic random media and the recent progress in anisotropic random media are reported.

抄録全体を表示

Inversion analyses play a central role in solid earth sciences, since observable quantities are very limited for the earth’s interior. Classical least squares method does not work well in these fields, since observed data are commonly inaccurate and/or insufficient. Least squares methods with additional conditions, such as damping or smoothing, have been widely used, but the weight of damping or smoothing have to be manually adjusted, if we do not take in a probabilistic point of view. In 1980, H. Akaike proposed a Bayesian Information Criterion (ABIC), where smoothness constraint is regarded as prior information that is combined with information from observed data by Bayes’ rule, and then the optimal weight between the information from observed data and prior constraint is objectively determined by minimizing ABIC. ABIC had been introduced to geophysics by several studies. Among them, Yabuki and Matsu’urea (1992) has been the most influential. The inverse method developed by them has been widely applied to various problems in seismology and geodesy. Recently, it has become clear that the inverse method must be further developed beyond the framework given by Yabuki and Matsu’ura (1992). Generalization has been performed in dealing with prior constraints, hyperparameters, and observed data.

抄録全体を表示

It has long been believed that only large earthquakes excite free oscillations of the solid Earth. In 1998 a few Japanese groups reported existence of Earth’s background free oscillations even on seismically quiet days. The excited modes are almost exclusively fundamental spheroidal modes with amplitudes of about 0.5 nGal (10^-11 m/s²), and they fluctuate persistently in little correlation with their neighboring modes. Cumulative effects of many small earthquakes are too small to explain the amplitudes. These features suggest that the background free oscillations are excited persistently by random disturbances globally distributed near the Earth’ surface. The intensities of these modes clearly show annual and semiannual variations with the largest peak in July and a secondary peak in January. The observed amplitudes of some modes are anomalously large relative to the adjacent modes. These are the modes that are theoretically expected to be coupled with the acoustic modes of the atmospheric free oscillations. All of these features suggest that atmospheric disturbance is one of the most likely excitation sources of this phenomenon. Assuming that atmospheric pressure disturbance acting on the Earth’s surface is a primary excitation source, some quantitative comparison has been made between the atmospheric pressure disturbance and Earth’s background free oscillations. Shortly after the discovery, some groups proposed oceanic excitation mechanism. This mechanism is based on two observations; (1) The typical frequency of Earth's background free oscillations of about 0.01 Hz coincides with that of ocean bottom pressure sensors. (2) The excitation sources are dominated in the northern Pacific in winter of the northern hemisphere and in the circum-Antarctic in winter of the southern hemisphere. The source distribution is consistent with oceanic wave height data. Based on these observations oceanic excitation mechanism by ocean infragravity waves through their nonlinear interaction was proposed. However, the proposed mechanisms work efficiently only in shore regions, and hence may not be consistent with the observed spatial extent of the excitation sources, and they cannot explain observed acoustic resonance. At present there is little consensus about the excitation mechanism that can explain all of the observed features. These observations suggest that excitation sources are superposition of oceanic and atmospheric phenomena. The phenomena of background free oscillations should be understood as those in a single system of the atmosphere, ocean and solid Earth.

抄録全体を表示

Change in regional seismic activity has been generally considered to be closely related to stress and fault strength or their spatio-temporal changes. To explain the causal relations between seismic activity and stress state, one must introduce some assumptions or hypotheses based on a physical scheme. In Shizuoka Prefecture, central Japan, small but distinct changes of seismic activity in the locked Tokai slab, which evidently divided the focal region into activated zones and quiescent ones, were observed almost simultaneously with the slow slip event beneath Lake Hamana from 2000 to 2005. Here, I propose the following hypothesis to associate the seismic activity change with the stress condition. The slow slip may have promoted a quasi-static sliding on the rim of the asperities, resulting in stress redistribution on the focal region. The zones possibly activated by stress concentration were then assigned to asperities. Based on this consideration, I specified several major asperities on the anticipated rupture area of the future Tokai earthquake. I thus conclude that monitoring seismic activity change together with an adequate mechanical hypothesis may help us understand the locking state along the plate interface with respect to the spatio-temporal changes in stress rate.

抄録全体を表示

In recent years, many researchers in seismicity analysis have focused on earthquake triggering or fault interactions.To evaluate these phenomena quantitatively, models for the spatio-temporal distribution of seismic activity have been constructed. This review paper introduces two important models : the ETAS (epidemic-type aftershock sequence) model developed by Y. Ogata and a seismicity model based on rate- and state-dependent friction law by J. Dieterich. The former is a pure statistical model, while the latter includes some physical parameters in its derivation.These two models now play separate role in probabilistic earthquake forecasting. For further achievement in the probabilistic earthquake forecasting, the integration of the two contrasting models should be considered. Further development of models would give new insights into earthquake occurrence.

抄録全体を表示

This paper reviews probabilistic prediction methods for recurrent earthquakes and aftershocks. The long-term probabilities for large earthquake have been officially published by the Earthquake Research Committee of Japan and Working Group for California Earthquake Probabilities. However they have not been statistically tested yet due to the long earthquake recurrence interval. Based on the theories of the large and small sample and Bayesian approach, various models for calculating probabilities are introduced in this paper. Several statistical tests for forecast probabilities are also described here. Probabilistic predictions are experimentally performed for 93 sequences of small repeating events with nearly identical waveform with 34 models on the renewal process. Testing for probabilities with log-likelihood and Brier score shows that models on the Bayesian approach and the small sample theory with lognormal distribution are better than others on the large sample theory, and they are significantly better than models on the Poisson process. The sophisticated model, such as the time predictable model, including physical aspects should be tested by comparing with the better formulated models on the renewal process. Models for predicting aftershocks relating to the large earthquake are introduced and statistically discussed. It is shown that the Bayesian approach is also useful for predicting aftershock activities in early stage. Finally the author writes down the prospects of probabilistic prediction and hopes that the method for estimating short-term seismicity would be improved by experimental forecasts and tests, like Regional Earthquake Likelihood Models (RELM) and Collaboratory for the Study of Earthquake Predictability (CSEP), in the future.

抄録全体を表示

Recent availability of a large number of ocean bottom seismographs (OBSs), a large volume of air-gun array and a long streamer cable for academics provide several new findings of lithospheric scale structures in subduction zones. In this paper, I summarize results of active source marine seismic studies which have been mainly carried out by JAMSTEC in the Nankai seismogeinc zone and the Izu-Bonin subduction factory. One of the most striking findings in the Nankai trough is several scales of subducted seamounts/ridges. An important aspect, from a point of view of a seismogenic process, is that those structures strongly controled the rupture propagations of the 1944 Tonankai and the 1946 Nankai mega-thrust earthquakes. A 1000 km-long profile along the volcanic front in the Izu-Bonin arc provides new seismological constraints on growth of continental crust in an intra-oceanic arc; i.e., 1) bulk composition of the Izu-Bonin arc crust is more mafic than typical continental crust; 2) middle crust with seismic velocities similar to continental crust lies predominantly beneath basaltic arc volcanoes; 3) bulk composition beneath basaltic volcanoes changes little at thick and thin arc segments; and 4) a process to return lower crustal components to the mantle, such as delamination, is required for an arc crust to evolve into continental crust.

抄録全体を表示

Recent seismic expeditions with controlled sources in Japan provided important constraints on the deformation styles and physical properties of the island arc crust and uppermantle. The crustal structure in the Japanese Islands is characterized by an upper part with a large velocity variation (5.5-6.1 km/s), a middle part with a velocity of 6.2-6.5 km/s overlying a lower part whose velocity is 6.6-7.0 km/s. In many cases, most shallow microearthquakes are concentrated in the upper crust and upper half of middle crust while the lower half of middle crust and lower crust are reflective with very low seismicity. The uppermost mantle is characterized by a low P_n velocity of 7.5-7.9 km/s. Several observations on P_mP phase indicate that the Moho is not a sharp boundary with a distinct velocity jump, but forms a transition zone from the upper mantle to the lower crustal materials. A detailed structural section across the NE Japan Arc from intensive onshore-offshore profiling in 1997-1998 revealed crustal deformation associated with the Miocene backarc spreading of the Sea of Japan. The backbone range of this arc shows a pop-up structure formed by inversion tectonics due to the present compressive stress regime. Crustal thinning associated with the backarc spreading is very clear west of this pop-up structure where the crust deduces in thickness from 30 to 25 km. A section across the SW Japan arc elucidated the detailed subduction geometry of the Philippine Sea Plate and inland crustal evolution associated with processes of accretion and magmatic intrusion. The Outer Zone south of the Median Tectonic Line is characterized by northward dipping structures of accretionary complexes, while the lower part of the crust in the Inner Zone is quite reflective, probably modified and homogenized due to the magmatic intrusion at Cretaceous time. Clear structural images obtained for arc-arc collision zones in central Hokkaido and easternmost part of the SW Japan arc provides direct evidence of crustal delamination. The structure in Hokkaido strongly indicates that the delamination of the Kuril forearc occurs at its brittle-ductile transition zone.

抄録全体を表示

Since late 1990’s, a dense nationwide seismograph network has been constructed in the Japanese Islands with an average station separation of ∼ 20 km, which has produced the highest quality data in the world and contributed to enhance the understanding of seismotectonics and volcanotectonics there. Travel-time tomography using such high-quality data has provided two important constraints on water-transportation paths in subduction zones. One is the evidence for hydrous minerals in and immediately above the slab. The hydrated oceanic crust is imaged as a low-velocity layer to a depth of 40 km for the Philippine Sea slab and 70-130 km for the Pacific slab. Another low-velocity layer is revealed immediately above the Pacific slab down to a depth of ∼ 110 km, which might correspond to a hydrous layer through which water is carried to deeper depths. The other is seismological imaging of mantle upwelling. Mantle return flows induced by the subduction of the Pacific and Philippine Sea slabs are imaged in Tohoku and Kyushu, respectively, whereas that probably generated by the subduction of both the Pacific and Philippine Sea slabs is apparent in central Japan. A large upwelling in the upper mantle revealed in the Chugoku district might be the origin of Quaternary volcanism there.

抄録全体を表示

Seismic anisotropy is critical to infer geodynamics such as mantle convection, plate tectonics, and the evolutionary processes of the crust. S-wave splitting analysis has been extensively used in the anisotropy study of the Earth’s interior, including Japan subduction zone, but the resolution of the splitting data is too poor to investigate the detailed vertical structure. To obtain 3D anisotropic velocity structure beneath the Japan Islands, anisotropic tomography has been applied with P-wave travel-time data obtained by seismic networks deployed over the Japan Islands. The resultant 3D anisotropic structure shows that the P-wave anisotropy exists not only in the upper crust and mantle wedge but also in the lower crust and subducting slabs, where anisotropy has not been thoroughly studied. Although the P-wave anisotropic structure is basically consistent with S-wave anisotropy obtained from S-wave analyses, disagreement between P-and S-wave anisotropies is seen in the upper mantle beneath northeast Japan and the upper crust of the region along the Itoigawa-Shizuoka Tectonic Line. This result requires further studies to construct an anisotropy model that allows disagreement between P-and S-wave anisotropies, and/or revisiting the S-wave anisotropy structure with a sophisticated method that could provide a high-resolution structure of S-wave anisotropy.

抄録全体を表示

Receiver function analyses have been developed since the late 1970s. Receiver functions are calculated by deconvolving the vertical component from the horizontal components of teleseismic P waveforms in order to remove the complexities of the source time functions and to extract the structural information beneath seismic stations. The receiver functions are sensitive to impedance discontinuities of the earth’s interior, especially to the shear wave velocity discontinuities. Therefore, the receiver function analyses have advantages in estimating the shear wave velocity structure and imaging the shear wave velocity discontinuities. In this paper we introduce several basic topics of the receiver function analyses. Firstly, we discuss the method of the calculation of the receiver functions, pointing out weaknesses in the water level method and the multitaper method and introducing the extended-time multitaper method. Secondly, we describe the inversions of the receiver functions for the shear wave velocity structure. We show two examples with the genetic algorithm. Thirdly, we introduce the way and several examples to image the shear wave velocity discontinuities with the use of the receiver functions. Finally, we discuss issues in the receiver function analyses which have not been addressed yet but should be in the near future.

抄録全体を表示

Recent observational studies about structure of the earth by using scattered waves provided us with knowledge that elastic parameters like a velocity in the earth’s crust and upper mantle are heterogeneous in the range from few tens meter to kilometer scale. In addition, the heterogeneity varies from region to region with 3-4 orders in mean free path. Especially, in volcanic area, the scattering strength becomes extremely strong. In this review, we introduce researches for small scale heterogeneities. There are two major approaches to estimate it. One is analysis of envelope around direct wave arrival. Most of the studies are based on Markov approximation. The method can solve relatively longer scale length than wavelength of seismic waves. Another is in coda part using backward scattered waves. For this analysis, there are models depending on strength of heterogeneity (i.e., single-, multiple-scattering, and diffusion models). Based on the models, many studies have estimated relative scattering strength by using data in regional seismic network and/or seismic array. Spectral structure of heterogeneous property in short wavelength is useful to get character of a medium. Various approaches for scattered wave provide the spectral structure in wide wave number range. This may help us to understand mechanism of earthquake occurrence and volcanic eruption.

抄録全体を表示

Seismic reflection method is the most powerful and successful tool to understand the structure and the physical properties in the shallow crust, especially in sedimentary layers. Since it plays major role to explore hydrocarbon resources, the oil industry has been taking the initiative in developing new technologies for both data acquisition and processing. In 1990s, 3D survey became a routine technology. It provides the detailed geological model of complicated subsurface structures with high resolution. This paper reviews the recent technical trends on seismic reflection method mainly developed and utilized in the oil industry. The topics include high-density and high-resolution data acquisition and processing of 3D seismic data, direct estimation of physical properties of subsurface layers from seismic data, and time-lapse survey for monitoring subsurface fluid flow. Their concepts and techniques have been and will continue to be introduced to earth science and earthquake studies. New promising analysis method such as the waveform inversion and the virtual source method are also reviewed.

抄録全体を表示

Owing to significant developments in instrumentations, data processing methods and modeling schemes, prospecting electrical conductivity structure with the aid of the magnetotelluric method has become one of the standard geophysical techniques for elucidating dynamics of subsurface crustal activities such as earthquake occurrence and volcanic eruptions. In this paper, after brief introduction of the magnetotelluric method, recent 10-yaer progress of the method and its experiments in crustal active areas in Japan were reviewed. Since electrical conductivity is especially sensitive to the existence of the interstitial fluids and their connectivity, together with existence of high temperature areas, the survey results deepened our understandings for generation mechanisms of earthquakes, seismic faults, localized strain accumulations and volcanic eruptions.

抄録全体を表示

This paper addresses some issues related to relationship between lithospheric thermal regime and depth extent of seismicity. Thickness of seismogenic crustal layer correlates with surface heat flow in most intraplate seismic areas of the world. Although inverse relationship between heat flow and the base of seismogenic zone is obvious, quantitative relationships are less certain. Compilation of previous studies shows that temperatures at the base of seismogenic zone appear to be distributed from about 250° to 450° over a large depth interval, 5-30 km, at different tectonic settings. It supports that temperature is one of factors governing the focal depth. Variations in lithology, slip rate, pore pressure, and focal mechanism may account for the temperature difference. Geothermal gradient data and bottom depth of magnetized layers are also useful proxy to reveal the base of seismogenic zone. These multidisciplinary data may provide a useful indicator of lithospheric thermal structure and improve the correlation with depth limit of seismicity.

抄録全体を表示

Three-dimensional crustal structures in and around the Japanese Islands were calculated from gravity data. In this calculation we adopt the crustal structure along the Oga-Kesennuma line, obtained from explosion seismological observation, as a constraint. Gravity anomalies employed in this study is the Slab Residual Gravity Anomaly, which is computed by removing gravity contribution arising from subducting plates from the observed gravity anomalies. For computational ease, we divided the Japanese Islands into five blocks; Hokkaido, Tohoku, Chubu, Shikoku-Chugoku and Kyushu. In each block, we generated cross sections of gravity anomalies along latitudinal and longitudinal directions with 20 arc-min intervals. Two-dimensional Talwani’s method was applied to estimate density model structures. For each block, we applied the following steps. At first, we estimate a density model along profiles in latitudinal direction, and then estimate along profiles in longitudinal direction. We repeated these steps until crustal structures beneath cross points merged into some stable values. Combining structures calculated for adjacent blocks, we draw a three-dimensional distribution of surface layer, Conrad and Moho discontinuities. The crustal structure thus estimated was generally concordant with both existing explosion seismological and inversion crustal models, but at the same time showed some significant differences in intermediate scales.

抄録全体を表示

New global and Japanese seismic projects have promoted the seismological studies of deep Earth structure in Japan. Global and regional upper mantle structures have been extensively elucidated toward a finer scale. The mantle transition zones beneath cold subducted slabs and hot upwellings are widely investigated with new seismic experiments. The lower mantle and D” structures are examined with a global tomography and regional seismic arrays. Comparing to the mantle studies, the core studies are still inactive in Japan. Although seismic data increases, the geographical coverage is still uneven or the gap rather widens. New and challenging seismic experiments are required to overcome the previous achievements. In order to construct an interesting scientific plan, we need new points of view for the Earth dynamics collaborating with the other disciplines.

抄録全体を表示

Stagnant slabs have been identified at a depth of about 660 km from recent seismic tomography. In this study I briefly review 2-D numerical simulations of mantle convection, which have been constructed to reveal physical mechanisms of slab stagnation. The effects of the Clapeyron slope of the phase transformation from ringwoodite to perovskite + magnesiowüstite, trench retreat velocity, viscosity jump at the 660-km phase boundary, buoyancy due to a possible metastable olivine wedge inside a slab, and low viscosity due to grain-size reduction associated with the non-equilibrium phase transformation of olivine inside a slab on slab stagnation are presented. Then, I raise some remaining problems on slab dynamics in the mantle transition zone, and show directions toward future research.

抄録全体を表示

Phase transitions in pyrolitic mantle and MORB compositions have been investigated by high-pressure and -temperature experiments up to the conditions of the Earth's core-mantle boundary. Most of the seismic wave velocity discontinuities are primarily attributed to the phase changes in the pyrolitic mantle but some are possibly related to those in subducted MORB crust. Recent discoveries of the spin transition of iron and the post-perovskite phase transition have profound implications for the nature and dynamics in the lower mantle.

抄録全体を表示

The interior structures of planets and satellites have been investigated through various methods. Density is the most fundamental information for estimating the composition of solar system bodies. Moment of inertia expressing the mass concentration is a key for investigating the interior structure. It can be obtained from spacecraft flyby with a planet or a satellite. Interior density structure of a planet is evaluated from gravity field which is estimated from orbital tracking of spacecraft around the planet. KAGUYA first successfully obtained the accurate far-side gravity field of the Moon. Detailed interior structure can be estimated from seismic methods. Direct seismic measurements had been developed on the nearside of the Moon by the Apollo project. Seismograms of moonquakes, reflecting less interior dissipation, are different from those of the Earth. Radar sounders have been successful in measuring subsurface structure of the Moon and Mars. The state of the planetary interior, especially of the core, can be discussed using the measurement of variation of planetary rotation. The study of the planetary interior is now being extended into extrasolar planets.

抄録全体を表示

Seismic source process has been estimated using inversion of seismic waveforms and geodetic data. However, seismic source models for the same earthquake are often quite different from one another. We briefly noted on representations of the spatio-temporal slip-rate distribution, observation equation, constraints and property of near-source and tele-seismic data, and introduced two possible reasons for this discrepancy: geophysical observations are always contaminated by systematical error; observations are always insufficient to estimate a fine image of seismic source process. We are here concerned with the method for estimation of proper seismic source model from such observations. To mitigate the effect of systematic errors, it is essential to take the data covariance components into account. To obtain a fine and reasonable seismic source model from insufficient observations, it is necessary to introduce suitable prior constraints and then determine their optimal relative weights on the observed data using objective method.

抄録全体を表示

Recent progress on our understanding of earthquake rupture dynamics based on seismological observations is reviewed. One of the most important advances is the physical understanding of breakdown process that occurs at the propagating rupture front of earthquakes. The parameters such as breakdown stress drop, slip-weakening distance and fracture energy, which control the progress of dynamic rupture propagation, have now been able to be estimated. Fracture energy and hence slip-weakening distance are found to be scale dependent. To achieve such scale dependence, several different physical processes might microscopically develop, one of which dominates depending on the environmental situation at the source. Then its macroscopic behavior becomes quite simple and seems to be modeled with a single law.

抄録全体を表示

New types of slow earthquakes have been revealed associated with the subduction of the Philippine Sea plate in southwest Japan based on the development of the dense seismic observation network. At the transition zone, which lies at the deeper extension of the seismogenic zone, deep low-frequency tremors, deep very-low-frequency earthquakes, and short-term slow slip events occur along the slab strike. The short-term slow slip event is the main phenomenon as a stick-slip on the plate interface and other seismic phenomena might be triggered events reflecting the source size. Considering the area of migration and activity of the deep low-frequency tremors, which indicate a transient slip, we can divide the deep slow earthquake belt into segments, each of which has a regular recurrence interval of 6 or 3 months. In each episode, the tremor activity migrates according to the propagation of the transient slip within the segment. Sometimes the migration reaches to the neighboring segment, and a connecting fault rupture occurs. These migrating phenomena are very similar to the rupture process of mega-thrust earthquakes. On the other hand, shallow very-low-frequency earthquakes have been detected in the accretionary prism near the Nankai trough. These events are considered seismic slow slip events at the reverse fault system reflecting the deformation of the accretionary prism. Watching these slow earthquakes is important for the monitoring of the plate motion.

抄録全体を表示

This paper reviews recent progress on the scaling studies for earthquakes, including some slow earthquakes. While a standard scaling law for macroscopic parameters with geometrical similarity is valid for most cases, several studies suggest its limits and exceptions. The scaling of seismic energy has been a hot topic in this decade. We review new kinds of scaling studies on seismic source complexity, focusing on three aspects of earthquake evolution. Finally we compare the scaling law for slow earthquakes along the Nankai subduction zone with various similar phenomena worldwide.

抄録全体を表示

The stress inversion is a statistical method to estimate tectonic stress fields from a set of observed data. The methodology of stress inversion, originally developed in geology, is based on the Wallace-Bott hypothesis; the direction of fault slip is parallel to the resolved stress vector on a preexisting fault. In the conventional methods based on this hypothesis, fault slip data have been used as observed data. Since the constitutive property of the source region is unknown, we cannot relate fault slip with stress by means of strain. This is an essential problem in the fault slip data inversion. The conventional methods have another technical but inherent problem that inverted stress patterns depend on the way of area partitioning. The method of CMT data inversion using ABIC resolved both the essential and the technical problems in the fault slip data inversion. In this method, instead of fault slip data, we use CMT data of seismic events, which can be directly related with stress fields without any knowledge of actual processes in the source region. Hierarchic, flexible Bayesian modeling with hyper-parameters and rational model selection with ABIC enable us to resolve the model dependence of inverted stress patterns.

抄録全体を表示

Discovery of small repeating earthquakes and development of rate- and state-dependent friction laws have brought a modification of the asperity model that was originally proposed by Kanamori’s group. In the modified asperity model, ‘asperity’ is a synonym of ‘seismic patch’on the plate boundary. The area other than asperities on the boundary is supposed to slip aseismically. Careful analyses of interplate large earthquakes also support the model. This model gives a basis of current long-term earthquake prediction. Further detailed analyses of the relation between the pore pressure and rupture process and the interaction among the asperities and aseismic areas are mandatory for the improvement in the accuracy of the prediction.

抄録全体を表示

History of studies of intraslab seismicity and its generation mechanisms, since the work by Wadati, is reviewed. Through the initial stage of studies of morphology of the intraslab seismicity, there followed the stage of discussion of focal mechanisms and terminal depths by mechanics and temperature of the slab. In 1970s, double seismic zones were discovered and their generation mechanisms were discussed in terms of bending or thermal stresses. In recent years, to overcome a difficulty that very high pressure prevents intraslab seismicity, dehydration embrittlement and phase transformation have been invoked for the mechanisms of intermediate and deep earthquakes, respectively. If the intermediate seismicity represents dehydration, it may give us a key to understand the distribution of fluids to the upper plate and to the seismogenic interplate thrusts, and finally to understand tectonics and volcanism in subduction zones.

抄録全体を表示

I review recent progresses in Japan about the process by which intraplate earthquakes are generated, in particular about the stress accumulation process on intraplate earthquake faults. It is thought that a part of convergent rates between subducting oceanic plates and the inland plate is consumed in the intraplate region. The deformation in the intraplate region in Japan, which occurs mainly in the concentrated deformation zone (NKTZ), results from the weak zone in the lower crust that is characterized by a low viscosity region. Low velocity and high attenuation anomalies are detected in the weak zone. Anomalous stress fields are also estimated around an active fault or fault zone, which are thought to be due to the deformation in the lower crust immediately beneath the active fault or fault zone. These results suggest that intraplate earthquakes are generated by the heterogeneity in the lower crust.

抄録全体を表示

We review recent advances in quantification of the sources of volcano-seismic signals. Since the 1990s, there have been remarkable advances in quantitative understanding of the sources of volcano-seismic signals. These were made by integrated contributions from observation, analysis, and modeling studies in volcano seismology. We focus on the following topics in this review paper: Broadband seismic observations on active volcanoes, waveform inversions of volcano-seismic signals, and source models based on resonances of a fluid-filled crack, bubble dynamics, and repeated fracture and healing of magma. Broadband seismic observations revealed the general existence of very-long-period seismic signals with oscillation periods longer than a few seconds beneath active volcanoes. Development of waveform inversion methods enabled us to estimate source mechanisms of volcano-seismic signals, which indicate that crack geometry is the most common in volcano-seismic sources. Source models have been successfully used to interpret dynamic interactions between volcanic fluids and the surrounding solid rock. Quantitative studies of volcano seismology are now reaching a mature stage, and links with geological and geochemical studies will be important future directions to achieve a better understanding of magmatic and hydrothermal processes beneath active volcanoes.

抄録全体を表示

There are some systematic variations in size and recurrence interval of earthquakes in subduction zones. Such variations, especially variation in recurrence time depending on earthquake size, are incorporated in hazard predictions for earthquakes in many countries including Japan, although the physical mechanism has not been revealed yet. We model an interplate earthquake as release of slip deficit with high slip rate. The slip deficit accumulation occurs on an area where slip velocity is significantly lower than the surrounding area mainly due to mechanical coupling of plate boundary. Spatial and temporal variation in slip on a plate boundary can be described mathematically with equations for stress change due to slip, a relation between stress, strength and slip velocity, and an evolution law of strength. Numerical simulations based on such model with heterogeneous frictional property qualitatively reproduce the variation pattern of size and time interval of great interplate earthquakes in some subduction zones. The recurrence time variation depending on size for the Nankai subduction zone arises from the specific layout of inhomogeneities of the region. Since such variations are seen in other subduction zones, there may exist other mechanisms related to specific or general patterns of heterogeneity in frictional properties.

抄録全体を表示

We review the recent researches of numerical simulations on faulting, which is interpreted in this paper as the evolution of the state of the fault plane and the evolution of fault structure. The theme includes the fault constitutive (friction) law, the properties of the gauge particles, the initial phase of the rupture, the dynamic rupture process, the interaction of the fault segments, the fault zone dynamics, and so on. Many numerical methods have been developed : boundary integral equation methods (BIEM), finite difference methods (FDM), finite or spectral element methods (FEM, SEM) as well as distinct element methods (DEM), discrete element methods (again DEM) or lattice solid models (LSM). The fault dynamics should be solved as a complex non-linear system, which shows multiple hierarchical structures on its property and behavior. The researches have progressively advanced since the 1990’s both numerically and physically thanks to high performance computing environments. The interaction at small scales is modeled to provide a large scale property of the fault. The dynamic rupture has been actively studied especially for the effect on the fault geometry evolution or due to the existed fault structure. The (quasi-)static and the initial processes of the fault movement have been explored in a seismic cycle. The effect of fluid or heat has been taken into account in the mechanics. All these efforts help us to understand the phenomena and the unified understanding (simulation) over different spatio-temporal scales is more and more expected.

抄録全体を表示

Recent high-resolution observations of crustal movements have revealed the occurrence of slow slip events (SSEs) along the deep parts of subduction interfaces. This report reviews the possible models for SSEs using rate- and state-dependent friction laws. SSEs can be modeled under the condition close to the stability transition. The triggered SSEs can be modeled considering a conditionally unstable cell. The self-triggered SSEs can be reproduced in a region that is unstable but close to being stable. In this case, the ranges of the constitutive law parameters for reproducing SSEs are limited. Further, SSEs can also be modeled by considering the frictional property of an unstable-stable transition zone that exhibits velocity weakening at low slip velocity and velocity strengthening at high slip velocity; this model is proposed on the basis of the results obtained in an experiment using halite around an unstable-stable transition zone. By considering this frictional property, Shibazaki and Shimamoto have reproduced short-term SSEs that are similar to the observed SSEs. This friction law needs to be verified experimentally under conditions that are relevant to the fault zones of SSEs. It is theoretically expected that for slip failure processes the propagation velocity is proportional to the slip velocity. This relationship appears to hold for observed SSEs. Therefore, SSEs can be regarded as slip failure processes occurring at deep subduction plate interfaces.

抄録全体を表示

This article describes construction of the characterized source model for strong motion prediction of scenario earthquakes. Based on the heterogeneous slip models obtained from strong motion data, an empirical relationship between seismic moment and asperity area size are observed. The asperity area coincides to the strong motion generation area that generates broad-band ground motions. Following these observations, the characterized source model is constructed based on the asperity model. The `recipes` are also reviewed for strong motion prediction.

抄録全体を表示

The 2003 Tokachi-oki earthquake (M_w = 8.0) caused the severe damage to oil storage tanks by liquid sloshing, especially at Tomakomai in Hokkaido, northern part of Japan, where large liquid sloshing was excited by the long-period strong ground motions. In view of severe damage to oil storage tanks and mega earthquake threats such as Tokai, Tonankai, and Nankai earthquakes, many studies have been carried out associated with source models of huge earthquakes, S-wave velocity structural models, and simulation methods. Based on these knowledge, strong ground motions including a long-period range, from a few to over ten seconds, were predicted by several researchers for future earthquakes in the Nankai Trough region and provided to Japan Society of Civil Engineers and Architectural Institute of Japan in order to examine seismic safety of various structures such as high-rise buildings, base-isolated buildings, suspension bridges, and oil storage tanks. Since validity of predicted ground motions are hardly verified in the long-period range, further works are expected about heterogeneous source models, three-dimensional subsurface structural models, and the effects of their fluctuations on long-period ground motion characteristics.

抄録全体を表示

The Japan islands are in a complex tectonic setting with various subducting plates, and most of urban areas are located over large-scale sedimentary basins. Since the sediments filling basins amplify ground motions and their velocity structures complicate the propagation of seismic waves, it is important for the prediction of strong ground motion and seismic hazard to determine the three-dimensional (3-D) velocity structures of these urban basins. This importance motivated various organizations to carry out extensive geophysical experiments and geological investigations, and velocity structure models are being constructed all over Japan. We have already proposed a standard procedure for modeling a regional 3-D velocity structure in Japan, simultaneously and sequentially using various kinds of datasets such as the extensive refraction/ reflection experiments, gravity surveys, surface geology, borehole logging data, microtremor surveys, and earthquake ground motion records. We applied the procedure to the Tokyo metropolitan area for constructing a reference 3-D velocity structure model. As this application confirmed the validity of the standard procedure, it is then applied to the central, eastern, and western parts of Japan in 2006 to 2008, to construct the Japan Integrated Velocity Structure Model. The velocity structure and source modelings have been dramatically improved after the 1995 Kobe earthquake with the advance of ground motion simulations based on substantial data of strong motion observation. The quality of broadband ground motion simulations enable us to predict realistic strong ground motions. The National Seismic Hazard Maps released in 2005 as a result of comprehensive research on long-term evaluation of earthquake occurrence and strong ground motion prediction are expected to be proved useful in earthquake disaster mitigation and scientific outreach to the public. It is also quite important for the simulation of long-period ground motion and its seismic hazard to validate the three-dimensional (3-D) velocity structure for the whole Japan islands. Since we are threatened by future earthquakes whose probabilities of occurrence are estimated to be high in long-term evaluation, long-period ground motions from future Tokai, Tonankai, Nankai, and Miyagi-oki earthquakes and their response spectra will be computed by using this Japan Integrated Velocity Structure Model, and combined into the Long-Period Ground Motion Hazard Maps in 2009 and later by the Headquarters for Earthquake Research Promotion. These hazard maps for the megathrust earthquakes may be proved to be touchstones of strong motion seismology.

抄録全体を表示

Effects of site responses are quite influential on strong ground motion characteristics. Dense strong motion networks, most of which were deployed after the 1995 Hyogo-ken Nanbu earthquake, have contributed significantly to recent site response studies. In this paper, we reviewed published papers after the 1990’s on site responses most of which were objected to broadband strong motion predictions in time domain. We focused our attention mainly to one-dimensional theoretical studies and empirical ones. One-dimensional theoretical site responses of horizontal motions for S waves and vertical motions for P waves were main concerns in the 1990’s. Vertical array records or borehole records were inverted for both sediments’ Q_p and Q_s based on one-dimensional (1-D) wave propagation theory assuming vertical incident P and S waves, respectively. In the 2000’s three-component site responses for both P and S waves based on 1-D wave propagation theory assuming obliquely incident P and S waves became to be a primary target. The three-component site responses for P and S waves were applied to strong motion predictions by statistical Green’s function method and provided several successful examples. However, we have concluded that more quantitative study is necessary to understand the mechanisms of three-component site responses, especially for three components of P waves and the vertical component of S waves. We also summarized recent studies on nonlinear site responses, in which observed records were simulated well by revised equivalent linear analyses, time-domain total stress analyses, or effective stress analyses. For empirical site response studies, estimation methods for both Fourier amplification spectra and phase spectra of site responses have been proposed and revised since the 1990’s. In addition, a new method to estimate empirical site responses in time domain using a wavelet analysis has been recently proposed. Estimated empirical site responses were applied to strong motion predictions through statistical Green’s function method and showed their good applicability. Empirical site response factors obtained from records of horizontal components for S waves were inverted for 1-D structures at observation sites. The agreements between them were very good or good for about one half of the sites. This result suggests that empirical site responses have advantage over 1-D theoretical site responses, because they include the site effects which cannot be represented by simple 1-D models. However, empirical site responses can be obtained only at strong motion observation sites. We should study the relationship between 1-D three-component site responses assuming obliquely incident P and S waves with the corresponding empirical site responses in a more quantitative manner. Then we will see more clearly the contribution of the basin effects, as well as the source and path effects.

抄録全体を表示

Recent studies on ground motion attenuation relations in the United States and Japan are reviewed. In the United States, the Next Generation of Ground Motion Attenuation Models (NGA) Project was initiated to develop the updated relations. In the proposed models for peaks and response spectra, effects of faulting style, hanging wall, fault top depth, nonlinear site amplification and basin depth are incorporated. In Japan, owing to recent development of strong motion instrumentation such as the K-NET, many updated attenuation relations not only for peaks and spectral response but also for instrumental seismic intensity and energy spectra have been individually developed. The terms considered in the relations are similar to those for the models of the NGA project. The variance of ground motion in the attenuation relations is also being discussed earnestly because the variance controls the results of seismic hazard analyses. Further efforts should be continued to improve the attenuation relations in response to increasing near-filed ground motion recordings and progress of theoretical studies on strong ground motion.

抄録全体を表示

Monitoring of crustal movements with a dense continuous GPS network called GEONET started in 1990’s in the Japanese islands and drastically changed observational studies of crustal movements. This paper briefly summarizes new findings and knowledge about seismotectonics of the Japanese islands obtained from observation of crustal movements since 1990’s. High precision (∼1 mm/yr), good spatio-temporal resolution (25 km and 1 day), and dramatic increase of data (>1200 daily coordinate solutions) due to continuous observation have enabled us to monitor minute crustal movements related to tectonic loading processes at subduction zones as well as inland active faults in a quantitative manner. On the other hand, continuous GPS data cover only a rather short time period and comparison of these data with those from conventional surveys and geological investigation show both similarities and dissimilarities. Integrated analysis of observation data obtained from different time scales is an important key to understand the whole tectonic processes. As for seismic deformation, we can separate postseismic transient motion from coseismic displacement with continuous observation. Now coseismic deformation can be resolved very quickly and precisely, facilitating the estimation of a fault model. Postseismic deformations were studied in detail using GPS network data, and their significance in seismic cycle has been clarified. Spatio-temporal distribution of afterslip provides information about frictional property on the fault plane and its heterogeneity, which is important in understanding seismogenic processes. Episodic slow slip events not preceded by major seismic events was first discovered by the Japanese continuous GPS network, and it has been shown slow slip is a common phenomenon in various subduction zones although their physical mechanism and tectonic implications have not been elucidated yet. Continuous GPS observation has significantly widened our scope about active deformation processes of the Japanese islands and provided a new opportunity to make comprehensive understanding as well as a quantitative forecast of crustal activities.

抄録全体を表示

Studies in Japan on earthquake source processes and tsunami generation using tsunami data during recent 15 yeas are reviewed. The new tsunami observation systems including ocean bottom tsunami meters, GPS tsunami meters, and ultrasonic wave meters, have been developed recently. The methods using the tsunami waveform data observed by those advanced systems have also been developed and the detail source processes of the recent large earthquakes are studied using those methods and data. Studies on 2004 Sumatra-Andaman great earthquake, which caused the worst tsunami disaster and affected the countries around the Indian Ocean, are reviewed. Next, studies on source models of historical large earthquakes using tsunami data are reviewed. Researches on the mechanism of tsunami earthquakes are also reviewed. Finally, tsunami studies on submarine landslides triggered by earthquakes are reviewed.

抄録全体を表示

Recent studies on modern, historical and pre-historical tsunami deposits are reviewed. Tsunami deposits are boulders, sand and mud deposited in coastal areas during the passage of a tsunami. Modern tsunami deposits have been studied to identify the landward limit of the tsunami inundation and to estimate the tsunami flow pattern. Buried tsunami deposits are unique data for recurrence interval and size distribution of historical and pre-historical tsunamis that have been generated by large earthquakes along subduction zones. For identifying buried tsunami deposits, we have studied to establish criteria for distinguish tsunami deposits from the other coastal event deposits. A series of remarkable works was done along the northeastern coast of Hokkaido facing the Kuril Trench. Time-space distribution of tsunami deposits suggested that unusually large earthquakes have been occurred in past 7000 years with about 500 years intervals. The last event occurred in the 17th century. The event is regarded as a multi-segment earthquake that have been never known in this area.

抄録全体を表示