Shigeki Horiuchi

TOMOGRAPHIC IMAGING OF P AND S WAVE VELOCITY STRUCTURE BENEATH NORTHEASTERN JAPAN

The seismic body wave tomography method has been improved and extended to adapt to a general velocity structure with a number of complexly shaped seismic velocity discontinuities (SVDs) and with three-dimensional variations in the velocities in the modeling space. An efficient three dimensional ray tracing algorithm which iteratively uses the pseudobending technique and Snells law is developed. The large and sparse system of observation equations is solved by using the LSQR algorithm. This method is applied to 18,679 arrival times from 470 shallow and intermediate-depth earthquakes in order to study P and S wave tomographic images beneath northeastern Japan. In addition to first P and S wave arrivals, clear later arrivals of SP waves converted at the Moho and PS and SP waves converted at the upper boundary of the subducted Pacific plate (UBPP) are also used in the inversion. The UBPP, Conrad and Moho are taken as three SVDs, and their depth distributions obtained by previous studies are used. High-resolution P and S wave tomographic images down to a depth of 200 km have been determined. Large velocity variations amounting to 6% for P wave and 10% for S wave are revealed in the crust and upper mantle. In the crust low-velocity (low-V) zones exist beneath active volcanoes. In the upper mantle the low-V zones dip toward the west from the volcanic front. A high-velocity (high- V) zone corresponding to the subducted Pacific plate is clearly delineated. Most earthquakes in the lower plane of the double-planed deep seismic zone are found to occur in relatively high-V areas. The obtained tomographic images are also found to explain other seismological observations well.

Seismic structure of the northeastern Japan convergent margin: A synthesis

Many studies recently made on the basis of seismic observations have revealed a detailed structure of the crust and the upper mantle beneath the northeastern Japan arc and its relationship to seismic and volcanic activity. Spatial distributions of the depths to the Conrad and the Moho discontinuities, estimated from shallow earthquake data and seismic explosion data, show that both discontinuities are deep in the middle of the land area and shallow toward the coastlines of the Japan Sea and the Pacific Ocean. The Pn velocity has a lateral variation; it is as low as ∼7.5 km/s beneath the land area, while that beneath the Japan Sea and the Pacific Ocean is 8.0–8.2 km/s. It changes abruptly at the transition zones, which are located along the coastlines. Precise structure and location of the subducted Pacific plate beneath the land area is inferred from converted or reflected seismic waves at the top or bottom of the plate. The Pacific plate is composed of a thin (∼5 km) low-velocity upper layer and a thick high-velocity lower layer, its total thickness being 80–90 km. The upper plane seismicity of the double seismic zone is confined to the thin low-velocity upper layer, which probably corresponds to the subducted former oceanic crust. The lower plane seismicity lies at the middle of the high-velocity lower layer, and the lower half of the plate below it is incapable of generating earthquakes. The shallower portion of the upper surface of the plate beneath the Pacific Ocean, along which major seismicity with low-angle thrust faultings is actually occurring, is also located by seismic observations on land and in the sea. The Pacific plate subducts at an extremely low angle of ∼5° for the first ∼25-km depth, and then the dip steepens rather abruptly to ∼30°. Normal-fault type events at the top of the plate have not been detected in the portion where the downward-bending is the largest, but have been ditected near the trench axis, where it is rather small. Tomographic inversions for seismic velocity structure clearly delineate the inclined high-velocity Pacific plate with a thickness of 80–90 km and low-velocity zones in the crust and the mantle wedge beneath active volcanoes. Seismic attenuation tomography also shows similar zones of low-Q value beneath active volcanoes, although its spatial resolution is much lower. The low-velocity zones with 2–6% velocity lows are continuously distributed from the upper crust just beneath active volcanoes to a depth of 100–150 km in the mantle wedge, and are approximately parallel to the dip of the underlying Pacific plate. These low-velocity zones probably reflect the pathway of magma ascent from a depth in the mantle wedge to the Earths surface, corresponding to a portion of the subduction-induced secondary mantle wedge flow.

An Automatic Processing System for Broadcasting Earthquake Alarms

In Japan, most of the damage caused by large earthquakes is concentrated within areas less than about 100 km from the focal regions. We have developed an earthquake alarm system that determines earthquake parameters within a few seconds of the P wave’s arrival at the closest station, and then transmits the earthquake information before the S -wave arrival in areas of possible serious earthquake damage. Since an earthquake alarm system requires the determination of reliable earthquake parameters as quickly as possible, it is unreasonable to wait until waveform data from numerous stations have been collected for analysis. For this reason, we have developed a novel method of determining the hypocentral location by using the arrival times for only a few stations, as well as the lack of P -wave arrivals at other stations at a given time. The use of not-yet-arrived data makes it possible not only to determine reliable hypocenter parameters within a few seconds, but also to detect erroneous arrival-time readings and remove them automatically. Since the available waveform data increases with time, the system was designed to redetermine earthquake parameters every second. Our method was deployed in a real-time system starting in July 2002. The real-time system locates 10–20 events per day, which includes a few felt earthquakes occurring in and around Japan. It was shown from the waveform data for about 100 felt earthquakes that almost all the events, except those far from the network, could be located within a few seconds, when most of seismic energy has not even arrived at some stations close by. We recently started widespread broadcasting of earthquake information in real-time by using a satellite transmission system.

Three-dimensional attenuation structure beneath the northeastern Japan arc estimated from spectra of small earthquakes

Abstract A fine-scale three-dimensional attenuation structure beneath the northeastern Japan arc has been obtained using a joint inversion for source parameters, site response and Q values. Data used for this study are P wave spectra of microearthquakes occurring in the subducting Pacific plate. The subducting plate is found to have higher Q values than the overlying mantle. There is an along-arc variation in the distribution of low-Q zones in the crust and the mantle wedge within the study area. For the southern part of the study area, low-Q zones are distributed in the crust beneath the volcanic front. They extend into the mantle wedge and become deep toward the west, that is, toward the backarc region. On the other hand, the northern part of the study area has low-Q zones at two depth ranges; at shallower depths just beneath the volcanic front and at the deeper part of the mantle wedge beneath the backarc region. The region without active volcanoes along the volcanic front has relatively high mantle Q values. Seismic activity in the lower plane of the double-planed deep seismic zone is high beneath this relatively high-Q region of the mantle wedge. Spatial distribution of low-Q zones found in the present study is nearly consistent with that of the low-velocity zones estimated from studies of travel times tomography. Partial melting zones in the mantle wedge, whose existence has been suggested from the study of seismic velocity structure and laboratory experiments, are located in the low Q zones obtained in the present study. These results suggest that spatial distribution of Q values is closely related to heterogeneous distribution of temperature in the mantle wedge.

Seismic velocity structure of the crust beneath the Japan Islands

Abstract More than 13,000 arrival times of 562 local shallow earthquakes selected from the Japan University Network Earthquake Catalog are used to investigate the seismic velocity structure of the crust beneath the Japan Islands. We simultaneously determined the hypocenter parameters, P- and S-wave station corrections and depth distributions of the Conrad and Moho discontinuities beneath the whole of the Japan Islands by applying an inversion method. The P- and S-wave station corrections show similar distribution patterns. They are negative (relatively late arrivals of seismic waves) in Hokkaido and positive (relatively early arrivals of seismic waves) in western Japan. In eastern Honshu, they are positive along the coast of the Pacific Ocean and negative in the land area and along the coast of the Japan Sea. The Conrad discontinuity is located at a depth range from 12 to 22 km. In Hokkaido, the Conrad is shallow in the eastern pan and becomes deep toward the west. In NE Honshu the Conrad is deep beneath the land area, and becomes shallow toward the surrounding seas. In western Japan, the Conrad is shallow in the northern part and becomes deep toward the south. The Moho is located in a depth range from 25 to 40 km. In Hokkaido, the Moho has the shape of a circular cone with a maximum depth of 36 km. In most parts of the Japan Islands, the Moho is deep beneath the land area and becomes shallow toward the surrounding seas. The Moho is deepest, up to 40 km, beneath the central part of the Chubu District.

3-D seismic velocity structure of the crust and the uppermost mantle in the northeastern Japan Arc

3-D seismic velocity structure of the crust and the uppermost mantle beneath the northeastern Japan Arc is investigated by using arrival time data from local earthquakes. We use a velocity model composed of three layers corresponding to the upper crust, the lower crust and the uppermost mantle, respectively. Taking into account the observed regional variation of Pn-velocity, the uppermost mantle is further divided into three parts by two Pn-velocity boundaries near the coasts of the Japan Sea and the Pacific Ocean. The velocities within the upper crust, the lower crust and the three parts of the uppermost mantle are assumed to be constant. Locations of two Pn-velocity boundaries and depth distributions of the Conrad and the Moho discontinuities are expressed by continuous functions of unknown parameters. Station corrections and hypocenters are also introduced to be unknowns. These unknown parameters are determined by inverting arrival time data of P- and S-wave first arrivals and clear later arrivals of Pg- and P∗-waves. The Pn-velocity boundary between the land and the Pacific Ocean is located approximately along the Pacific coastline and that between the land and the Japan Sea is nearly along the Japan Sea coastline. The Conrad and the Moho discontinuities are at depths ranging from 14 to 20 km and from 27 to 36 km, respectively. The Conrad and the Moho depths have similar spatial distributions. They are deep beneath the land and become shallower toward the western and the eastern coastlines. Beneath the land, they are shallow in the central part of the Tohoku District and become deeper toward both the north and the south directions. In the north, they become shallow again.

Seismic activity and deformation process of the overriding plate in the northeastern Japan subduction zone

Abstract We estimated the temperature distribution within the crust of the northeastern Japan arc from P wave velocity perturbations obtained by travel time tomography. By comparing the estimated temperature distribution with the focal depth distribution of shallow, precisely relocated microearthquakes, we found that the brittle to ductile or stick-slip to stable-sliding transition occurs at the ∼400°C isotherm and that the transition depth has considerable lateral variations. The brittle seismogenic zone, the upper portion of the crust, becomes locally thin in the P wave low-velocity areas, where the temperature is estimated to be relatively high. Concentration of shallow microearthquakes, high topography and relatively large contractile deformation of the crust are also observed in these low-velocity areas. Active faults are not distributed in the low-velocity areas but lie just along the edge of those areas or outside them. All these observations suggest that earthquake occurrence and deformation within the crust is governed, to a considerable degree, by the thermal regime of this volcanic arc, which is characterized by a horizontally inhomogeneous distribution of temperature.

Discrimination of fault planes from auxiliary planes based on simultaneous determination of stress tensor and a large number of fault plane solutions

It is well known that there is a large difference in focal mechanism solutions of events even if they occur within a certain small area. Considering earthquake-generating stress to be uniform in a studied area, we developed a new method for simultaneously determining the stress tensor and the orientation of fault planes for many events by the use of polarity data of P waves. First, the number of inconsistent stations is calculated for various values of three parameters in the focal mechanism solution, with intervals of 10°. This calculation is made for all the events used. Then, four parameters defining the stress tensor are determined by a grid search using the data calculated above with the assumption that the slip direction in the fault plane should be parallel to the direction where the shear stress on the plane becomes maximum. A numerical test is made by applying the present method to an artificial data set composed of 25 events with 25 × 25 readings. It is found that a stable solution of the stress tensor is determined by the inversion and about 60% of fault planes are distinguished from auxiliary planes. There are no events whose auxiliary planes are incorrectly determined as fault planes. This method is used to estimate the earthquake-generating stress in the aftershock area of the 1984 western Nagano prefecture earthquake, central Japan. The results obtained show that the maximum principal stress is horizontal in the direction of N80°W-S80°E. There are strike-slip events having fault planes almost perpendicular to the fault plane of the main shock.

Comment on “Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records” by Aldo Zollo, Maria Lancieri, and Stefan Nielsen

] An early warning system should determine the loca-tion and magnitude of an earthquake as rapidly as possiblein order to broadcast an alarm to regions that will undergosevere ground shaking. It was recently claimed by Zollo etal. [2006] that earthquake size could be determined fromonly the first 2-seconds of P- or S-wave strong-motion data;this represents a fraction of the rupture time for larger M >7 events. Using this relatively short amount of data, themethod of analysis was to find the peak ground displace-ment (PGD), which was reported to scale with earthquakemagnitude; such rapid information would play an importantand much needed role in an earthquake early warning(EEW) system. Here we perform a similar analysis onstrong-motion data from the KiK-net and K-NET arrays inJapan and find no compelling evidence that the peak grounddisplacement during the first couple of seconds of P-wave isrelated to the eventual size of a large earthquake.[

Simultaneous estimation of attenuation structure, source parameters and site response spectra—application to the northeastern part of Honshu, Japan

Abstract We have developed an inversion method for estimating Q structure by the use of a large number of waveform data from earthquakes. We divide an area to be analyzed into small blocks and simultaneously estimate the Q value in each block, source parameters for each event and site response spectrum for each station. The source spectrum is assumed to be expressed by the ω2 model, and values of corner frequencies and low-frequency levels of source spectra are estimated. Amplitudes are calculated by taking the orientation of focal mechanism solutions into account. A trade-off among estimated parameters is tested by numerical experiments using three sets of artificial data. In a case when artificial data include error of 15° in the orientation of the P axis of focal mechanism solutions, the inverted result shows that the original checkerboard pattern of Q structure is well reconstructed, although estimated Q −1 values are about 10% higher than those of the original one. The result without the corrections for radiation effect shows that the checkerboard pattern is only reconstructed in the blocks having a sufficient number of ray paths. We applied this inversion method to a real data set in the northeastern part of Honshu, Japan. Estimated Q P structure shows that low Q P regions are distributed beneath the volcanic front and they move towards the west with increasing depth. Previous studies of three-dimensional seismic velocity structure in this area show that P-wave low-velocity regions exist in the crust beneath the volcanic front and they are inclined towards the west in the mantle wedge. The low Q P regions estimated in this study coincide in general with the P-wave low-velocity regions.