S31A-0196
Crustal heterogeneities deduced from wideband and Network MT measurements around the Niigata-Kobe Tectonic Zone, Chubu District, Japan
Wideband magnetotelluric (MT) soundings were carried out around the concentrated deformation zone, Chubu District, Japan (NKTZ: Niigata-Kobe Tectonic Zone, e.g. Sagiya et al., 2000). The NKTZ becomes one of important target areas in "the 2nd new Program of and Observation for Earthquake Prediction" (Hirata, 2004). A multidisciplinary research around the NKTZ, especially the Atotsugawa fault, using dense GPS, seismological observations and investigation of crustal resistivity structure has been started since 2004. In Oct. 2004, we obtained the MT data at 30 sites across the central part of the NKTZ in which the Ushikubi, Mozumi-Sukenobu, Atotsugawa, Takayama-Oppara faults are located. Apparent resistivity and phase in TM mode, phase in TE mode and tipper were used for two-dimensional inversions. Obtained model shows following inhomogeneities in the middle and lower crust. (A) Beneath the NKTZ, an almost 10km thick resistive block is detected. (B) This resistive block gets thin gradually to approximately 5km thickness at the Atotsugawa and Mozumi-Sukenobu faults which are centrally located in the NKTZ. (C) In the lower crust, beneath these faults, conductors are found, respectively. These conductors may represent fluids-filled week zones (e.g. Iio et al., 2002). To reveal heterogeneity along the fault plane, we carried out MT survey along the Atotsugawa fault in Oct. 2005. A seismic gap and a creep-like crustal movement were observed along the Atotsugawa fault (e.g. Ito et al., 2007; Geographical Survey Institute, Japan, 2002). Obtained preliminary inversion result shows lateral inhomogeneity correlated with heterogeneity in seismicity along the fault. The relatively low seismicity region on the Atotsugawa fault plane seems to be imaged as resistive body. In addition to these conventional wideband MT surveys, we have performed Network-MT surveys across and along the NKTZ since Dec. 2005, in order to determine the deeper structure down to the upper mantle in the wider area of Chubu district. The EM response functions across the NKTZ from Noto Peninsula to the south of Nagano prefecture show remarkable phase value enhancements in the longest period range (about 10000 s) in Noto area and in the shortest period range (about 10 s) in the NKTZ. The former and the latter probably indicate upper mantle and middle crust conductors, respectively. A 2-D inversion result along the Atotsugawa Fault indicates that seismogenic zone shallower than 10km is relatively resistive and the slipping zone beneath it down to 20km is relatively conductive. This feature is almost the same as that obtained from the conventional MT survey. Beneath the conductive zone, a moderately resistive layer is revealed.
S31A-0197
Crustal heterogeneities on electrical resistivity in the source regions of the 2004 Niigata earthquake and the 2007 Noto earthquake
The 2004 Niigata Earthquake of Mj6.8 occurred on 23 Oct. 2004 in the backarc area of the central main Japan island. Focal mechanism of this earthquake is the reverse fault type with a strike of approximately N35E and a dip of approximately 60 degree. In long-term velocity field revealed by an analysis of Japanese nationwide GPS array (GEONET) (Heki and Miyazaki, 2001), North Eastern Japan(NEJ) and South Western Japan(SWJ) respectively acts almost rigidly and large strain rate is restricted in rather narrow zones. Niigata-Kobe Tectonic Zone (NKTZ: e.g. Sagiya et al., 2000) forms a motional boundary between NEJ and SWJ running along the back-arc side of the central Japan, where the present earthquake focal area is located. Across its source region, wideband magnetotelluric (MT) survey was performed in Nov. and Dec., 2004 in southern and northern part of the focal area, respectively. Since phase tensor analysis indicates that 3-D features are dominant in the longer periods for both of the profiles, we tried to determine a 3-D structure by inverting full-tensor data for both profiles with the aid of a DASOCC 3-D code. Results from the 3-D interpretation are as follows: (1) the main shock together with aftershocks occurred in relatively resistive portion beneath the surface conductive thick sedimentary layer. (2) The north profile is found to be located on EW trended ridge structure. (3) Lower crustal moderately high conductivity area exists beneath eastern side of main- and aftershock zones. 3 years after the Niigata Earthquake, a damaging earthquake (Mj6.9) occurred on 25 March 2007 near the west coast of the Noto Peninsula which is located in the north of the NKTZ. We also performed wideband MT survey for the onshore area of the source region immediately after the mainshock. We first constructed 2-D resistivity models along five profiles using only the TM mode responses. Significant characteristics of the resistivity models are: (1) Beneath the mainshock hypocenter, there is a conductive body which spreads to the eastern edge of the active aftershock region. (2) A resistive zone is located in the gap of the aftershock distribution between the mainshock hypocenter and the largest eastern aftershock. (3) One of the largest aftershock occurred at the boundary of the resistive zone described above. These results suggest that the deep conductors represent fluid- filled zones and the lateral heterogeneity could have controlled the slip distribution on the fault plane. In addition to these 2-D results, we will present a preliminary 3-D interpretation.
S31A-0198
Fault Zone Behavior Observed from Crossing Fault 7-level TCDP Boreholes Seismometers
We investigated the waveforms recorded by 7-level borehole seismometers (Taiwan Chelunpu-Fault Drilling Project, TCDP 7-level BHS), which were placed from the depth of about 950m to 1300m crossing the fault zone associated with the 1999 Chi-Chi earthquake. The seismometers were installed over the depth range of hanging wall and footwall with the depth interval of about 50-60m, and a seismometer at the depth of 1111km as identified to be the fresh slip zone for the 1999 Chi-Chi earthquake. The TCDP 7-level BHS recorded nice waveforms for micro events with magnitude down to -0.5. The detail velocity structure in the boreholes was constructed from logging data. Even though the 7-level seismometers are only 50 m apart in depth, the seismometers located above the low velocity zone (LVZ), which might be related to a shear zone, show distinct different in waveform compared to other seismometers. The shear zone is about 20m in thickness, which contribute to a Stoneley waves like wavelet in horizontal component. These waves might be related to the fault zone permeability or viscosity. In addition to the observations for regular events, a cross fault experiment of fluid injection test (FIT) was carried out after the drilling and the completed installation of the seismometers to understand the in-situ hydraulic behavior of the fault zone. A high pressure fluid (~4MPa) was injected in hole-B with chemical and gas observations and monitoring in hole-A. We observed distinct feature in the seismograms, which only showing a triangle source type P-wave without S-wave, recorded by TCDP BHS during FIT. Whether this observation was in an association of new open cracks after high pressure Fluid Injection Test (FIT) is carrying on. With the high resolution TCDP 7-level BHS and FIT in an active fault zone, we try to understand dynamics of the fault zone and the estimation on in-situ permeability from FIT related triggered events. The observations would also give hint to fault zone healing process through time.
S31A-0199
Effects of Mining-Induced Stress Perturbations on Pre-Existing Faults Near a Deep South African Gold Mine
For over a century, mining-induced earthquakes have been recorded in the deep underground mining region of Witwatersrand Basin in South Africa. The TauTona gold mine experiences an appreciable number of mining- induced earthquakes and is the site of the Natural Earthquake Laboratory in South African Mines (NELSAM) project. In this work, we constrain the virgin (i.e. unperturbed) state of stress near the TauTona gold mine using an iterative forward modeling approach that combines observations of drilling induced borehole failures in borehole images, boundary element modeling of the mining-induced stress perturbations, and forward modeling of borehole failures based on the results of the boundary element modeling. We find that the state of stress is a normal faulting regime with principal stress orientations that are slightly deviated from vertical and horizontal and therefore denoted with a (*). The maximum principal stress, Sv*, is deviated 0° to 20° from vertical towards the NNW and has a magnitude gradient of 27 ± 0.3 MPa/km. The intermediate principal stress, SHmax*, is inclined 0° to 20° from horizontal with an azimuth of 145° to 168° and has a magnitude gradient of 21.5 to 26 MPa/km. The least principal stress, Shmin*, is inclined 0° to 10° from horizontal with an azimuth of 235° to 258° and has a magnitude gradient of 13 to 15.5 MPa/km. Using this constrained stress state, we investigate the likelihood of faulting to occur both on pre-existing fault planes that are optimally-oriented to the virgin stress state and on faults affected by the mining-perturbed stress field, the latter of which is calculated with boundary element modeling. The far-field stress state indicates that the crust is in a state of frictional faulting equilibrium, such that normal faulting is likely to occur on faults striking SSE and NNW and have a dip angle of approximately 45° to 80°. The mining-induced stress perturbation affects faults relatively closer to the mining excavation. We analyze active faults observed in borehole image log data and mapped in the TauTona access tunnels. In the borehole image logs, we find evidence of recent slip on faults that have become critically stressed due to the mining-induced stress perturbation. In our investigation of the Pretorius Fault Zone (PFZ), an ancient (Pre-Cambrian) fault system, we find that stress perturbations due to recent mining advances may be responsible for reactivating segments of the PFZ that were observed to slip during a M2.2 event recorded in December 2004.
S31A-0200
Seismic velocity structure and hypocentral distribution in the Western Nagano prefecture by using the dense seismic network data
We carried out the seismic tomography to investigate the reason why the earthquake swarm occurs in the hypocentral region of the 1984 Western Nagano prefecture earthquake (Mj6.8). The seismic activity of the earthquake swarm is very high from 1976 to the present. A dense seismic network was installed in this region at 1995. The sampling frequency of this recording system is 10 kHz. The initial number of stations is 6. However, the number of stations gradually increased, the maximum number is 57. The average station spacing is 1 km or 2-4 km near the center or in the periphery of the network, respectively. We used 14,226 events that occurred between October 1995 and February 2005. The number of travel time data of P and S waves is 250,212 and 209,656, respectively. All arrival times are manually picked. The accuracies of P and S arrival times are about 1 ms and 30 ms, respectively. These accuracies are quite high because of the 10 kHz sampling frequency and good signal-to-noise ratio. We conducted an initial hypocenter determination and the 1 dimensional inversion before the 3 dimensional inversion (Seismic tomography), then we obtained more precise hypocentral distribution and velocity structure. The grid intervals in the 3 dimensional inversion are 1km*1km*1km in the vicinity of the area where most of events occur. We used the Pseudo Bending method (Um and Thurber, 1987) and the LSQR method (Paige and Saunders, 1982) for the ray tracing and the matrix calculation, respectively. To estimate the resolutions at each grid point, we carried out the checker board resolution test. We found that most of events are distributed along a lot of planes or lines. Those planes or lines decline almost the same dip angle or perpendicular to the dip angle. The relationship between the hypocentral distribution and the velocity anomalies shows a few distinct features as follows. First, the hypocenters are located around the upper part of the low P wave velocity anomaly. Second, the hypocenters adjoin to the low P and S wave velocities anomaly. Some velocity anomalies seem to reflect the existence of gas or fluid. Thus, the gas or fluid is related to the occurrence of the earthquake swarm in the Western Nagano prefecture.
S31A-0201
New Insights On The Val D'Agri Normal Fault Systems (Southern Italy) By Using A Dense Temporary Seismic Network
Three-dimensional Vp and Vp/Vs images of the upper crust in the Val d"Agri area (Southern Apennines seismic belt), down to 12 km depth, were computed by traveltime inversion of P (9269) and S (8726) waves from about 700 low-magnitude earthquakes. Data were recorded by a dense temporary seismic network composed by 36 digital three-component continuously recording stations, deployed over about 1500 km2 with an average receivers spacing of 5 km. The Val d"Agri basin is one of the most seismically active area in the central Mediterranean region. It was struck by a large destructive earthquake in 1857 (Me=7.0) and ongoing seismic activity testifies the high seismic hazard. Despite of the intense hydrocarbon exploration and field surveys, the geometry of the seismogenic sources is still debated. Two contrasting seismotectonic models are present in the literature, claiming for NE or SW dipping basin bounding normal-fault systems. We recorded, in thirteen months, about 2000 low-magnitude earthquakes (0.5<ML<2.7) testifying a high rate of background seismic release (about 3.3e-03 events/day/km2) almost continuous in time. The seismogenic volume is about 12 km thick. Earthquakes are mostly located beneath the south-western margin of the basin. Seismicity concentrates in the shallow crust (3-7 km) within a high-velocity (up to 6.5 km/s) body, clustering at the boundary between high and low Vp/Vs regions. We interpret the body with both Vp and Vp/Vs positive anomalies as the highly fractured and fluid saturated Mesozoic carbonates of the Apulian Platform, the oil reservoir of the region. Our 3D hypocentral locations and focal mechanism solutions strongly support the presence of an active NE-dipping normal-fault system bounding the basin to SW.
S31A-0202
Heterogeneous structure in and around 2005 West off Fukuoka Earthquake inferred from dense seismic observation
The 2005 West Off Fukuoka Prefecture earthquake (Fukuoka EQ.) occurred on March 20, 2005, with magnitude of 7.0 in JMA scale in the offshore region of Fukuoka Prefecture, Kyushu, Japan. This earthquake damaged many buildings in the surrounding area, especially on Genkai Island as well as in and around Fukuoka City. The earthquake fault was a left-lateral strike-slip having a nearly vertical fault plane and a strike in the WNW-ESE direction. The largest aftershock with a magnitude of 5.8 (JMA) followed 1 month later. In order to obtain both detail seismic activity and inhomogeneous structure, we performed dense seismic observation in and around the aftershock area. The velocity structure was inferred from P and S arrival time data by the Double Difference tomography method, signifying high-velocity regions found on the edge of the aftershock area and on the area with large co-seismic slip. We achieved attenuation structure in the aftershock area from spectral data. Regions with high attenuation were found at the southeastern edge and below the earthquake fault. The scatterer distribution inferred from the seismic array observation also showed existence of strong heterogeneity at the edge. Hypocenter distribution relocated using three dimensional velocity structure implied that the alignment of the aftershock distribution in shallower part is slightly different from the deeper ones. This shows the difference has similar tendency to difference between focal mechanisms by CMT solution and that from polarities of P wave first motions. These results showed the rupture at the beginning of the earthquake had different behavior from the major slip generating large seismic wave. An active fault called Kego fault lies at southeastern of the fault of the Fukuoka EQ. The direction of the Kego fault is not similar to that of Fukuoka EQ in strike direction, which means that a segment boundary in the fault system exists at the part. The rupture of the mainshock stopped at the boundary. In addition, we revealed that strong heterogeneity found at the boundary by velocity, attenuation and scatterer distribution. These results suggested that heterogeneity could control the faulting of Fukuoka EQ., and gave us insights in terms of the stress accumulation process for the forthcoming earthquake at the Kego fault.
S31A-0203
Determining Three Dimension Q Attenuation Structure beneath Kyushu Island, Japan exerting Waveform Spectra of Microearthquakes
Examining the anelastic structure in the uppermost mantle and the crust continues to be a significant problem in seismology. In particular, the observation of attenuation heterogeneities around subduction zones and active fault has implication for the dynamic oceanic lithosphere and asthenosphere. Seismic attenuation can be estimated by extracting the amplitude-frequency information contained in seismic waveforms and it provides an important insight into the nature of heterogeneities structure and composition of the earth's interior. The study area at Kyushu Island, in South Part of Japan is characterized by subduction from Philippine Sea Slab and Eurasian Plate; volcanic front seen in islands arcs runs through the central part of Kyushu Island. Futagawa- Hinagu Fault zone, which is one of the active faults in Japan, lies in the middle of Kyushu, in which the seismic activity of shallow earthquakes is high. The fault is furrow from southwest of Aso volcano until Yatsushiro city, with the type of right-lateral strike slip. For investigating the 3D wave attenuation Q structure of the crust and the upper mantle in Kyushu Island, we examined the seismic wave spectra for micro earthquakes (Magnitude 2.5 4.5) observed at seismic stations by Kyushu Univ., JMA, and Hi-net. In this study selected waveform were used for spectral analysis concerning strong lateral heterogeneities, to estimate Q values by the spectral decay of the spectrum. Used for estimating path attenuation, we obtained the best result with amplitude spectra determined for a 128 - sample window around seismic-wave arrivals and having signal-to-noise ratio greater than 2 over the continuous frequency range. Solving for Q structure, an inversion program were developed with referring Tsumura, 2000.,by given a set of path attenuation, source parameter, site response and velocity model, the Q structure can be distinguished the difference between fore-arc zone and back arc zone around Kyushu Island, especially from the Philippine Sea Slab (subduction). A record of an event distance from the network suggested that high attenuation (Low-Q) was exposed under volcanic areas (in the crust and mantle wedge) along the central of Kyushu Island, and for low attenuation (High- Q) was showing the figure of subducting Philippine Sea Slab. The variations in seismic attenuation can be attributed to a high attenuation (low-Q) zone in the mantle underneath the volcanic area of Kyushu Area; it was formalize by performing the Q structure inversion. Keywords: Attenuation, Q, 3-D structure, Kyushu Island
S31A-0204
Spatial Clustering and Repeating of Seismic Events Observed Along the 1976 Tangshan Fault, North China
The magnitude 7.8 earthquake that struck the city of Tangshan,~160 km east of Beijing, on July 28, 1976 is one of the most destructive events in terms of the loss of life. The main shock occurred along NNE trending strike-slip fault system. We investigated the spatial and temporal distribution of microearthquakes occurring along this fault during 2001-2006 with waveform data recorded by the Beijing metropolitan digital Seismic Network. The relocated seismicity with the double difference method reveals a dextral bend in the middle of the fault that divides the fault into the southern and northern segments. More than 85 percent of the earthquakes were found in two clusters in the northern segment where relatively small coseismic slips were observed during the 1976 earthquake. The b values calculated from the seismicity occurring in the northern and southern segment are 1.03±0.02 and 0.85±0.03, respectively. The distinct seismicity and b values are probably the collective effect of the fault geometry and the regional stress field that has a NNE-SSW oriented compression. The southern segment thus experiences a lower shear stress but a higher normal stress compared to the northern segment, resulting a smaller number of earthquakes with relatively larger magnitude that explains the observed lower seismicity and b value. Using cross-correlation and fine relocation analyses, we also identified a total of 21 doublets and 25 multiplets that make up >50% of the total seismicity. Most of the sequences are aperiodic with recurrence intervals varying from a few minutes to hundreds of days. Based on a quasi-periodic sequence, we obtained a fault slip rate of ~2 mm/yr at ~15 km, which is consistent with surface GPS measurements. The high percentage of similar and repeating events suggests that the intraplate Tangshan fault is likely to have a well developed weak zone which is a characteristic of many mature interplate faults.
S31A-0205
Complex fault system of recent earthquakes in Central Japan by dense seismic observations
Some earthquakes of the transverse fault occurred in Central Japan, recently. A temporal dense seismic array was deployed the day after the main shocks to determine the major source faults responsible for the main shock and large aftershocks. Using the high-resolution seismic data, it was clarified that they were not one fault plane but two or more fault planes. For example, in a case of the 2004 Niigata prefecture Chuetsu Earthquake two parallel faults dipping steeply to the west located 5 km apart, and another dipping eastward and oriented perpendicular to the west-dipping faults. The strong heterogeneity of the crust is related to the complex geological and tectonic evolution of the area. The complexity of the fault system was clarified by a dense seismic observation. We will show the following cases: the 2004 Niigata Prefecture Chuetsu Earthquake, the 2004 off Kii Peninsula Earthquake, the 2007 Noto Hanto Earthquake and the 2007 Niigata Prefecture Chuetsu Offshore Earthquake. And we have started a new project, the Special Project for Earthquake Disaster Mitigation in the Metropolitan Tokyo area. The core item of this project is a dense seismic array observation in metropolitan area, which is called the MeSO-net (Metropolitan Seismic Observation network). It is a purpose of this observation to clarify the more detailed geometry and physical properties of the Philippine Sea Plate beneath the Tokyo area.
S31A-0206
The 2006 Acadia National Park, Maine Earthquake Sequence
A sequence of small to moderate sized earthquakes shook the Bar Harbor, Maine region in Acadia National Park starting on September 2006. The earthquake sequence started with a magnitude 3.4 event on September 22, 2006 which was preceded by four earthquakes with magnitude ranges of 1.8 to 2.4. Many aftershocks of magnitude about 2 followed the September 22 shock, then a magnitude Mw 4.0 earthquake occurred in the same epicentral area on October 3, 2006. A four-station local seismographic network was deployed on October 5, 2006, by a field crew from Lamont-Doherty Earth Observatory (LDEO) in collaboration with staff at Acadia National Park around Mount Desert Island for aftershock monitoring. The focal mechanism of the mainshock on October 3, 2006 shows the P-axis with trend=86 degree and plunge=10 degree and the T-axis with trend=289 degree and plunge=79 degree. Hence, the P-axis orientation is consistent with the stress axis in the northeastern United States - E-W to ENE-WSW trending horizontal compression. During October 5 through November 30, 2006 about 45 aftershocks with magnitude ranging from Mw 0.4 to 2.4 were located by the local network. Preliminary locations of the aftershocks indicate that these shocks occurred at three clusters around Champlain Mountain close to the east coast of Mount Desert Island. The focal depths of the aftershocks range from 0.5 to 2.5 km. The hypocenters show a complex distribution and do not delineate a simple fault plane. The static stress-drops of 0.1-10 MPa are obtained for these aftershocks, whereas the source radii estimates range 0.05-0.2 km for aftershocks with seismic moment 1.0e+10 Nm to 1.0e+13 Nm (Mw 0.4-2.4).
S31A-0207
Relations between rupture velocity changes, seismic radiation, and fault geometry during the Kokoxili earthquake (Tibet, 2001/11/14)
Determining how rupture velocity varies during an earthquake provides decisive information on the nature of the rupture process and the associated seismic radiation. We present here an innovative analysis of the 400-km- long Kokoxili earthquake (Tibet, 14 November 2001) using the HIMNT broadband stations located in Nepal as a dense seismic array. This approach reveals that the speed of the propagating rupture tip reached extremely high values, very close to the velocity of seismic compressional waves, over a long fault segment. Although theoretically known since the 1970"s, direct evidences for the existence of this regime had not been provided up to now. We show that the acceleration or deceleration locations related to the transition between low and high velocity regimes are not randomly located but correlate very well with fault geometrical complexities. These transitions are shown to generate strong high frequency radiations and are therefore a very serious candidate to explain the origin of the damaging waves induced by earthquakes.
S31A-0208
High-Resolution Source Imaging of the 2007 Niigataken Chuetsu-oki Earthquake From Dense Strong-Motion Networks: Focus on the First 5 Seconds of the Rupture Process
The 2007 Niigataken Chuetsu-oki earthquake (Mw6.6) occurred on 16 July 2007 (JST) in the western offshore of Niigata Prefecture, Japan, which strongly shook the near-source area and was recorded with dense strong- motion observation networks. Although no station was in the source region because this earthquake occurred in offshore, many strong-motion stations recorded the seismic motion near the source region. The local strong- motion records of this earthquake have a remarkable feature: the records show several seconds of small but increasing amplitude arrival ("initial rupture phase") followed by the onset of the main energy release ("main rupture phase"). In this study we get from these records the high resolution image of the source process during the five seconds after the rupture initiation of this earthquake. To investigate source rupture process, waveform inversions are now popular, but these source inversion methods do not have so high resolution. Recently another approach with high resolution, the back-projection method (e.g., Ishii et al., 2005; Yamamoto and Takenaka, 2006) was exploited, which images earthquake rupture directly from dense network data. The back- projection method is useful to teleseismic array data of an earthquake with very large dimension such as the 2004 Sumatra-Andaman earthquake (Mw9.3), as mentioned by Ishii et al. (2005, Nature). The back-projection method was also adapted for analyzing local strong-motion records of a middle earthquake, the 2005 Fukuoka earthquake (Mw 6.6), by Yamamoto and Takenaka (2006, AGU Fall Meeting). Their imaging technique is useful to middle to large earthquakes. Here we applied their technique to near-source strong-motion records of the 2007 Chuetsu-oki earthquake. We used P-wave portion of UD motion records and derived the following detailed process during the five seconds after the rupture initiation: The rupture plane is a NW-dipping one which is one of the P-nodal planes in the focal mechanism solution. The rupture mainly propagated toward the direction of Kashiwazaki City (around N200E). The main rupture (breaking of the asperity) began at a position of 4 km apart and southwest above the original hypocenter at 2.1 seconds after the origin time. The asperity distributed toward the direction of Kashiwazaki City on the fault plane. This spatial configuration suggests that the rupture front then swept the asperity upward and to the southwest, so that Kashiwazaki City directly suffered the strong effects of the forward rupture directivity ("killer pulse"). (Acknowledgements) We used the strong-motion records supplied by the National Institute for Earth Science and Disaster Prevention (NIED; K-NET, KiK-net, F-net), and the Japan Meteorological Agency (JMA). This study are patially supported by "Special Project for Earthquake Disaster Mitigation in Metropolitan Tokyo Area".
S31A-0209
Seismic source characterization by ionospheric sounding from Gound Positioning System data
Imaging the terrestrial ionosphere is becoming possible since the installation of dense GPS networks, with a temporal and spatial resolution allowing the detection of ionospheric seismic waves. Since the 1960s, ionospheric seismic waves are detectable almost punctually after large shallow earthquakes, with current minimum magnitude of 6.5. Most recently, the use of dense networks gave the way to a global visualization of the horizontal propagation of co-seismic ionospheric disturbances. Such a use of a Global Positioning System array, and the sounding capability of the method above the ocean, prove the potential of this method as a complement to more traditional techniques used in seismology. From now on, after imaging seismic waves in the ionosphere, the challenge is the characterization of the seismic source, whose rupture involves coupling mechanisms between the moving solid earth and its surrounding atmosphere. The study presented here is based on the Total Electronic Content variations mapped close to the source and shortly after the Tokachi-Oki earthquake (M=8.3) that occurred on September, 25, 2003, in Japan. The first fundamental source parameters derived from 1 Hz sampled data will be reminded here. The rupture process is then pre-modelled in reference to the co-seismic displacements estimated by other techniques. Therefore, a modelling of the horizontal propagation of acoustic waves generated by three aligned separated sources is developed. The preliminary results of the subsequent GPS data inversion tests will be presented. Finally, for physical modelling of the vertical propagation, we used ray tracing in the atmosphere, in order to study the effects of the near-field pulse spreading in acoustic domain as well as the redistribution of the charged particles under geomagnetic dependency.
S31A-0210
Heterogeneity in the Crust and Uppermost Mantle beneath southwestern Japan Imaged by Coda Envelope Inversion
We have investigated short-wavelength heterogeneous structures in the crust and uppermost mantle beneath the Chugoku and Shikoku regions, southwestern Japan, in which the Philippine Sea plate subducts along the Nankai trough beneath the overriding plate. A large number of observed envelopes of S coda were analyzed by an envelope inversion method [Asano and Hasegawa, 2004] in order to estimate 3-D distribution of scattering coefficients. It is known that the seismic coda is a superposition of seismic waves scattered by short-wavelength heterogeneous structures. Estimated distribution of scattering coefficients provides us information about the short-wavelength heterogeneous structures, which is hardly imaged by travel-time tomography. We analyzed 3553 three-component seismograms recorded at 143 Hi-net stations installed with about 20- to 30-km separation. The target area is covered with 18 areas for inversion grid having 4139 node points with horizontal and vertical separations of 8 and 10 km in the central part, and with horizontal and vertical separations of 16 and 20 km in the surroundings. Scattering coefficient values in each node point are unknown parameters to be estimated. Obtained results show that there exist several zones with large scattering coefficients (LSZs). In the upper crust, LSZs are distributed in northen part of the Chugoku region and eastern part of Shikoku region, which is approximately corresponding to seismically active regions. In the lower potion, most predominant LSZs are widely distributed at depths of about 30 km and deeper beneath the central part of the Chugoku region. These LSZs are thick and extend to other LSZs located near the top of the subducting Philippine Sea slab. Shiomi et al. [2006] found that the depth of continental Moho is relatively deeper in the central part of the Chugoku region than in the surroundings by receiver function analysis; Matsumoto [2006] showed that bottom depth of seismogenic layer is deep and heat-flow value is low in this region. Therefore it seems that the LSZs in this region show distribution of reflective and/or heterogeneous structures in lower part of such cold and thick crust suggested by these previous studies.
S31A-0211
Interferometric seismic imaging of sedimentary basins using local-earthquake coda
In recent years, the correlation-type reciprocity theorem for one-way wave fields has been used to derive relations between the reflection and transmission responses of an arbitrary 3-D inhomogeneous medium with distributed sources below irregular layers. This concept is equivalent to the extraction of pseudo-shot records related to surface-related multiples at all receiver stations. Based on this generalized relation, the interferometric seismic imaging(ISI) in the presence of passive seismic sources can be simulated by cross-correlating the transmission responses recorded at dense receiver array. The ISI inherently realizes symmetric pseudo-shot-receiver sampling, and prevents irregularities of offset distribution in CMP ensembles. We have investigated the possible application of ISI to local earthquake data, which ensures the basic assumption of the correlation-type reciprocity theorem that the seismic source is located beneath the all reflectors to be reconstructed. Synthetic seismograms simulated by the elastic pseudospectral method for a simple 2-D model of sedimentary basin are given to investigate the application of ISI approach to first-order free-surface multiples in local- earthquake coda. The numerical modeling results demonstrate the potential imaging capabilities of ISI for the sedimentary basin structure with a high spatial resolution. ISI has also been applied to local earthquake data accidentally recorded during a seismic reflection survey across the Kanto basin, central Japan. We compare the basement structure of Kanto basin estimated from local-earthquake coda with the prestack migrated profile of active seismic experiment.
S31A-0212
Temporal variations of crustal structure in the source region of the 2007 Noto Hanto and Niigata-ken Chuetsu-oki Earthquakes, central Japan, with passive image interferometry
The passive image interferometry technique (Sens-Sch{\" o}nfelder and Wegler, 2006) is preliminarily applied to the continuous seismic waveform data obtained around the source region of the 2007 Noto Peninsula Earthquake (M6.9, Noto EQ) and 2007 Niigata-ken Chuetsu-oki Earthquake (M6.8, Chuetsu-oki EQ), central Japan, to detect the temporal variation of the subsurface structure around the source region. We computed the autocorrelation function (ACF) of band-pass (1.5 or 2.0 Hz ~ 10.0 Hz) filtered seismic noise portion recorded with each short-period seismometer at several seismic stations for each one day. There are two remarkable features in the observed ACFs. One is that phases with a respective predominant frequency (around 2.5 Hz ~ 3.5 Hz) in ACF at one station are coherent over different days, though the shape of ACF at each station is different to each other. The other is that the temporal evolutions of particular phases in ACF are observed around the occurrence of the earthquake in some cases. In general, the increase lag time of phases in ACF is caused by decrease of seismic wave velocity, and vice versa. Change of wave velocity will be caused by several reasons. Some possible candidates proposed in the previous studies are as follows; (1) change in stress in the Earth's interior caused by the earthquake, (2) change in the near-surface material properties such as generation of damage or cracks caused by the strong shaking of the earthquake, and (3) change of degree of water saturation in the near-surface. Sudden changes of ACF are detected associated with the occurrence of the main shocks. Followings are some results together with epicentral distances (EPD) and changes of volumetric strain caused by mainshocks. In the source region of the Noto EQ, station N.TGIH (EPD = 4km, Dilatation) exhibits the increase of lag time of the phases. ACFfs of the two stations around the source region of the Chuetsu-oki EQ also show the change of lag time. Station IZUMOZ (EPD = 9 km, Contraction) exhibits increase of lag time while station N.KZKF (EPD = 30 km, Dilatation) shows decrease of the lag time of the ACFs. Additionally, in previous study, Wegler and Sens-Sh{\" o}nfelder (2007) detected the increase of lag time at the station N.KZKF associated with the occurrence of 2004 Niigata-ken Chuetsu earthquake (M6.8, Chuetsu EQ), whose EPD is 31 km and located in the area of dilatation. If we assume the change of lag time of the ACFs are caused by change of stress, the tendencies observed in the source region of Noto EQ and Chuetsu EQ are consistent, while that of Chuetsu-oki EQ is not. Temporal variations of ACFs associated with these three M6 class inland earthquakes indicate that change of seismic velocity structure associated with mainshock can be detected with this method. At present, however, it is hard to give concrete interpretations to the observed features. It is essential to make quantitative analyses to identify what the phases in ACF observed at each stations mean, as well as identifying the cause of temporal evolution of ACFs. In some stations, temporal evolution of ACFs preceding the mainshock is also detected, which would be of great importance for understanding the stress state before occurrence of earthquakes.
S31A-0213
Crustal heterogeneity and its relation to seismic activity in the Kinki district, southwest Japan
It is essential to estimate detailed heterogeneous structures in and around the earthquake source region, in order to understand the generating process of earthquakes. Microseismicity is very high in the central to northern part of the Kinki district, southwest Japan. We will discuss the crustal heterogeneity in this region from a viewpoint of delineating its correlation with the earthquake occurrence in space and time. We estimated a 3-D distribution of relative scattering coefficients in the Kinki district by inversion of coda envelopes from local earthquakes. We analyzed 828 seismograms from 63 events which occurred in 2003, recorded at 50 stations of Kyoto Univ., Hi-net, and JMA. The result shows the existence of a remarkable scattering zone at a depth from 20 to 30 km, just below the high microseismicity area. The strong scattering zone is well correlated with the location of S-wave reflectors estimated by previous studies (Katao, 1993), and has revealed the heterogeneous structures in more detail. Deep low-frequency earthquakes occur at a depth of 30-40 km just below this strong scattering zone (Ohmi, 2002). From these, we consider that this scattering zone should have a strong influence on the active microearthquakes just above it, for example as supplying fluids to the seismogenic zone in the upper crust. We have a data base of earthquakes in recent 30 years in this region. We will also discuss the correlations between the temporal variation in microseismicity and that in scattering images, aiming at detecting some evidence of crustal fluids affecting the earthquake occurrence.
S31A-0214
3-D Seismogenic Stress Fields in and around Japan, Inferred from the CMT Data Inversion
We developed a robust inversion method to estimate the pattern of the stress fields related to earthquake generation (seismogenic stress fields) from the centroid moment tensors (CMT) of seismic events by using Akaikefs Bayesian information criterion (ABIC). The CMT solution of a seismic event is conventionally defined by the surface integral of 2-D moment tensor density over a rupture area. Applying Gauss' divergence theorem, we can transform the surface integral of 2-D moment tensor density into the volume integral of stress release over a finite elastic region surrounding the dynamic rupture area. The volume integral representation of CMT is more essential than the surface integral representation, because dynamic rupture growth is controlled by energy flow into the rupture zone from the surrounding region storing elastic strain energy. Since the occurrence of an earthquake releases some part of the seismogenic stress field around its hypocenter, we can relate CMT data with the seismogenic stress field with the volume integral representation. We represent the CMT of a seismic event by a weighted volume integral of the true but unknown seismogenic stress field. The weighting function is taken to be a 3-D Gaussian-type distribution with its peak at the hypocenter and variance proportional to the two- thirds power of the seismic moment. Representing each component of the seismogenic stress field by the superposition of a finite number of 3-D basis functions, we obtain a set of linear observation equations to be solved for the expansion coefficients (model parameters). We introduce prior constraint on the roughness of the seismogenetic stress field and combine it with observed data to construct a Bayesian model with a hierarchic flexible structure controlled by hyper-parameters. The optimum values of the hyper-parameters are objectively determined from observed data by minimizing ABIC. Given the optimum values of the hyper-parameters, we can obtain the best estimates of model parameters by using a maximum likelihood algorithm. We applied the inversion method to observed CMT data in and around Japan (the NIED Moment Tensor Catalogue, 1997.1.31-2007.1.31) to reveal the 3-D patterns of seismogenic stress fields associated with plate subduction. We compared the stress patterns in the northeast Japan arc and the Ryukyu arc. In both regions the stress patterns in the shallow part of the oceanic plates and the descending slabs are characterized by normal and reverse faulting with strikes parallel to the trench axes, respectively. However, the stress patterns in the back- arc regions are quite different from each other: reverse faulting in the northeast Japan arc and normal faulting in the Ryukyu arc. The difference in the stress pattern results from the difference in the tectonic loading mechanism.
S31A-0215
Strain Concentration Zones, Inland Earthquakes and Seismogenic Layers in the Crust of Central Honshu, Japan
Strain concentration zones have been found from surveys by a nation-wide dense GPS network in recent years in Honshu, Japan. The zones seem to have a large potential to cause disastrous earthquakes in the future, since large historical earthquakes occurred in and near the zones. Besides, Some moderate sized earthquakes took place for the last three years in one of the zones called the Niigata-Kobe Tectonic Line (NKTZ). However, no clear indication relating to the zones except for the abrupt variations in GPS displacement vectors along the zones. There exists an active fault system, called the Atotsugawa Fault System, in a part of NTTZ, in northwest of Chubu Honshu. In and around the fault zone, seismic observations have been continued for more than 30 years by the Kamitakara Observatory, Kyoto University. As a result, detailed seismicity has been revealed in the fault zone, showing a clear change in depth distribution across the fault zone; depths are shallow along the fault and become deep towards both sides of the fault, especially deeper in the NW direction. Furthermore, seismic velocity structure was also obtained from seismic refraction and wide-angle reflection studies. The depth of the upper and middle crust becomes deep towards NW from the fault zone. The variations in the cutoff depths of seismicity coincident well with the fault zone and eventually the stress concentration zone along the Atotsugawa fault system. The tendency is the same as in the area of the 2004 Chuetsu earthquake, occurred in NTTZ. In general, the stress concentration zones seem to locate in the areas along the change in depths of the seismogenic layer, although the reason of the strain concentration is not yet clear. The seismogenic layer is well correlated with thermal structure, as shallower the cutoff depth, the hotter the area. Furthermore, near both sides along the strike of a large inland event, cutoff depths of aftershocks become shallower. Thus, variations in depths of the seismogenic layer seem to affect the strain concentration in the upper crust. Dense network of earthquake observation made us enable to get high accuracy hypocenters on shore over Japanese Islands. Therefore, comparing the cutoff depths of seismicity with strain concentration areas from GPS over central Honshu, the latter seems to locate along the steeply changing areas of the cutoff depth of seismicity. The main strain concentration zones are NKTZ, the area along the Median Tectonic Line, and on the coast from Kanto to Tokai districts. However, since all steep changing areas of cutoff depths do not necessarily correspond to the stress concentration zone, the deeper crust, in which many dipping reflectors lie, also seems to play an important role to the formation of the strain concentration.
S31A-0216
Co-seismic strain changes inferred from the groundwater level in Chinese mainland induced by the 2004 Sumatra Earthquake
The 2004 great Sumatra earthquake induced prominent water level steps in at least 50 wells of the Earthquake Monitoring Network in Chinese Mainland. We compiled the water level data and analyzed the amplitude of the water level. We found that the amplitude of well water level changes was not consistent with the epicentral distance but the wells contributed along the tectonics. We use earth tides information recorded by water level to standardize the water level steps and name the standardized water level changes as equivalent strain. We found that the amplitudes of those equivalent strain changes showed prominent difference between the west and the east. The amplitudes in the west of China are larger than those in the east of China. We compared the equivalent strain of the well water level with the co-seismic surface displace measured by GPS, the result showed that the trends of them are consistent with each other. We compared the equivalent strain with the distribution of apparent stress in China which was inferred from the NEIC broadband radiated energy catalogue and the Harvard CMT catalogue from January 1987 to December 1998 finished by Wu Zhongliang et al. in 2002. The amplitudes and their distribution are well consistent with each other. That means that the co-seismic strain from ground water was controlled by the tectonic stress state that was expressed by apparent stress, which can be as one of the mechanism of induced groundwater changes by remote earthquakes. Our conclusion is that the seismic induced groundwater changes can be as one of the approaches for active tectonic investigation. This research is supported by National Natural Science Foundation of China with Contracts No.40674024 and 40374019.
S31A-0217
The Metropolitan Seisimic network (MeSO-net) for Detection of Mega-thrust and Intra-slab Earthquakes beneath Tokyo Metropolitan Area, JAPAN
In central Japan the Philippine Sea plate (PSP) subducts beneath the Tokyo Metropolitan area, the Kanto region, where it causes mega-thrust earthquakes, such as the 1703 Genroku earthquake (M8.0) and the 1923 Kanto earthquake (M7.9). Assessment in Kanto of the seismic hazard produced by the Philippine Sea Plate (PSP) mega-thrust earthquakes requires identification of all significant faults and possible earthquake scenarios and rupture behavior, regional characterizations of PSP geometry and the overlying Honshu arc physical properties (e.g., seismic wave velocities, densities, attenuation), and local near-surface seismic site effects. The Kanto Basin is covered with thick Neogene to Quaternary sediments interfering precise observations. Furthermore culture noises in the Kanto region are quite high because of the highly concentrated population. Those environments prevent us to make a clear image of the PSP plate until now. To achieve stable high sensitivity seismic observation(Hi-net) avoiding surface ground noise, seismographs are installed at the bottom of the borehole of more than 100m depth(maximum 3,500m). The average interval of Hi-net over the Tokyo metropolitan area is, however, about 25km, which is not sufficient to image the internal structure of depicts of the PSP. The velocity structure in the Kanto region is elucidated by a temporary highly dense seismic array observation in the Boso peninsula using double-difference tomography. (Hagiwara et al., 2006) The results of this array revealed detailes layered structure: there are an oceanic crust and the slab of the Philippine Sea plate. The configuration of the top and bottom boundaries of the Philippine Sea Plate colliding with the Pacific plate was elucidated. It was shown that, although containing large noises, the highly dense array data enables us to depict a detailed underground structure. Thus we proposed highly dense seismic array observation in the Tokyo metropolitan area. This array is named the MeSO-net (Metropolitan Seismic Observation network ) with 400 stations and 2-5km station interval under the Special Project for Earthquake Disaster Mitigation in the Metropolitan Tokyo area (2007 - 2011), which we have just started. Deployment of the MeSO-net will be completed in four years (2007-2010). To achieve stable seismic observation avoiding surface ground noise by an easy method, seismographs are installed at the bottom of a borehole of 20m depth. The MeSO-net will give an accurate estimation of the plate boundaries of the PSP and the Pacific plates under the metropolitan area, resulting in clear understanding of the relation between a deformation of PSP and intra-slab M7+ earthquakes. The MeSo-net will also present a high resolution tomographic image to show a low velocity zone which may a possible internal failure of the slab, i.e., a possible source region of the M7+ intra-slab earthquake. Our Special Project will contribute the next assessment of the seismic hazard in the Tokyo metropolitan area.
S31A-0218
Metropolitan Seismic Observation Network (MeSO-net) in Japan
Beneath the metropolitan Tokyo area, the Philippine Sea Plate (PSP) subducts and causes damaged mega- thrust earthquakes. The Dai-Dai-Toku Project revealed the geometry of the upper surface of PSP, and estimated a rupture process and a ground motion of the 1923 Kanto earthquake. However, these results are not sufficient for the assessment of the entire picture of the seismic hazards beneath the metropolitan Tokyo area including those due to an intra-slab M7+ earthquake. So, we have started a new project, the Special Project for Earthquake Disaster Mitigation in the Metropolitan Tokyo area. Proving the more detailed geometry and physical properties (e.g. velocities, densities, attenuation) of PSP is very important to attain this issue. The core item of this project is the dense seismic array observation in metropolitan area, which is called the MeSO-net (Metropolitan Seismic Observation network). The MeSO-net consists of 400 stations including those in Hakone, a Data Center at ERI (Earthquake Research Institute, the University of Tokyo), and a Sub-Center at NIED (National Research Institute for Earth Science and Disaster Prevention), at which the data are backuped and integrated with the Hi-net data. In order to obtain the high resolution images of a velocity and Q structure, it is requested to construct a seismic network with a spacing of 2-5 km. The total number of seismic stations of the MeSO-net will be about 400. The MeSO-net is to be deployed in 4 years and will provide useful datasets. Since the metropolitan area is surrounded with cultural noises, seismographs are installed at the bottom of the 20-m-deep borehole. Seismographs, data of which are digitized at the bottom of observation well, are three- component acceleration type with the maximum measurable acceleration of ±2G (dynamic range of 135 dB or more). The seismic data with a sampling rate of 200 Hz are telemetered to the Data Center using UDP/IP protocol. We adopt an autonomous cooperative seismic telemetry system, which is a newly developed for the reliable data transmission in a large seismic observation network, such as the MeSO-net. The new system enables us to easily maintain and quickly recover from troubles.
S31A-0219
Quiescence of repeating earthquakes preceding M-6 class interplate earthquakes
Understanding of processes preceding large earthquake generation is important. Interplate repeating earthquakes are considered as repetition of ruptures on the small patches distributed on the plate interface and can be used to monitor relative slips on the plate boundary and determine configuration of plate interfaces. At the Kanto district, M-6 class interplate earthquakes often occurred on the subducting Pacific plate (PAC). We studied temporal evolution of plate slips based on small repeating earthquakes. We conducted tentative repeating earthquake analysis for events from 2003 to 2006 in addition to the previous result of Kimura et al. (2003). Adding new results, we calculated cumulative slip history for the past 27 years from 1979 to 2006. At the eastern Kanto district, five M-6 class interplate earthquakes occurred on the PAC for the past 27 years (1989, 1990, 2000, 2005a, and 2005b event). These events and their aftershocks are distributed in a seismic cluster with size of approximately 17 x 17 km. In this cluster, numerous repeating earthquakes have been found and we calculated cumulative slip history by using them. In the slip history, deceleration of plate slip with time period of 2 to 4 years preceding M-6 class events are clearly observed for 3 of 5 events (1989, 2000, 2005a events). Deceleration of plate slip corresponds to quiescence of repeating earthquakes. Our results indicate that repeating earthquakes within 17 x 17 km region show quiescence preceding M-6 class earthquake. For the other 2 events, time interval from the previous event is not long enough, and it is difficult to judge whether slip rate decelerates or not. We interpret this phenomenon of quiescence of repeating earthquakes as that degree of locking on the plate interface increases with time before occurrence of the M-6 class earthquake. As the degree of plate locking increases, plate boundary slip is likely to decelerate near the source region.