Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Post-glacial rebound wikipedia , lookup
Shear wave splitting wikipedia , lookup
Earthquake engineering wikipedia , lookup
Oceanic trench wikipedia , lookup
Magnetotellurics wikipedia , lookup
Mantle plume wikipedia , lookup
Seismometer wikipedia , lookup
Reflection seismology wikipedia , lookup
Plate tectonics wikipedia , lookup
Seismic inversion wikipedia , lookup
Tectonophysics 388 (2004) 7 – 20 www.elsevier.com/locate/tecto Configuration of subducting Philippine Sea plate and crustal structure in the central Japan region Takashi Iidakaa,*, Tetsuya Takedaa, Eiji Kurashimoa, Tomonori Kawamuraa, Yoshiyuki Kanedab, Takaya Iwasakia a Earthquake Research Institute, University of Tokyo, Yayoi 1-1-1-, Bunkyo, Tokyo 113-0032, Japan b Japan Marine Science and Technology Center, Natsushima 2-15, Yokosuka, Japan Received 30 September 2003; received in revised form 3 February 2004; accepted 13 June 2004 Available online 2 September 2004 Abstract A seismic experiment with six explosive sources and 391 seismic stations was conducted in August 2001 in the central Japan region. The crustal velocity structure for the central part of Japan and configuration of the subducting Philippine Sea plate were revealed. A large lateral variation of the thickness of the sedimentary layer was observed, and the P-wave velocity values below the sedimentary layer obtained were 5.3–5.8 km/s. P-wave velocity values for the lower part of upper crust and lower crust were estimated to be 6.0–6.4 and 6.6–6.8 km/s, respectively. The reflected wave from the upper boundary of the subducting Philippine Sea plate was observed on the record sections of several shots. The configuration of the subducting Philippine Sea slab was revealed for depths of 20–35 km. The dip angle of the Philippine Sea plate was estimated to be 268 for a depth range of about 20–26 km. Below this depth, the upper boundary of the subducting Philippine Sea plate is distorted over a depth range of 26–33 km. A large variation of the reflected-wave amplitude with depth along the subducting plate was observed. At a depth of about 20–26 km, the amplitude of the reflected wave is not large, and is explained by the reflected wave at the upper boundary of the subducting oceanic crust. However, the reflected wave from reflection points deeper than 26 km showed a large amplitude that cannot be explained by several reliable velocity models. Some unique seismic structures have to be considered to explain the observed data. Such unique structures will provide important information to know the mechanism of inter-plate earthquakes. D 2004 Elsevier B.V. All rights reserved. Keywords: Philippine Sea plate; Reflected waves; Tokai region; Plate boundary; Crustal structure 1. Introduction * Corresponding author. Tel.: +81 358415804; fax: +81 356897234. E-mail address: [email protected] (T. Iidaka). 0040-1951/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2004.07.002 In the Tokai region, central Japan, the Philippine Sea plate is descending beneath a continental plate at a velocity of several cm/year. In this area, large inter-plate earthquakes have occurred repeatedly. A knowledge of the physical properties at the plate boundaries will help 8 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 to understand the mechanism of large inter-plate earthquakes. The Tokai region is very important for understanding the mechanism of large inter-plate earthquakes. Many numerical simulations of large earthquakes have focused on this area (e.g., Kato and Hirasawa, 1999). In numerical simulation studies the configuration of the upper boundary of the subducting plate and the physical property at the plate boundary are important. The geometry of the subducting plate is one of the very important parameters in the numerical simulation, but it has not yet been well determined. Little is known of physical properties at the plate boundary. The detailed configuration of the subducting Philippine Sea plate was not clear because it had been Fig. 1. Location map of the profile line in 2001 experiment (A–AV), which crosses central Japan from north to south. The profile line traverses the island arc from the Philippine Sea to the Japan Sea. 391 seismic stations are located on the line. The profile line of the experiment in 1985 (B–BV) with an east–west direction is also shown by small symbols (Matsu’ura et al., 1991). Shot points are shown by stars. The research area is shown in the inset. The thick purple line in the Philippine Sea is the profile line of the experiment of JAMSTEC. The seismic stations of the experiment of Aoki et al. (1972) are shown by triangles. The orange lines crossing central Japan are profile lines analyzed by Takeda (1997). T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 estimated from the distribution of micro-earthquakes (e.g., Ishida, 1992; Yamazaki and Ooida, 1985; Harada et al., 1998). Seismic reflection and refraction studies are required to know the detailed configuration of the subducting Philippine Sea plate. Several seismic experiments with explosive sources have been done in this area (Aoki et al., 1972; Matsu’ura et al., 1991; Iidaka et al., 2003) (Fig. 1). In the study of Matsu’ura et al. (1991), the seismic survey line was located parallel to the Nankai trough with a length of 53 km (B–BV in Fig. 1). Two later arrivals were observed and interpreted as reflected waves at the boundaries with depths of about 12–16 and 23–28 km. The boundary with a depth of 23–28 km was considered to be the upper boundary of the subducting Philippine Sea plate. However, it is difficult to know the configuration of the subducting plate because the survey line was located parallel to the Nankai trough. A seismic experiment with the profile line along the dip direction of the descending plate is required to reveal the configuration of the subducting plate. Iidaka et al. (2003) researched the crustal structure and the configuration of the subducting Philippine Sea plate by forward-modeling using data from an active source seismic experiment. Clear reflected waves from the upper boundary of the subducting Philippine Sea plate were observed. However, the configuration of the subducting plate was not estimated well because of the trade-off between the dip-angle of the boundary and velocity structure. In the Nankai region, which is located west of the Tokai region, huge earthquakes have occurred cyclically. A low-velocity layer at the top of the subducting Philippine Sea plate had been reported from active source experiments (Kodaira et al., 2000, 2002; Kurashimo et al., 2003) and ScSp phase studies (e.g., Nakanishi, 1980; Nakanishi et al., 1981) in the Nankai region. A detailed analysis of the reflected waves is required to reveal the physical properties at the upper boundary of the subducting plate. A joint seismic experiment was conducted in the Tokai and Chubu areas, central Japan, in August, 2001 (Fig. 1). The seismic experiment consisted of three parts: (1) refraction study with explosive sources by the Research Group for Seismic Expedition in Central Japan, which is organized by universities, JAMSTEC (Japan Marine Science and Technology Center), and 9 other government organizations (Iidaka et al., 2003), (2) refraction and reflection surveys in the oceanic area by JAMSTEC (Kodaira et al., 2003), (3) reflection study with CDP cable by several universities (Sato et al., 2001). Here, we show the results of the refraction study in central Japan. The objectives of the experiment are to know the large-scale structural variation of the island-arc crust across central Japan, and to know the configuration of the subducting Philippine Sea plate. In the study, we research the lateral heterogeneity of the subducting Philippine Sea plate and crustal structure using data from seismic experiments. 2. Data The seismic experiment was conducted on 25 and 26 August, 2001. A 261.6-km-long profile extended in the N–S direction to traverse the island-arc of Japan from the south coast to the north coast (Fig. 1). We put 391 seismic stations along the survey line. Six explosive sources were shot on the seismic survey line. The charge sizes of the shots were 500 kg for J1, J2, J3, J4, and J5, and 100 kg for T6. Verticalcomponent sensors with a natural period of 2.0 Hz were used at 328 seismic stations. Three-component geophones (4.5 Hz) were used at 63 seismic stations. The average spacing of the seismic stations was 669 m. Digital recorders were used at the seismic stations with a sampling frequency of 100 Hz. 3. Analysis 3.1. Crustal structure in the Tokai–Chubu region A velocity model was constructed from observed data by the following process. (1) First arrivals and later phases at each seismic record were picked. (2) The observed arrival time data located very close to the shot points were used to estimate the velocity structure of the shallowest layer. (3) The seismic structure was estimated by forward modeling using a ray-tracing method (Zelt and Smith, 1992). The record sections are shown with a reduction velocity of 6 km/s (Fig. 2). The amplitude of the waveforms is normalized by the maximum amplitude value of each trace. In the record section of shot J1 (Fig. 10 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 Fig. 2. Record sections of shots J1 (a), J3 (b) and J5 (c). In the record section of J1, delayed arrivals appear at the Tonami basin (20–40 km in distance). The remarkable two later-phases (L1, L2) are shown by arrows in the record section of J5 (200 to 20 and 100–0 km). T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 11 Fig. 3. The reflected waves at the upper boundary of the subducting Philippine Sea plate (arrows) for shots of J4 (a) and T6 (b). 2a), the arrival times are delayed at ranges of 20–40 km. Pn arrivals are first arrivals at ranges greater than 130 km. For shot J3 in the centre of the profile, the apparent velocity is almost 6 km/s with some perturbation at ranges greater than 20 km (Fig. 2b). In the record section of J5 (Fig. 2c), two clear later arrivals are shown at ranges of 200 to 20 km and 100 to 0 km. The amplitude of the later arrivals, which are observed at ranges of 200 to 20 km from shot J5, is very large compared to the first arrivals (Figs. 2c and 3). The obtained P-wave seismic structure is shown in Fig. 4. The observed and modeled travel times are shown in Fig. 5. The shallower part of the crustal structure is well constrained by the travel times of the Fig. 4. P-wave velocity structure model at central Japan. Several reflectors are shown in the crust. L2 in the crust is the boundary of geological segments. L1 is the upper boundary of the subducting Philippine Sea slab. The reflected waves of the boundaries (L1 and L2) are shown in Fig. 2c. 12 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 Pg phase. A lateral large variation of the sedimentary layer was found. The Tonami basin is located in the northern part of the survey line. The delay of arrivals at profile distances of 20–40 km can be explained by the structure with the low-velocity Tonami basin (Figs. 4 and 5a). The thickness of the sedimentary layer beneath the Tonami basin is estimated to be ~3 km. A layer with a P-wave velocity of 5.3–5.8 km/s and a thickness of ~5 km is located beneath the sedimentary layer along the length of the profile. The P-wave velocity for the lower part of the upper crust is estimated to be 6.0–6.4 km/s. The lower crust P-wave velocity is 6.6–6.8 km/s. The thickness of the lower crust is about 10 km. The uppermost mantle velocity in the central Japan region is estimated from the arrival time data of the Pn phase. In this experiment, arrival-time data with apparent velocity values of 7.7–8.0 km/s are observed at ranges of about 140–170 km for shots of J1 and J2. The velocity of the uppermost mantle is 7.6–7.9 km/s. 3.2. Configuration of the subducting Philippine Sea plate Two remarkable clear later arrivals are observed on the record section of shot J5, which is the southernmost shot point (Fig. 2c). The origins of the later phases were investigated. The two later arrivals are explained by the reflected waves at two boundaries beneath J5. The upper boundary is located at a depth range of about 10–20 km (L2 in Fig. 4). Seismic experiments with explosive sources have been done in this area (Matsu’ura et al., 1991; Sato et al., 2001; Kawamura et al., 2003). Sato et al. (2001) concluded the reflector was the lithological boundary between the Southern Shimanto belt and the Northern Shimanto belt. The deeper boundary was located at depths of 20– 35 km (Iidaka et al., 2003). The location of the deeper boundary (Figs. 4 and 5) is consistent with that of the upper boundary of the subducting Philippine Sea plate, which was estimated from iso-depth lines on a seismicity map (Yamazaki and Ooida, 1985). The later phase was identified as a reflected wave at the upper boundary of the subducting Philippine Sea plate. The configuration of the subducting Philippine Sea plate was estimated by Iidaka et al. (2003). However, the configuration of the subducting plate was not estimated well because the reflected wave was only observed at a record section of one shot. It was very difficult to locate accurately the subducting Philippine Sea plate because of the trade-off between the dip-angle of the boundary and velocity structure. To estimate the location of the subducting Philippine Sea plate precisely, the reflected waves have to be observed at record sections of several shots at different locations. In this study, the reflected waves are detected on the record sections of shots J4 and T6 in addition to shot J5 record section (Figs. 3 and 5e). The configuration of the subducting Philippine Sea plate is revealed. The dip angle of the Philippine Sea plate obtained is 268 at the depth range of 20–26 km (Fig. 4). A flat Philippine Sea plate model with a dip angle of 268 explains most of the observed data (Fig. 5). However, the theoretical arrival-time curve does not fit well with the observed values at profile distances of 170–200 km (Fig. 5f-2). A distorted plate boundary is required to explain the observed traveltime data. An adequate velocity-model with a distorted plate boundary is obtained by forward modeling with the ray-tracing method (Fig. 6). A structural model, with the upper boundary of the Philippine Sea plate distorted at the depth range of 26–33 km, was derived. 3.3. Depth variation of the reflected-wave amplitude A remarkable reflected wave at the upper boundary of the subducting Philippine Sea plate is observed on the record section of the shot J5 (Figs. 2 and 3). We researched the amplitude of the reflected waves. Usually, site effects on amplitude data are expected to be large. We have to estimate the crustal response using seismic data with different sources. However, the seismic stations were set up to observe explosive shots, and were operated only for several hours. No far-field earthquake was observed at the seismic stations. We therefore investigated the amplitude ratio of the reflected wave at the upper boundary of the subducting Philippine Sea plate to the first arrival wave. The waveforms of the reflected wave at the upper boundary of the subducting plate are not impulsive, so T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 13 14 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 Fig. 5. Observed travel times and calculated travel times using the P-wave velocity model in Fig. 4 are shown by colored bars and black symbols, respectively. (a), (b), (c), (d), (e), and (f-2) show the travel time data for the shots J1, J2, J3, J4, T6, and J5, respectively. (f-1) shows the ray diagram of shot J5. T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 15 Fig. 6. Ray diagram using a model with the distorted upper boundary of Philippine Sea plate. The reflection points of the large-amplitude reflected wave are located at the depth of 26–35 km (white line). The reflection points, where the amplitude data can be explained by the reflected waves at the layer with the velocity of 6.5 km/s, are located on the yellow line. a time window with a length of 1.0 s is used for selecting the first arrivals and reflected waves. Maximum amplitude values of the two-phases are picked for each window and compared. The amplitude ratio (the maximum amplitude within the timewindow of the reflected wave/the maximum amplitude within the time-window of the first arrival) is plotted along with distance from the shot in Fig. 7. At a distance of 0–30 km, the amplitude ratio obtained is less than 1.0. At a distance of 30–70 km, however, the amplitude ratio is about 1.0–8.0. The large amplituderatio values are observed with a large perturbation. At the seismic stations with a distance greater than 70 km from the shot, the amplitude ratio obtained is about 1.0–4.0. The amplitude of the observed body wave O(f) is expressed by the following form: Oð f Þ ¼ A4B4S ð f Þ4Pð f Þ4C ð f Þ4I ð f Þ A is radiation pattern, B is geometrical spreading factor, S( f ) is source spectrum, P( f ) is path effect, C( f ) is crustal response, and I( f ) is instrumental response. ( f ) is frequency. We are able to assume that the S( f ), C( f ), and I( f ) parameters are cancelled using the amplitude ratio of the reflected wave to the wave of first arrivals. The value A is not considered here because an explosive source is used. B and P( f ) were evaluated using a ray-tracing method. A raytracing code of SEIS83 (Cerveny and Psencik, 1983) was used for the calculation. 16 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 The attenuation effect along the ray path has to be considered. Iwasaki et al. (1994) obtained a detailed crustal structure in the Tohoku region, northern Japan, using data from explosive sources. The Q structure was estimated by analyzing amplitude data of the explosive sources. The following Q structure model is used for the calculation based on the results of Iwasaki et al. (1994). The Q value of 100 is assumed for the sedimentary layer. The Q value of 150 is given to the layer beneath the sediment. The lower part of uppercrust and the lower-crust are assumed to have Q values of 300 and 400, respectively. The Q value at the uppermost mantle is assumed to be 600. The Q value of the subducting Philippine Sea slab is given to be 2000. The geometry of the subducting Philippine Sea plate with the flat plate boundary model (Fig. 4) was used for the calculation. The distorted plate boundary model (Fig. 6) is much better than flat plate boundary model for explaining the observed arrival-time data. The distorted plate boundary model has to be used for the calculation to evaluate amplitude accurately. However, the ray-tracing method is used in this calculation. It is difficult to evaluate the focusingand defocusing-effects on amplitude data using ray theory. The simple assumption (i.e., flat plate boundary) was therefore adopted for the calculation. In this study, a thin layer with a thickness of 1 km was assumed to be located at the top of the subducting plate based on the results of previous studies (Kurashimo et al., 2002; Kodaira et al., 2002). Several velocity models with different P-wave velocities at the thin layer were examined to explain the observed amplitude data. The amplitude ratio is calculated for the models with P-wave velocity values of 8.0, 6.5, 4.0, 3.0, and 2.0 km/s at the thin layer (Fig. 7). In the distance range of 0–30 km, the observed amplitude ratio of the reflected wave to the first arrival can be explained by a model with a P-wave velocity Fig. 7. Amplitude ratio data of reflected waves at the upper boundary of the subducting Philippine Sea plate to first arrival waves. (a) for ranges 0–30 km, (b) for ranges 30–70 km, and (c) for ranges 70–160 km. The observed data (red lines and symbols) are compared to the calculated data. The velocity model for the calculated data has a thin layer with different velocity values at the top of the slab. Velocity values are 8.0 km/s (Model-a), 6.5 km/s (Model-b), 4.0 km/s (Model-c), 3.0 km/s (Model-d), and 2.0 km/s (Model-e). T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 of 6.5 km/s at the upper boundary of the subducting slab. The velocity is consistent with that of the subducting oceanic crust at the Philippine Sea plate (e.g., Kurashimo et al., 2002). At distances of 30–70 km, large amplitude ratios (1.0–8.0) are observed. In the calculation, the amplitude of the reflected wave increases, as the velocity of the thin layer decreases. However, the observed amplitude ratio value is much larger than that of the low-velocity layer with a P-wave velocity of 2.0 km/s. In the distance range of 70–160 km, the observed amplitude data have ratios within 1.0–4.0, but the theoretical amplitude ratios are less than 1.0, even though an extremely low velocity model (V p=2.0 km/s) is used. The rays which arrive at distances of 0–30 km are reflected at depths of about 20–26 km, and have amplitudes consistent with those of reflected waves at the boundary between the mantle wedge and the subducting oceanic crust. The rays arriving at the seismic stations at distances of 70 km and 160 km are reflected at depths of about 30 km and 35 km, respectively. At reflection points with depths of 26–30 km, extremely large amplitude ratios of 1.0–8.0 are observed. At the depth range of 30–35 km, the reflected wave amplitude is several times (1.0–4.0) larger than that of first arrivals. At reflection points deeper than 26 km, the amplitude ratio data suggest that the amplitude of the reflected wave is much larger than that of the model with an extremely low-velocity layer (V p=2.0 km/s) located at the top of the slab. 4. Discussion The crustal structure in this area was researched by Aoki et al. (1972). The major difference between the model of Aoki et al. (1972) and our model is the thickness of the lower crust. A crustal model in the Chubu region obtained by Aoki et al. (1972) showed a lower crust with a thickness of 6 km, which was estimated from data from seismic stations with spacings of about 10 km. The spatial density of the seismic stations was insufficient to detect refracted waves from the lower crust. In this study, the average spacing of the seismic stations is 669 m, which is sufficient to detect a refracted wave at the lower crust, and a thickness of about 12 17 km was derived. The seismic velocity model obtained here explains well the first arrival time data at a distance range of 70–110 km for the shots of J5 and T6 (Fig. 5). The uppermost mantle velocity in the central Japan region is estimated from the arrival time data of the Pn phase. In this experiment, arrival-time data with apparent velocity values of 7.7–8.0 km/s are observed at the epicentral distance of about 140–170 km for shots of J1 and J2. The velocity of the uppermost mantle obtained is 7.6–7.9 km/s. The values are consistent with those of the previous studies (Aoki et al., 1972; Takeda, 1997). The boundary L2 is identified as the lithological boundary between the Southern Shimanto belt and the Northern Shimanto belt. The location of the L2 boundary and the velocity value at the shallower part of the boundary were well defined because many rays traveled through the region. However, the velocity value at the lower part of the L2 boundary was not clear because few rays traveled through the area. The location of the boundary was only derived and the velocity change at the boundary was not estimated. The location of the upper boundary of the Philippine Sea slab was estimated by Yamazaki and Ooida (1985) using a seismicity map with the assumption that the upper boundary of the slab is located just above the locations of micro-earthquakes. The upper boundary of the Philippine Sea slab obtained here is 2–5 km shallower than that of Yamazaki and Ooida (1985). In the Shikoku region, central part of the Nankai trough, Kurashimo et al. (2002) suggested that the upper boundary of the Philippine Sea slab was located at 5–10 km above the seismic region of micro-earthquakes from an analysis of seismic refraction data. The upper boundary of the subducting slab seems to be located a few kilometers above the seismic zone. The dip angle of the subducting Philippine Sea plate of 268 is estimated at the depth range of 20–26 km. In this depth range, the reflected waves at the upper boundary of the subducting Philippine Sea plate are observed on the record sections of shots J4, J5, and T6. The reflected waves, for which the reflection points are deeper than 26 km, are only observed at the record section of shot J5. It was difficult to estimate the dip angle of the slab that is deeper than 26 km. 18 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 The record section data of shot J5 (Fig. 6) suggests that the subducting Philippine Sea plate deeper than 26 km is distorted. The shape of the distortion is not well defined because of the trade-offs of several parameters (i.e., velocity and shape, etc.). In the Nankai region, located west of this research area, large inter-plate earthquakes have occurred cyclically. Seismic experiments with explosive sources have been done in this area (e.g., Kodaira et al., 2000, 2002; Kurashimo et al., 2002, 2003), and a seismic velocity structure model of the subducting Philippine Sea plate obtained. The reflected wave observed at the upper boundary of the subducting plate has an amplitude that is much larger than that of first arrival. Kurashimo et al. (2003) tried to estimate the reflection coefficient at the boundary using the Amplitude Variation with Offset (AVO) method. A thin layer with a thickness of 200 m and velocity of 4.0 km/s is required at the top of the subducting Philippine Sea plate. In the study of Kodaira et al. (2002), the amplitude analysis of the reflected waves at the subducting slab suggested that a thin lowvelocity layer with a velocity of 3 km/s was located at the top of the slab with a depth range of about 15–30 km. Therefore, previous evidence suggests the existence of an extremely low velocity zone at the top of the subducting Philippine Sea plate. In this study, the amplitude data analyzed by the velocity model is consistent with oceanic crust at the top of the subducting Philippine Sea plate within the depth range of 20–26 km. At a depth deeper than 26 km, however, the amplitude ratio is extremely large compared to those of the theoretical models with large variations of P-wave velocity (2.0–8.0 km/s). To explain the observed large amplitude data, the following models were considered: (1) layered structure at the top of the subducting slab, (2) focusing effect on the reflected waves. In this analysis, the maximum amplitude values are obtained for a time window with a length of 1 s, because the reflected waves are not an impulsive phase. If the uppermost part of the subducting Philippine Sea plate has a layered structure, and consists of many thin layers, the amplitude of the reflected wave will be increased by the interference of the reflected waves at the many layers. If the large amplitude reflected wave is caused by a layering structure at the top of the plate boundary, a spectrum analysis of the large amplitude waveform should suggest a dominant frequency corresponding to the layering structure. However, the FFT analysis of the reflected wave did not show any significant peak in spectrum data. If the upper boundary of the Philippine Sea plate is distorted, some perturbation should be observed in the arrival-time data of the reflected waves, and a large amplitude reflected-wave might be observed by the focusing effect at the distorted interface. A distorted upper boundary model is required to explain the observed travel-time data (Fig. 6). The ray path of the reflected wave, which arrives at profile distances of 170–200 km, reflects at the upper boundary of the slab with the depths of 26–30 km. The depth range is the large amplitude ratio zone (Figs. 6 and 7b). The large-amplitude reflected waves might be caused by the focusing effect. In the Tokai region, a seismic survey with refraction tomography and pre-stack depth migration of wideangle seismic data imaged a trough parallel cyclic ridge subduction in the oceanic area (Kodaira et al., 2003). If the subducted ridge is located at a depth of 26–30 km, the upper boundary should be distorted. The distorted boundary might cause a large amplitude reflected-wave due to the focusing effect. However, it has not been modeled well because the configuration of the distorted boundary was not obtained accurately, and the ray-tracing method is not suitable for evaluating the focusing effect on amplitude data. 5. Conclusions A seismic experiment with six explosive sources and 391 seismic stations was conducted in central Japan. The crustal structure of the area was obtained. A large lateral variation in the thickness of the sedimentary layer was found. Beneath the Tonami basin, the thickness of the sediment is estimated to be ~3 km. The layer located beneath the sedimentary layer has a P-wave velocity of 5.3– 5.8 km/s. The P-wave velocity of the lower part of the upper crust is estimated to be 6.0–6.4 km/s. The lower crust P-wave velocity is 6.6–6.8 km/s. The P-wave velocity at the uppermost mantle is 7.6–7.9 km/s. T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 Large-amplitude reflected waves at the two boundaries are observed. One of the boundaries is located at depths of 10–20 km, which is the lithological boundary between the Southern Shimanto belt and the Northern Shimanto belt. The other boundary at depths of about 20–35 km is interpreted to be the upper boundary of the subducting Philippine Sea plate. The dip angle of the plate is estimated to be 268 at the depth range of 20–26 km. The upper boundary of the subducting Philippine Sea plate could be distorted at locations deeper than 26 km. To know the physical properties of the upper boundary of the Philippine Sea plate, the amplitude ratio of the reflected wave to first arrival was researched. At depths of 20–26 km, the amplitude ratio data is consistent with the wave being reflected at the interface of the oceanic crust of the slab with a P-wave velocity of 6.5 km/s. However, the amplitude ratio is much larger than that of acceptable velocity models at depths deeper than 26 km, which is the depth range at which the upper boundary of the slab is distorted. The large-amplitude reflected waves might be caused by the distorted upper boundary of the subducting Philippine Sea plate with a focusing effect. However, the amplitude data has not been modeled well because the configuration of the distorted boundary was not obtained accurately, and the ray-tracing method is not suitable for evaluating the focusing effect on amplitude data. Some unique seismic structure has to be considered to explain the observed data. A unique structure will provide important information for understanding the mechanism of inter-plate earthquakes. More detailed analyses of the waveform are required to reveal the physical properties at the boundary, and provide good information for understanding the mechanism of inter-plate earthquakes. Acknowledgments This experiment was supported by a grant from JAMSTEC. We thank all members of the Research Group for Seismic Expedition in Central Japan for the data acquisition employed in this experiment. RAYINV and SEIS83 computer program codes were used in the calculation. 19 References Aoki, H., Tada, T., Sasaki, Y., Ooida, T., Muramatsu, I., Shimamura, H., Furuya, I., 1972. Crustal structure in the profile across central Japan as derived from explosion seismic observations. J. Phys. Earth 20, 197 – 223. Cerveny, V., Psencik, I., 1983. Program Package SEIS83. Harada, S., Yoshida, A., Aketagawa, T., 1998. Configuration of the Philippine Sea slab and seismic activity in the Tokai region. Bull. Earthq. Res. Inst. 73, 291 – 304. Iidaka, T., Iwasaki, T., Takeda, T., Moriya, T., Kumakawa, I., Kurashimo, E., Kawamura, T., Yamazaki, F., Koike, K., Aoki, G., 2003. Configuration of subducting Philippine Sea plate and crustal structure in the central Japan region. Geophys. Res. Lett. 30, 23-1 – 23-4. Ishida, M., 1992. Geometry and relative motion of the Philippine Sea Plate and Pacific plate beneath the Kanto-Tokai district, Japan. J. Geophys. Res. 97, 489 – 513. Iwasaki, T., Yoshii, T., Moriya, T., Kobayashi, A., Nishiwaki, M., Tsutsui, T., Iidaka, T., Ikami, A., Masuda, T., 1994. Precise P and S wave velocity structure in the Kitakami massif, Northern Honshu, Japan, from a seismic refraction experiment. J. Geophys. Res. 99, 22,187 – 22,204. Kato, N., Hirasawa, T., 1999. A modeling for possible crustal deformation prior to a coming large interplate earthquake in the Tokai district, central Japan. Bull. Seismol. Soc. Am. 89, 1401 – 1417. Kawamura, T., Onishi, M., Kurashimo, E., Ikawa, T., Ito, T., 2003. Deep seismic reflection experiment using a dense receiver and sparse shot technique for imaging the deep structure of the Median Tectonic Line (MTL) in east Shikoku, Japan. Earth Planets Space 55, 549 – 557. Kodaira, S., Takahashi, N., Nakanishi, A., Miura, S., Kaneda, Y., 2000. Subducted seamount imaged in the rupture zone of the 1946 Nankaido earthquake. Science 289, 104 – 106. Kodaira, S., Kurashimo, E., Park, J.-O., Takahashi, N., Nakanishi, A., Miura, S., Iwasaki, T., Hirata, N., Ito, K., Kaneda, Y., 2002. Structural factors controlling the rupture process of a megathrust earthquake at the Nankai trough seismogenic zone. Geophys. J. Int. 149, 815 – 835. Kodaira, S., Nakanishi, A., Park, J.-O., Ito, A., Tsuru, T., Kaneda, Y., 2003. Cyclic ridge subduction at an inter-plate locked zone off central Japan. Geophys. Res. Lett. 30, 72-1 – 72-4. Kurashimo, E., Tokunaga, M., Hirata, N., Iwasaki, T., Kodaira, S., Kaneda, Y., Ito, K., Nishida, R., Kimura, S., Ikawa, T., 2002. Geometry of the subducting Philippine Sea plate and the crustal and upper mantle structure beneath eastern Shikoku Island revealed by seismic refraction/wide-angle reflection profiling. Zisin 54, 489 – 505. Kurashimo, E., Hirata, N., Iwasaki, T., 2003. Physical properties of the top of the subducting Philippine Sea plate beneath the SW Japan arc by AVO analysis. Programme and Abstracts, Seismix2003, p. 90. Matsu’ura, R.S., Yoshii, T., Moriya, T., Miyamachi, H., Sasaki, Y., Ikami, A., Ishida, M., 1991. Crustal structure of a seismicrefraction profile across the Median and Akaishi tectonic lines, central Japan. Bull. Earthq. Res. Inst. 66, 497 – 516. 20 T. Iidaka et al. / Tectonophysics 388 (2004) 7–20 Nakanishi, I., 1980. Precursors to ScS phases and dipping interface in the upper mantle beneath southwestern Japan. Tectonophysics 69, 1 – 35. Nakanishi, I., Suyehiro, K., Yokota, T., 1981. Regional variations of amplitudes of ScSp phases observed in the Japanese Island. Geophys. J. R. Astron. Soc. 67, 615 – 634. Sato, H., Ito, T., Miller, K., Iwasaki, T., Hirata, N., Ohishi, M., Kaip, G., Kato, N., Kikuchi, S., Kwiatkowski, A., Kurashimo, E., Kawamura, T., 2001. Seismic reflection image of Lithospheric structure beneath Shidara, using explosive sources from the 2001 deep seismic profiling in central Japan. Eos, Trans, AGU 82, F1152. Takeda, T., 1997. Reanalysis for seismic refraction data in Nagano Basin Area, Japan—crustal structure in central Japan, MS thesis, University of Tokyo, p. 26. Yamazaki, F., Ooida, T., 1985. Configuration of subducting Philippine Sea plate beneath the Chubu district, central Japan. Zisin 38, 193 – 201. Zelt, C.A., Smith, R.B., 1992. Seismic travel inversion for 2-D crustal velocity structure. Geophys. J. Int. 108, 16 – 34.