MTL Wakamiya Aoyagi Hakushu Hoouzan Onajika-toge Shimotsuburai Ichinose SEISMICITY AND CRUSTALL STRUCTURE ALONG THE SOUTHERN JAPANESE ALPS SEGMENT OF THE ITOIGAWA-SHIZUOKA TECTONIC LINE Panayotopoulos Y.[1]; Hirata N.[1]; Sato H.[1]; Iwasaki T.[1]; Kato A.[1]; Imanishi K.[2]; Cho I.[2]; Kuwahara Y.[2] [1] Earthquake Research Institute, The University of Tokyo, P.O , Yayoi, Bunkyo ward, Tokyo Japan [2] Geological Survey of Japan, AIST Tsukuba Central 1,Tsukuba, Ibaraki , Japan Tel: , Fax: Introduction Large intraplate earthquakes are a common phenomenon in the Japanese islands, with the 2004 Niigata Prefecture earthquake being a very recent example. These events take a heavy toll of human lives and severely damage the infrastructure of the affected area. The wide area along the Itoigawa-Sizuoka Tectonic Line (ISTL) is considered to be with the highest probability of a major inland event occurrence in the near future (fig 1). The tectonic relation between the different parts of the ISTL fault system and the surrounding geological units is yet to be full explained. In order to understand the active tectonics that control the earthquake genesis mechanisms, it is essential to investigate the microseismic activity near the faults, the deep layout of active faults and the seismic velocity structure of the earth crust in the surrounding area. In order to reveal seismic activity that might be related to the southern segment of the ISTL fault system, we have deployed a dense seismic station array of 60 stations (fig 2). We combined our stations with the permanent seismic station network in the area and precisely relocated all the seismicity that occurred during the period of this project. In this presentation we correlate the observed seismicity with the deeper parts of the faults and use it to estimate the 3d seismic wave velocity structure of the area. This study is supported by the Japanese Ministry of Education, Culture, Sports, Science and Technology under a special project titled “A high priority investigation in the ISTL fault zone”. 2.Geological setting The Itoigawa-Sizuoka Tectonic Line (ISTL) is a major geological structure that divides Japan into NE and SW parts. It was formed as a normal fault in the early Miocene and represents the southwestern boundary of the northern Fossa Mangna sedimentary basin (fig 1). During the Pliocene the ISTL is reactivated as a reverse fault due to tectonic inversion after the collision of the Izu-Bonin arc with the Japanese arc. The deep of the ISTL fault system is not constant along its segments but changes from an East dipping thrust fault in the north to a strike slip fault in the center and to a west dipping trust fault to the south. The central and northern parts of the ISTL have a slip rate of mm/yr and 3.0 mm/yr respectively, which represents one of the highest slip rates in all off the Japanese islands. Paleoseismological studies in the area suggest an average recurrence interval of 1250 – 1500 years for major events. The southern part of the ISTL fault system constitutes the boundary between the Cretaceous-Tertiary accretionary prism to the west and the Izu-Bonin units to the east. It is consisted of two parallel thrust fault systems identified by most researchers as the geological ISTL ( Hakushu, Hoouzan, Onajika-toge faults) on the western side, and the neotectonically active ISTL ( Shimotsuburai, Ichinose faults ) on the eastern side (fig 6b). The southern segments of the ISTL fault system have a lower slip rate than the northern ones, with an estimated mm/yr and a recurrence interval of approximately 5000 years. The last event has occurred somewhere between BC, which implies that this area has high probability of a major event occurrence in the near future. I.S.T.L N.F.M Paleozoic-Mesozoic rocks Miocene rocks  Yatsugatake volcano  Yatsugatake volcano Mt. Fuji  Crataceous-Tertiary rocks Crataceous-Tertiary rocks Izu-Bonin rocks Izu-Bonin rocks Fig. 1 ) Geological map of the surveyed area. 3.Observations In order to reveal seismic activity, which may be related to the ISTL activities, we have deployed 2 dense seismic station arrays in the south-central part of ISTL from 15 Sep to 23 Dec. 2005, with prospect to extend our investigation in the central and northern part in the coming years (fig 2). We deployed 2 parallel linear arrays consisted of 30 and 25 stations each at an average spacing of 1.5 km and 5 stations in the surrounding area. We used 3 -component 1-Hz seismometers and long-term off- line recorders with a sampling rate of 200 Hz. We have combined our stations with the temporary station array deployed by the Geological Survey of Japan (AIST), the local network in this area deployed by the Japanese Meteorological Agency (JMA), the Earthquake Research Institute, the University of Tokyo (ERI) and the National Research Institute for Earth Science and Disaster Prevention (NIED) to relocate the regional seismicity observed by the JMA (fig 2). In addition, we integrated into our observations the station network deployed for the purposes of the 2003 investigation. The 2003 temporary network run for a 2 month period from 25 August to 16 October 2003; consisting of 49 linear array stations cutting across the ISTL fault trace and 9 stations scattered in the surrounding area. In total we utilised 235 stations for the purposes of this investigation. 4.Dataset Our combined dataset consists of 89 events recorded during the 2003 observation and 345 events during the 2005 observation (fig 3). In addition, we have included in our dataset 2 vibroseis shots from the 2003 observation and 11 explosive shots that were scheduled for the same period as the 2005 observation. Several of these events have not been located by the JMA, and were estimated using our temporary stations. We manually re-picked P and S phases from the waveform data recorded by the local network and our station arrays. Thus we have obtained 9675 P phase arrivals and 9723 S phase arrivals, which is more than triple the phases picked by the Japan Meteorological Agency (JMA). We reprocessed these events using the Double Difference tomography method (Zhang and Thurber, 2003), in order to obtain a detailed 3d seismic wave structure of the crust in the surveyed area. 5.Tomography inversion We calculated the Vp, and Vs values at nodes of a 3D grid matrix, with 20 x 11 x 13 nodes at XYZ directions respectively (fig 4). The x axis of the coordinate system was tilted 170 to the north in order to be parallel with our station arrays. The grid points on x axis were set at 5km spacing for the central 60 km and 15 km or 20 km in the perimeter. On y axis we have lined them on 9 lines at -56, -32, -16, -8, 0, 8, 16, 32, 56 km, so that the -8 an 8 km lines overlap with our stations. At depth we have grid points at 0 、 3, 6, 9, 12,16, 20, 30, 35, 100, 200 km, counting from sea level. For the initial 1d velocity model, we selected the same 1d velocity model we utilized wile re- picking the events. This model represents the average P wave velocity structure of the crust and upper mantle underneath the Japanese arc, that is used by ERI for all of its earthquake determinations. The final 3D velocity model achieved a 64% reduction in the root-mean-square (RMS) travel time residuals, from an initial RMS of 297 ms to a final ms after 8 iterations. The calculated Vp, Vs values were plotted along 5 cross sections matching the -16, -8, 0,8 16 km lines to a depth of 20 km (fig 6a). The reliability of the tomography inversion results was checked by the means of a checkerboard resolution test. We assigned ±5% velocity perturbations of the initial velocity model and created a synthetic dataset using the source-receiver distribution of the real data. We then inverted the synthetic dataset with the initial unperturbated velocity model, using the same parameters as for the real data. The checkerboard patterns were reproduced along a 60 km region, up to a depth of 20 km (fig 5). x y Fig. 2) Seismic stations used in this study. Fig. 3) Distribution of the events used for the travel time inversion. Fig 4) Grid node distribution ： x: 5km spacing for the central 60km 、 15 km or 20 km in the perimeter. y: 9 lines at -54, -32, -16, -8, 0, 8, 16, 32, 54 km. At depth: 0 、 3, 6, 9, 12,16, 20, 30, 35, 100, 200 km Fig 5) Checkerboard resolution test. ±5% velocity perturbations of the initial velocity model. The checkerboard patterns were reproduced along a 60 km region, up to a depth of 20 km Velocity perturbationsVp resolutionVs resolution Wakamiya faultAoyagi faultWakamiya faultAoyagi fault Hakushu fault Hoouzan faultShimotsuburai fault Hoouzan faultShimotsuburai fault Onajika-toge faultShimotsuburai fault Onajika-toge faultShimotsuburai fault Onajika-toge faultIchinose fault Onajika-toge faultIchinose fault 6.Results and discussion The seismicity relocated buy the tomographic inversion, appears to be averagely 1.5 km shallower than the initial JMA determination (fig 7). These is mainly due to the fact that the 3D crustall structure in the area is not laterally homogeneous. It is clear in all of the Vp cross sections that there is a wide wedge like area of low velocities between -30 and 5 km and till approximately 15 km depth (fig 8). We identify this area as the Cretaceous-Tertiary accretionary prism of the Japanese arc. To the east of this area we observe a low Vp zone to a depth of 5-10 km which is followed by a sharp increase in the velocity distribution from km. This area represents the upper and middle to lower Izu-Bonin arc units respectivly. The relocated events on the cross section have lined up on a low / high Vp boundary between 5 to 20 km depth. The deeper extension of the Hakushu fault of the ISTL fault system in this region seems to be the boundary between these low Vp and high Vp zones, which can be identified as the accretionary prism units and the Izu-Bonin arc sequences. It forms a thrust fault that dips approximately 45 0 ~55 0 towards the west. The seismicity in the area may occur on the deeper extension of the Hakushu fault; and if so, this could be evidence that the deeper part of the geological ISTL is connected to the active fault at depth. The seismicity on the rest of the cross sections is not enough to reveal the deep structure of the rest of the ISTL fault system segments in the area. Conclusion We have conducted a temporary network array survey in the Southern Japanese Alps segment of the ISTL from 15 September 2005 to 23 December We have combined our stations with the routine network in the area in order to accurately estimate the local seismicity. We have processed these events according to the DD tomography method and estimated the crustal velocity structure in the area. The deeper extension of the ISTL in southern Alps region seems to be the boundary between low Vp and high Vp zones, which can be identified as the Japanese arc accretionary prism units and the Izu-Bonin arc sequences. ISTL forms a thrust fault that dips approximately 45 0 ~55 0 towards the west. The seismicity below the 8 km profile may occur on the deeper extension of the ISTL. If so, this could be evidence that the deeper part of the geological ISTL is connected to the active fault at depth. Fig. 6a) Relative location of the cross sections and the geological units in the area. Fig. 6b) Distribution of the ISTL fault segments in the Souther Japanese Alps region. Vp profilesVs profiles Fig. 7) Relocated events by the Double Difference tomographic inversion. The seismicity appears to be averagely 1.5 km shallower than the initial JMA determination Fig. 8) Vp and Vs cross sections. A wide wedge like area of low velocities between -30 and 5 km and 15 km depth is prominent in all of the profiles.