IASPEI

B.460 IUGG2003

lower crust and above-mentioned northward dipping reflectors are converged to the trench-ward dipping reflector forming wedge-thrust structures. This crustal structure demonstratedby reflection pattern suggests that the main part of the accretioanry material is injected intothe mid-crust of the overlying plate associated with wedge-thrusting since the lateCretaceous.

SS03/07A/A02-007 1100CRUSTAL SECTION ACROSS THE HIDAKA COLLISION ZONE, HOKKAIDO, JAPAN,AS INFERRED FROM SEISMIC REFRACTION/REFLECTION PROFILING

Takaya IWASAKI, Research Group of HOKKAIDO TRANSECT (Earthquake Research Insttitute,the University of Tokyo)

The Hidaka collision zone, central part of the Hokkaido Island, Japan, is known as anongoing collision zone between Kuril Forearc(KA) and the Northeast Japan Arc (NJA) sincemiddle Miocene. A multidisciplinary project of Hokkaido Transect in 1998-2000 revealedvarious scale structural heterogeneity across this coll ision zone by seismicrefraction/reflection profiling and very dense earthquake observation. A 227-km longrefraction profile was undertaken to determine the whole crustal structure from NJA to KA. Aseries of seismic reflection lines, whose total length is 138 km, were concentrated in thecentral part of the refraction line to get a clear image of crustal deformation associated withthe collision. The data obtained elucidated a complicated collision structure. In the easternpart of the profile, KA is covered with 0.3-4km thick highly deformed sedimentary layer,beneath which two eastward dipping reflectors are imaged in a depth range of 10-20 km,probably representing obducting middle or lower curst of KA. Actually its outcropped part hasa relatively higher velocity and Vp/Vs than those of the surrounding part. Beneath thesereflectors, another flat and westward dipping reflectors are situated at 25 and 25-27 kmdepths respectively. The obtained layer geometry forms wedge-like (crocodile) patterns,probably expressing that the crust of KA is delaminated into two or three segments beneaththe Hidaka Mountains. The western part of the profile (NJA), which belongs to the fold-and-thrust belt of the collision zone, is characterized by a very thick (more than 5-10 km)sedimentary package including two or more velocity reversals. Beneath this package, thecrystalline crust of NJA is traced with a slight eastward dip down to 20-25 km. The crustalimage obtained is much complicated as compared with that in the southernmost part of thecollision zone, where the crust of KA is delaminated into to segments forming a single wedgestructure, indicating a significant regional difference in style of crustal deformation.

SS03/07A/A02-008 1115CRUSTAL STRUCTURE AROUND THE NORTHERN PART OF THE ITOIGAWA-SHIZUOKA TECTONIC LINE (ISTL), CENTRAL JAPAN, FROM REFRACTION/WIDE-ANGLE REFLECTION DATA

Tetsuya TAKEDA, Takaya IWASAKI, Hiroshi SATO, Shin'ichi SAKAI, Takashi IIDAKA(Earthquake Research Institute, University of Tokyo)

The Itoigawa-Shizuoka Tectonic Line (ISTL) is a major structural boundary and one of themost active faults, which divides SW and NE Japan. To reveal the crustal structure aroundnorthern part of ISTL is important for better understanding of the development of Japaneseislands and estimation of strong ground motion produced from deeper extension of ISTLactive fault system. To reveal the crustal structure around the northern part of ISTL, therefraction/wide-angle reflection data from three seismic lines obtained in 1967-1991 were re-analyzed with adding a new seismic data carried out in Itoshizu 2002. The three dimensionalfeature of velocity structure of upper crust around northern part of ISTL is clearlydemonstrated. Some reflectors located in the middle to lower crust were obtained by wide-angle reflection data. From a wide view, crustal structure is different between the easternand western sides of ISTL. It is said geologically that ISTL played an important role in theMiocene opening of the Sea of Japan and bending of Japan island arc. It means that thedifference of reflector distribution bounded by ISTL indicates the possibility of horizontalmovement in the opening of the Sea of Japan or crustal deformation undergone by igneousactivity. In the velocity structure, the layer with 4km/sec is found to have 4km depth in thewestern side of ISTL. It is consistent with the fact that sedimentary rocks since Miocene isthickly distributed in this area. ISTL is found to be east-dipping from gravity and reflectiondata. To take account that ISTL is lifting up in the east as a reverse fault at present, we thinkthat ISTL was formed as east-dipping normal fault in mid-Miocene during initial phases andhas been reactivated as a reverse fault owing to subsequent crustal shortening. Additionally,by analyzing travel times of micro-earthquakes below ISTL, we expect to obtain moredetailed structure of the eastern side of ISTL, which is deformed strongly owing to crustalshortening.

SS03/07A/A02-009 1130REFRACTION AND WIDE-ANGLE REFLECTION STUDIES BY USE OF MCS AND OBSIN MARMARA SEA, TURKEY (SEISMARMARA 2001)

Hideki SHIMAMURA1, Anne BÉCEL2, Jean-Claude LÉPINE2, Tuncay TAYMAZ3, Yoshio MURAI1,Philippe CHARVIS4, Yuichi NISIMURA1, Mireille LAIGLE2, Alfred HIRN2, Serdar OZALAYBEY5

(1ISV, Hokkaido University, 2Sismologie Expérimentale, Institut de Physique du Globe de Paris, 3ITU,Istanbul Technical University, Maslak, Istanbul, 4IRD-Géosciences Azur, Villefranche, 5TUBITAK-MAM, Marmara Research Center, Gebze)

The SEISMARMARA survey 2001 was carried out as a multi-method approach of seismicstructure and activity, in the frame of the Turkish-French-Japan cooperation on the MarmaraSea, Turkey, after the 1999 Izmit earthquake occurrence on the North Anatolian Fault.MCS survey was made by the French N/O Nadir of Ifremer, and 37 OBS of ISV Hokka?Eowere deployed, in order to record airgun shots and natural earthquakes, and recovered by aTurkish vessel MTA-Sismik 1. We present the first results of modelling proceeds on the twolongest, East-West lines of Leg 1. . In order to assess effects of off-line propagation in thisregion of strong 3D structural variation, North-South cross-lines of the MCS grid can bedrawn in as well. Another E-W line along the southern rim of the Trough is modelled, tying inwith the Marmara-1 borehole that reached the Upper Cretaceous limestone, the basement tothe evolution of the Trough. The 8100 cu. in. capacity of the 12-airgun array of N/O Nadir,and its shooting in single-bubble mode, provided for signal strength, low frequency withrelatively short signal duration. The main regional line, striking E-W along the axis of theNorth Marmara Through, is 120 km long and has 13 OBS. The corresponding MCS stackwith 15-fold coverage, obtained at normal incidence with the 4.5 km long 360-channelsdigital streamer, is used for structural constraints for the shallow part in the refractionmodelling. The airgun array was also shot at 20 seconds interval to give a higher, 45-foldcoverage for MCS meant at finer resolution, and the comparison with velocity control givenby the streamer recording in shallow water will be used too Basement topography andstructure of the lower crust is sampled and modelled by these data sets. In the case of theline on the rim of the Trough, the velocity model can be compared to the MCS which gives areasonable structural image through the whole crust. On the axial line, new criticalconstraints are provided on the basement and deep structure across the succession ofbasins and highs, where the normal-incidence MCS image is complicated by strong multiples

and off-line echoes. These results will contribute to the analysis of the structural evolution ofthe Marmara trough with respect to the North Anatolian Fault, and bring constraints on thedepth limits of the seismogenic layer involved in earthquakes to occur in the Marmara seapart of the North Anatolian Fault. Derivation of the velocity structure provides a framenecessary to refine earthquake location procedures in general. More specifically, theanalysis of shot data on the 3 components of OBSs, give prime control on the local P and Svelocity heterogeneity under the OBS which will be used as input in a detailed study of thelocal earthquakes also recorded on these same OBSs.

SS03/07A/A02-010 1145REFLECTION OF THE CHANGES OF THE PHYSICAL PROPERTIES OF THE EARTHCRUST IN THE BULK-SOUND AND SHEAR WAVE-SPEED VARIATIONS

Alexei GORBATOV1, Brian L.N. KENNETT2 (1Pacific Region Data Centre, IFREE, JAMSTEC,2Research School of Earth Sciences, ANU, Australia.)

Examination of bulk-sound and shear wave speed variations at the active boundaries of thetectonic plates may provide valuable information about the compositional or physicalproperties of the Earth material and mechanic of collision processes because thisinformation can be directly compared with the results of laboratory experiments. Detailedglobal and regional tomographic inversions have been carried out using the arrival times forP and S wave ray paths with comparable coverage. The inversion was performed directly forbulk-sound and shear wave speed variations in the Earth´s mantle. Global inversion wasperformed using cells with sides of 2 by 2 degrees while for the regional tomography cellsfrom 0.5 to 2 degrees were chosen. A simulatanoeus inversion was undertaken for both theregional and global structures to minimize the influence of surrounding structures on theregional image. The nonlinearity of the inverse problem was taken into account by using ascheme which iteratively applied three-dimensional ray tracing and linearized inversion.Detailed regional tomography of the Western Pacific region reveals prominent correlationbetween bulk-sound and shear wave speed balance within the slab and the average age ofthe subducting lithosphere in the uppermost mantle of the subduction zones. Plates olderthan ~90Ma are more prominent on the shear wave speed images while younger slabs aremore distinctive in the bulk-sound wave speed. The 90Ma break would correlate with thesubduction of the full thickness of oceanic lithosphere and hence a mature thermal regime atthe trench. However this correlation can be strongly altered by thermo-mechanical processaffecting the subducting slab such as in the case of back-spreading and consequent rollback of the trench. This effect is well illustrated by subduction of old slab below Tonga-Kermadec. Roll-back of the trench is high below Tonga diminishing toward the Kermadecsubduction zone. Subducted slab below Tonga is dominated by bulk-sound anomalies whilethe shear wavespeed becomes more prominent below Kermadec. This behavior of bulk-sound and shear wave speed anomalies detected on the regional tomography of WesternPacific is consistent with the images on the global scale for the Pacific region in general.However, the most peculiar feature in this area is the strong anticorrelation between bulk-sound and shear wavespeed anomalies for the San Andreas Fault system where the zone ofstrongly positive bulk-sound anomalies correspond to negative shear wave speedperturbations. Observed anticorrelation may be explained by strong fracturation and partialwater saturation of this zone.

SS03/07A/A02-011 1200MELT GENERATING PROCESSES AT THE SOUTHEARN EAST PACIFIC RISEREVEALED BY THE ELECTRICAL CONDUCTIVITY STRUCTURE

Kiyoshi BABA1, Alan D. CHAVE2, Rob L. EVANS2, Greg HIRTH2, Randall L. MACKIE3 (1Institutefor Frontier Research on Earth Evolution, Japan Marine Science and Technology Center, 2Woods HoleOceanographic Institution, 3GSY-USA, Inc.)

Electrical conductivity structure models allow us to discuss mantle dynamics based on meltand water distribution beneath a mid-ocean ridge system. The electromagnetic data from theMantle ELectromagnetic and Tomography (MELT) experiment have been re-inverted for twosurvey lines. The major line crosses the East Pacific Rise at 17°S where the ridge segmentis thick suggesting an abundant magma supply. The northern line crosses at 15°45´S to thenorth of an overlapping spreading center on a magma-starved ridge segment. The data areinverted for a two-dimensional anisotropic conductivity structure that incorporates correctionfor three-dimensional topographic effects on the magnetotelluric responses. The modelspace allows for different conductivity values in the along-strike, cross-strike, and verticaldirections along with imposed constraints that the model be smooth and that the threeconductivities be as close together as possible. The strength of these constraints is variable,and hence a range of models from isotropic to anisotropic can be explored. For the southernline, two end-member models (one isotropic and one anistropic) were obtained which yieldsimilar RMS misfits. Both models have a major gross feature which is also seen in previousMELT inversions: an asymmetric conductivity structure with higher values to the west of theridge axis. In addition, the anisotropic model is more conductive in the cross-strike directionto the east of the rise at depths of 60-150 km, is quite resistive above 60 km, and has anarrow, highly conductive pipe in the vertical conductivity located immediately beneath theridge axis. First two features are well resolved required by the data. The last feature is notstrongly required by the data. But some tests show that it seems to be something real. Theanisotropic model is more consistent with other geophysical and laboratory data, and henceis preferred. These results suggest that melt exists over a wide region, but may be morehighly concentrated and connected in the vertical at the ridge axis. The deep ( 60 km)conductive region to the east of the ridgecrest is probably due to the preferred orientation ofolivine under wet conditions. The flat resistive-to-conductive boundary at 60 km agrees wellwith the inferred depth of the dry solidus of olivine. These results suggest that the extractionof water from olivine due to partial melting substantially reduces mantle conductivity. Thenorthern line data have been analyzed in the same manner as the southern line data. Theresulting conductivity structure wll be compared with the structure of the southern line. Thecomparison may reveal along axis variation of the melt generation and transportationsystem.

SS03/07A/A02-012 1215THE SIMULATOR EARTH'S CRUST MAKE BY BASE OF THEORY COVER

Stanislav G. DOLGIKH (Dep. Atmosphere and Ocean)

Today, be many simulator of Earth, but all are not describe processes, which run at crust,mantle and surface marking. We developed simulator earth’s crust base by theory cover.This is multi-ply spherical shell with real parameters of earth’s crust/ Shell laded outside andinternal forces. The model is flexible, that is why we may take into many processes whichrun on Earth (variation of pressure, natural oscillations of reservoirs? Wave processes ofocean and so on). When we will take into a count all this processes, we can count a stressof shell. We will get exertions, deformations, transference, so whale shell, and so this isvalues at a specific dot. In the upshot, we receive physical consistent dynamic model multi-ply shell. If we will take into distributed load (atmospheric and mantle pressure), we will canget significance elastic parameters earth’s crust and significance fundamental frequencies