Arc Deformation and Marginal Basin Opening: Japan Sea As

Arc Deformation and Marginal Basin Opening: Japan Sea as a Case Study Laurent Jolivet, Philippe Huchon, Xavier Le Pichon, Nicolas Chamot-Rooke, Jean-Charles Thomas, Jean-Pierre Brun

To cite this version:

Laurent Jolivet, Philippe Huchon, Xavier Le Pichon, Nicolas Chamot-Rooke, Jean-Charles Thomas, et al.. Arc Deformation and Marginal Basin Opening: Japan Sea as a Case Study. Journal of Geophysical Research : Solid Earth, American Geophysical Union, 1991, 96 (B3), pp.4367-4384. 10.1029/90JB02455. insu-00726512

HAL Id: insu-00726512 https://hal-insu.archives-ouvertes.fr/insu-00726512 Submitted on 30 Aug 2012

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 96, NO. B3, PAGES 4367-4384, MARCH 10, 1991

Arc Deformationand Marginal Basin Opening' JapanSea as a CaseStudy

LAURENT JOL1VET AND PHlI2PPE HUCHON

Ddpartementde Gd,ologie, Ecole normale supdrieure,

JEAN PIERRE BRUN

Laboratoire de Tectonique,Universitd, de RennesI,

XAVIER LE PICHON, NICOLAS CHAMOT-ROOKE AND JEAN CHARLES THOMAS

Ddpartementde Gdologie,Ecole normalesupdrieure,

We discussthe openingmechanism of the JapanSea in Miocenetime using (1) tectonicand published palcomagneticdata along the eastemmargin from the north of Hokkaido Island to Sado Island, (2) a mechanical model which is tested by small-scale physical modeling, and (3) crustal structureand bathymetric features in the Japan Sea which constrain our kinematic model and preopening reconstructions.Our main conclusionsare the following. The easternmargin of the JapanSea was, as a whole, a dextralshear zone about 100 km wide. This conclusionis supportedby the existenceof a ductile dextral shear zone in Central Hokkaido (Hidaka Mountains) and associatedbrittle deformation in western Hokkaidoand northeastemHonshu. The stressfield duringthe opening(which endedabout 12 Ma ago at the end of the middle Miocene) changesfrom right-lateral transpressionin the north to right-lateral transtensionin the south.The westernmargin, along the Korean peninsula,during the sameperiod, also was an active dextral shear zone. Palcomagneticresults indicate that clockwise rotations occurredin the southduring the openingand counterclockwiserotations in the north. We proposea model of right-lateral pull-apart deformationwith dockwise rotationsof rigid blocks in the southerntranstensional domain and counterclockwiserotations in the transpressionalone. Small-scale physical models show that the clockwiserotation in transtensionis possibleprovided that the easternboundary (Pacific side) is free of stress.The openingstopped and compressionsubsequently began about 12 Ma ago. Finally, we showthat the dextral shear,which is distributedover the whole JapanSea area, is accommodatedby N-S trending fight4ateral faults and rotation of blocks locatedbetween these right-lateral faults.

INTRODUCTION we propose for the opening is tested with small-scale experiments.Finally, this model is quantified in terms of Whether marginal basin opening occurs by slabretreat kinematicreconstructions. We do not discuss forces which led [Matsuda and Uyeda, 1971; Hilde et al., 1977; Chase, 1978; to the opening.We addressthe questionof how the JapanSea Uyeda and Kanamori, 1979; Taylor and Karner, 1983; Uyeda, openedand leave asidethe questionof why it opened. 1986] or by extension induced along transform faults [Hamburger and Isacks, 1988] or by intracontinental deformation [Tapponnier et al., 1982; 1986; Kimura and BACKGROUND Tamaki, 1986], it is assumedthat the adjacent arc does not behave as a rigid body during the openingbut, instead,suffers Two principalmethods have been used to study deformation extensive deformation [Otofuji et al., 1985, Tamaki, 1985]. at the scaleof the Japanarc. The first methodcharacterizes the We proposein this paper a model for the Miocene openingof deformationwhere it is localized,i.e., along the major faults, theJapan Sea (Figures 1and 2) which isbased on our study of anddescribes thestress and strain regime. This method does the deformationover the wholearc andpublished not usually give an estimateof themagnitude of the palcomagneticdata.Earlier Otofuji etal. [1985] proposed that displacementsalongthe faults because clear offset markers are the Japan Sea openedby rotation without significant north- usuallymissing on both sidesof the fault zone (it is indeed the soum translation of SOBLIIWC•...... tgIll •11L• 11UI casefor the Japanarc). The secondmethod uses palcomagnetic based on paleomagnetic data (see Figure 2, model 1). We measurements within the relatively less strained blocks bounded by the major faults to obtain the amount of rotation opposethis "bar door" opening model for a right-lateral pull- apartmodel which involvesrotations about a nearby pole. The and of latitudinalmotion. These two approachesare obviously complementary. geometryof a typical pull-apart supposesno "spinning"of the displaced blocks [see McKenzie, 1990]. In our model a Based on a palcomagneticstudy, Kawai et a/.[1971] first concluded that Japan had been bent. All the following componentof rotation about a nearbypole is addedto the N-S translationalong the major strike-slip faults. The mechanism paleomagnetic workshave confirmed thisearly conclusion althoughsignificant discrepancies exist. For example,Otofuji et al. [1985] concludedthat SW Japanhas rotated clockwise Copyright1991 by the AmericanGeophysical Union between15 and 13 Ma by 56ø about a pole locatedin the Tsushimastrait (between Kyushu and the Koreapeninsula), and Papernumber 90JB02455. that NE Japan (the northern part of Japan, north of the .....n 148-n'•'•'•/t•, /•,,/90JB02455505.00 Tanakura Tectonic line) has rotated rigidlv...... by 47 ø

4367 4368 JOIXVKF ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

Sea was describedas a major right-lateralshear zone for two reasons. First, the offshore structure shows a successionof N-S trendingridges and troughsbounded by vertical faults in the Tartary Strait, and right-lateralen 6chelontroughs along the westerncoast of Hokkaido and northernHonshu [Antipov et al., 1980; CCOP-IOC, 1980]. Second, a right-lateralshear zone, on land, the Hidaka ShearZone, was activeduring the opening.A similarargument was used for the westernmargin alongthe easterncoast of Korea [Kang, 1981;Reedman and Um, 1975;Schluter and Chun, 1974]. To testthis proposal we analyzedthe brittle deformationalong the easternmargin from Hokkaido and Rebun islands in the Tartary Strait to SW Hokkaido [Jolivet and Huchon, 1989], Oga Peninsula, Nihonkokuregion and Sadoisland (the data collectedon Sado Islandare herepresented for the fLrsttime) usingAngelier's [1984] methods.We alsoconsidered structural data concerning the bending of Mesozoic belts in Kyushu [Murata, 1987; Fabbri eta/., 1987]. These data, togetherwith thosepreviously discussed show that (1) the easternmargin of the JapanSea is a right-lateral shearzone which was activefrom the Oligocene(?) to Middle Miocene, (2) the stressfield evolvedin spacefrom a NE-SW horizontalcompression in Hokkaidowith simultaneousright- lateral and reversedisplacements (right-lateral transpression) to a NW-SE extensionin the southwith coexistenceof right-

'130• 150"E lateral and extensional displacements (right-lateral transtension)in agreementwith the initial pull-apartmodel, and (3) this deformationis contemporaneouswith the bending Fig. 1. Simplified structuralmap of the JapanSea and Okhotsk of SW Japanin the Middle Miocene. These results,considered Sea area. Key shows (1) oceaniccrust, (2) thinned continental togetherwith the palcomagneticdata, lead to a pull-apart crust, (3) outcrops of basement in the Okhotsk Sea. TF is model with mostly clockwise rotation to the south and counterclockwise rotations to the north. Tartary fault, HSZ is Hidaka ShearZone, KLZ is Kashevarov Linear Zone, YF is Yangsanfault, TF is Tsushimafault, YBk is Small-scalephysical experiments were undertakento see (1) Yamato Bank, YBs is Yamato Basin, OM is Oki-Monbetsu under which conditionsrotations are allowed and (2) what Tectonic Line, and EKF is East Korean Fault. senseof rotationcan be expectedin a right-lateralshear zone with a pull-apartbasin opening. We testedthe influence on the openingmechanism of the "free boundary"on the Pacific side: as suggestedby Tapponnier and Molnar [1976] and counterclockwiseabout a pole located north of Hokkaido Peltzer and Tapponnier [1988], the horizontalstress induced duringthe openingof the JapanSea (Figure2). Accordingto by subductionof the Pacificplate eastof Asia is weak and the Otofuji et a/.[1985], these large rotations were completed subduction does not resist to the eastward extrusion of within less than 1 Ma about 15 Ma ago. Faure and Lalevde continental blocks. We show that rotations are always [1987] usedcorrelations of preopeningstructures to proposea clockwise, e.g., right-lateral. Rotations of small dominolike similar model. blocks between major faults and rotation of large blocks On the other hand, on the basis of observations of the leading to basins opening are both observed. All these deformationin northern Japan and compilationof available manifestationsare enhancedby the presence of a free structuraldata, Lallemand and Jolivet [1985] and Jolivet et al. boundary. We conclude that the clockwise rotation of SW [1989] pointed out that major right-lateral faults were active Japanabout a polelocated in theTsushima strait is compatible during the opening along the eastern and westernmargins in terms of mechanicswith the pull-apart system.We also (Hidaka mountainsin Hokkaido and YangsanFault in South concludethat the counterclockwiserotation of northeastJapan Korea). These authors proposeda right-lateral, pull-apart is the result of the transpressionalcontext there. We then openingmodel (Figure 2). Earlier, Otsukiand Ehiro [1978] had propose a kinematic model which takes into account the used a similar approach,but with less data available at that geometryof the basinsand major faults. time they reacheda differentconclusion. There is an obviousconflict betweenthe pull-apartmodel, SUMMARY OF TttE BASIN STRUL-q75RE which involvesN-S translations,and the currentinterpretation of palcomagneticdata, which implies only rotations about The structureof the JapanSea is describedin detail by nearby poles. In order to resolve this conflict, we presenta Tamaki [1988]. Accordingto Ludwig et al. [1975], Honza threefold approach:(1) We establishbetter constraintson the [1979] and Tamaki [1985], oceaniccrust is distributedin three deformationof the arc and its timing by characterizationof the deepbasins (Figure 2). The largestand deepestis the Japan stressfield and studyof the behaviorof the majorfaults during basinwhich has the shapeof a parallelogram.It is located the opening.This allowsus to proposea generalmodel; (2) We betweenthe Eurasian margin to thenorth, the Korean margin test the mechanical:ealidity of this model using small-scale to the west, the Yamato Bank to the south, and the western analog experiments; (3) We propose preopening margin of Honshu and Hokkaido to the east. Between the reconstructionsof the Japan Sea based on this general Yamatobank and the islandof Honshuis the 2000-m-deep mechanismwith constraintsfrom the geometryof bathymetric Yamato Basin. West of the Oki ridge is the Tsushimabasin. features. Present-daytectonics involves a nascentunderthrusting of the In the initial paper proposing the pull-apart model JapanSea crust below the NE Japanarc [Fukaoand Furumoto, [Lallemandand Jolivet,1985], the easternmargin of the Japan 1975; Nakamura, 1983]. JOLIVETET AL.: ARCDEFORMATION AND MARGINALBASIN OPENING 4369

JAPAN

Oga Pen.

3 4 5 ,,, • • 9 10 11

Fig. 2. Tectonicmap of theJapan Sea (insert: three different models of opening,(l) Otofujiet al. [1985],(2) Jolivetet al. [1989], (3)Lallemand and Jolivet [1985]. Key shows(1) continentalblocks in the JapanSea afterTamak/ [1986], (2) thinnedcontinental crust, (3) oceaniccrust, (4) recentinfill of en 6chelongrabens on theeastern margin of the JapanSea, (5) onlandthrusts, (6) strikeof foldsaxes after Kimura et al. [1983],(7) activeoffshore thrusts after Tamaki [1988], (8) strike-slipfaults, (9) normalfaults, (10) subductionof the Pacificplate, (11) subductionof thePhilippine Sea plate. BTL is ButsuzoTectonic Line, HMT is HidakaMain Thrust,HTSZ is HidakaTartary Shear Zone [Lallemandand Jolivet,1985], IBR is Izu-BoninRidge, ISTL is Itoigawa-ShizuokaTectonic Line, KPR is Kyushu-PalauRidge, KTL is KamishiyubetsuTectonic Line [Kimura et al., 1982], kyb is Kita YamatoBank, MTL is MedianTectonic Line, nkr is North KoreanRidge, ntb is Nishi-TsugaruBasin, ob, is Oki Bank, ob is OkushiriBasin, or is Oki Ridge,ot is Oki Through,TF is TsushimaFault, 'I'TL is TanakuraTectonic Line, skr is South Korean Ridge, st is ShiribeshiTrough, yb is Yamatobank, Y-Fis YangsanFault, andZR is ZenisuRidge. 4370 JOL1VET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

The magnetic structureof the basin is poorly constrained. rigid rotation 15 Ma ago by 56ø about a pole located in lsezaki and Uyeda [1973], Kobayashiand lsezaki [1976], and Tsushimastrait. The directionof paleomagneticvectors are lsezaki [1986] publishedcontoured magnetic anomaly maps of indeedfairly consistentin the areasurveyed by Otofujiet al. the JapanSea. The E-W to ENE-WSW strike of the anomalies [1985],who are in favorof a rigid-bodyrotation. However the appearsclearly, at least in the Japanbasin, but it is impossible declinationprogressively becomes more northerlyto the east to identify the anomalieswith confidence.Tamaki [1988] also and northeast[ltoh, 1986; 1988; Itoh and lto, 1988]. Hirooka recognizedE-W trending anomalieswith a possibleeast to et al. [1986] proposedthat independentrotations of smaller west propagationof the spreadingcenter along strike, during a blocks may have occurred.Itoh [1988] concludedthat SW recentsurvey of the northernpart of the JapanSea. Japanhad first rotatedclockwise, as claimedby Otofujiet al., The directionsof the anomaliesdo not fit the fan shape and that the easternpart of it later rotatedcounterclockwise to openingof the northernJapan basin proposed by Otofuji et al. someextent due to bendingof the arc along the Itoigawa- [1985] basedon paleomagneticdata, whichpredicts fan-shaped Shizuoka Tectonic Line. Thus the magnitudesof these anomalies.Instead, the spreadingcenter seems to abut against rotationscannot be ascertainedas we only know the resultant the westernmargin of the Japanarc. We know that the amount rotation. of underthrustingalong this margin is small [Nakamura, 1983; Jolivet et a/.[1989] proposedthat these results could be Tamaki and Honza, 1984], which implies that the anomalies interpretedas due to a rigid rotationof SW Japanwith local were originaly formed perpendicular to the margin. This rotationsdue to Mioceneinternal deformation of the arc along implies a component of strike-slip or transform along the theMedian Tectonic Line whichis a Cretaceousand Paleogene margin. left-lateralfault [lchikawa, 1980;Otsuki and Ehiro, 1978]. But The Yamato Bank, a rifted continentalridge, separatesthe theEarly to MiddleMiocene deformation along this fault is not Yamato and JapanBasins. The Yamato Basin is thoughtto be well constrained and further observations are needed to slightly younger than the JapanBasin [Hilde and Wageman, ascertainthis conclusion.We shall see in the followingthat 1975]. Bathymetryand seismicprofiles show that the centerof 30ø of clockwiserotation of SW Japanis requiredto closethe the ridge is occupiedby a graben trending N50øE. Several basins fitting crustal features. This clockwise rotation is parallel grabens appear in the morphologyof the northern compatiblewith structuraldata of KyushuIsland (see below). margin of SW Japanand easternmargin of Korea. Along the More recently, McCabe et al. [1988] showedthat Early easternmargin of the Yamato and Japanbasins, grabens are Miocenesites of SouthKorea have been rotated clockwise by also observedtrending more northerly (N30øE). Several en 20 ø. The latter data either show local rotations east of the •chelon troughs (Figure 2: Shiribeshi trough, Okushiri and right-lateralYangsan fault (see later) or reducethe amountof Nishi TsugaruBasins) and the activeE-W compression[Fukao rotationof SW Japanwith respectto the Koreapeninsula. and Furumoto,1975] havebeen related by Nakamura [1983] to the nascentsubduction. They actually are older extensional NE Japan structuresreworked by the recent compression[Tamaki and Honza, 1984; Lallemand et al., 1985]. One of thesegrabens Otofuji et al. [1985] proposedthat northeastJapan rotated cropsout in the islandof Sado(Figure 3). Here the faultingcuts counterclockwise by 47 ø about a pole located north of Oligocene to Middle Miocene volcanics and volcanoclastics Hokkaido 15 Ma ago. Their data were obtained along the [Ganzawa, 1982] and is postdatedby Late Miocene marine easterncoast of the Japan Sea in Oga peninsulaand Asahi deposits. mountains(Nihonkoku in figure 3). Western Asahi shows an A similar tectonictiming is recordedalong the whole eastern easterly declination while eastern Asahi shows a westerly andsouthern margins of the JapanSea. For example,along the declination.They concludedthat the easterlydeclination is due easterncoast of Korea, Schluter and Chun [1974] have shown to local right-lateral rotations along the Nihonkoku-Miomote offshoreN-S trendingfaults parallel to the margin. Onland, shearzone while the westerlydeclination is due to rigid body NE-trendingen 6.chelonnormal faults cut the Early Miocene rotationof NE Japan.Their conclusionsare mainly basedon depositsof the Yangbuggroup, and are coveredwith the Yeonil the data obtained in the Oga peninsula where a westerly depositsof Middle Miocene age [Lee and Pouclet, 1988;Lee, declination of rocks older than 14 Ma is observed. When 1989]. The extensional deformation thus occurred sometime plottingall the paleomagneticdirections of the Early Miocene duringEarly andMiddle Miocene.This timingis in agreement sitespublished by various aukors in NE Japan[figure 4 of with the conclusionsof Tamaki [1986] which are basedon heat Jolivetet al., 1989], thesedirections are not homogeneousand flow, stratigraphy,and basementdepth: the JapanSea opened many sites are not rotated. Tosha [1983] and Tosha and duringLate Oligoceneto Early andMiddle Miocene. Harnano [1988], after extensive sampling of the Oga Tarnaki [1985] produceda map of the thicknessof the crust peninsula,concluded that a fast rotation of about 20 ø occurred in the JapanSea. Oceanicand continentalcrust and strongly about 15 Ma ago. Figure 8 of their paper shows that the attenuatedcontinental crust are shownfrom his datain figure2. westerly declination decreaseswhen the age of the sample This is a usefulguide for the preopeningreconstructions. increases.If the data are reliable this is incompatiblewith a rotation achievedin a single discretestage. One must assume REVIEW OF ONLAND PALEOMAONET•CDATA an early clockwise and a later counterclockwiserotation as suggestedby Celaya and McCabe [1987] and Tosha and SW Japan. Harnano [1988] (or, alternatively, a progressiveclockwise rotationby 300ø). Thus the exact amountof rotationis poorly We shall not considersites older than Early Tertiary in this known. In this paper we relate the counterclockwiserotation paper. Still there is considerable argument concerning the to the transpressionalstress field in Hokkaido and show that a interpretationof Early Tertiary paleomagneticdata. total rotation of 20ø is compatiblewith the geometry of the Moreau et a/.[1987] concludedthat a strongremagnetization opening. event may have occurred in Miocene time, but Tosha and Hamano [1988] arguedagainst this conclusion.A widespread DEFORMATION OF THE ARC DURING THE OPENING sampling of Miocene tuffs show easterly declinationsin SW PROCESS,FmLD APmOACI-• Japan[Otofuji and Matsuda, 1983; Otofuji et al., 1983; Otofuji and Matsuda, 1984; Sasajima, 1981; Otofuji et al., 1985; In the following we summarizethe smacturaldata gathered Hayashida, 1986]. Otofufi et al. [1985] proposeda clockwise along the easternmargin of the Japan Sea (for detail on the 142 ø 140 ø

Fig. 3. Bathymetricmap of the easternmargin of the JapanSea after [Maritime Safety Agency of Japan, 1980a,b) andthe main onland tertiary structures [after Jolivet and Huchon, 1989; Yamaft, 1989]. IN is Iwanai nappe.Open arrows: Cretaceous motion along strike-slip faults; solid arrows'Cenozoic motion. KTL is KamishiyubetsuTectonic Line, OB is OkushiriBasin, NTB is NishiTsugaru Trough, ST is ShiribeshiTrough. Dashedcurves show direction of C•Hmaxduring O!igocene to MiddleMiocene; dashed solid curves show fold axes. 4372 JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENrING deformation in the Hokkaido Central Belt see Jolivet and Huchon [1989 and referencestherein]), we review published data about other areas, and we conclude with a deformation mechanism at the scale of the arc.

Hold•ido Central Belt

The present-dayHokkaido Central Belt (HCB, Figures3 and 4) developedduring the Cenozoicin two main stages.A major, right-lateral,N-S trendingshear zone (Hidaka Mountains)was formed during Oligoceneto Middle Miocene. It was followed by an E-W shorteningwhich led to the formationof westward verging thrustsfrom Late Miocene to Present[Kimura et al., 1982; 1983; Jolivet and Miyashita, 1985; Kimura, 1986; Moriya, 1986; Watanabe, 1988; Jolivet and Huchon, 1989]. This Cenozoicsuccession of deformationpostdates a Mesozoic stacking of nappes in high-pressure, low-temperature metamorphicconditions which is outsideof the scopeof this paper [lshizukaet al., 1983; Jolivet, 1986b]. The formationof the right-lateral Hidaka Shear Zone was coeval with the eraplacementof the Iwanai nappe[Jolivet and Huchon,1989] toward the south and with the formation of en 6chelon folds and thrusts.These latter features can be followed all along the Hokkaido Central Belt. The analysisof the fault sets also revealsthe two successive stagesof deformationand allowsus to datethem. The younger sets cuts all the strata including the Pliocene [see Mitani, 1978; Yamagishiand Watanabe,1986]. The deducedprincipal compressirestress o 1, following Angelier's [1984] method, strikesE-W. Thesefaults are related to the recentE-W thrusting event. The older fault set is found in all strata from Middle Miocene downward.The principalcompressive stress strikes NE-SW which is compatiblewith the right-lateralshear along the Hokkaido Central Belt. The Oligoceneto Middle Miocene age of the right-lateral shear is also confirmed by the radiometric data obtained by Shibata [1968], Shibata and lshihara [1979] and Shibata et al. [1984] on synkinematic granites and migmatites in the Hidaka Mountains (40 to 17 Ma). The Hidaka Shear Zone is made of high-temperature metamorphic rocks [Komatsu et al., 1983; Osanai et al., 1986a; b] with a vertical N-S trending foliation. The synmetamorphicstrike-slip deformation reworked the original Mesozoicthrust contact between an ophiolitesequence and a granulitic basement[Jolivet, 1986a]. A high PT gradientis observed increasing from west to east [Nakano, 1981; lshizuka, 1981; Miyashita, 1983]. Throughoutthe zone, the foliation is vertical, exceptin narrow en •k:helonzones where it flattens toward the west. A conspicuous horizontal stretching lineation, with consistent evidence of synmetamorphicright-lateral shear, is observedeverywhere. Evidence of Miocene right-lateral ductile shear was also reportedby Watanabeand lwata [1985]and Watanabe [ 1988]in the northernHidaka Belt and also by Uda [1973; 1976] in the southernpart of the belt. Jolivet and Huchon [1989] concluded that the entire Fig. 4. Directionsof the main horizontalstresses deduced from Hokkaido Central Belt was a crustal-scale, transpressive, flower structure(a positive flower structurein the senseof fault sets analysis in Hokkaido and Sado. 1, strike of folds Harding 1985] with a narrow(10 km), deep,ductile, strike-slip axes. Key shows (2) main thrusts,(3) direction of maximum shear zone and a wide, shallower domain of en 6chelon thrusts compression[after Yamagishiand Watanabe, 1986] in the andfolds during the Oligoceneto Middle Miocene. Oshimapeninsula, (4) directionof minimumcompression after the same authors,(5) our own data, large arrowsfor the main horizontal compression, narrow ones for the other main East Coastof the JapanSea horizontalstress, compressional in (a) Hokkaido, extensional Oligoceneto Middle Miocenetuffs and lavas(Green Tuffs, in (b) Sado, Large numbers(1 and 2) refer to the successive Ingle [1975]; Ganzawa [1982; 1983; 1987]; lijima et al., stages recorded in Sado when these have been clearly [1988]) are widely exposedalong the easternmargin of the differentiated.The open arrow along the KTL representsthe JapanSea andrecord a deformationcontemporaneous with the right-lateral motion active after the end of Middle Miocene. opening.From RebunIsland to SadoIsland (Figures 3 and 4), ATL is Abashiri Tectonic Line, KTL is Kamishiyubetsu the Oligocene to Middle Miocene horizontal maximum Tectonic Line, HMT is Hidaka Main Thrust. JOLIVETET AL.: ARC DEFORMATIONAND MARGINAL BASIN OPENING 4373

compressivestress 03Hmax), deduced from the analysisof This is in good agreementwith the resultsof Otsuki [1989] in faults sets, trends NE-SW. However, this stresscorresponds to NE Honshu,based on the directionsof dykesand metalliferous the maximum compressivestress C•l in the north where the veins. stressregime is transpressive,while it correspondsto the The secondobservation is that this deformationcorresponds intermediatestress c•2 in the south,where the stressfield is to right-lateral transpressionto the north coeval with the transtensive.In the north, thrustsfaults and strike-slip faults deformationof the Hidaka Shear Zone, and to right-lateral are associatedwith folds, whereas in the south normal faults transtensionto the south. This spatial evolution is also and strike-slipfaults coexist. There is a clear transitionfrom observedin time on Sado Island although more precise right-lateraltranspression to right-lateraltranstension from stratigraphicdata would be neededto confirmthis theory.The north to south. third observation is that the Cretaceous left-lateral shear zones Figure4b showsa summaryof the datafrom SadoIsland. such as the Nihonkoku-Miomote and Tanakura tectonic line are This island is divided in two by a graben.This is one of many reactivatedas right-lateralfaults during the Tertiary.Thefourth grabenswhich are now below sea level along the eastern observationis that this right-lateral shear ended before the marginof the JapanSea. The tectonicdata on thisextensional Late Miocene.All thesecharacteristics are compatiblewith the deformationshed light on the evolutionof the entiremargin. environmentof the marginof a right-lateralpull-apart basin The deformationshown in Figure 4b doesnot affect the Late settledonto a pre-faultedcontinental basement. Miocenedeposits which postdate the formationof the graben. The northernextension of this systemcan be followed in The outcropsshow the associationof strike-slipand normal Sakhalin(Figure 1; Rozhdestvenskiy,1982; 1986; Kirnura et faults and the principal horizontal compressive stress al., 1983] as a well developeddextral strike-slip system along consistentlystrikes NE-SW. Althoughthe stratigraphicdata the Okhotsk-Amurplate boundary.Deep narrowbasins of the are not preciseenough to concludetwo discretestages, we Tartary strait west of Sakhalin [Antipov et al., 1980] have observedat many outcropsthe successionfrom a strike-slip probably been formed in this strike-slip environment.From movement to a normal movement with a strike-slip Hokkaidoand northwardthe right-lateralshear zone is located component.In some cases, the earlier strike-slip faults betweentwo continentalblocks. The deformationalong this crosscutfiat-lying reverse faults compatiblewith the same intracontinental boundary is transpressional. South of directionof C•Hmax.It seemsthat the deformationmechanism Hokkaidothe strike-slipshear zone is in a backarcposition. was first transpressive then transtensive with a larger Duringthe Early andMiddle Miocene the Japanarc faced only componentof extensionwith time. the Pacific plate, the PhilippineSea Plate being far to the In the Oga peninsulathe directionof C•Hmaxwas determined south [Jolivet et al., 1989]. This transition from an from a dykeswarm. We observedin severalplaces right-lateral intracontinentalcompressional context to a back-arcsetting en 6chelon relationshipsbetween dykes as well as reverse explains the transition from transpression to transtension faults giving the directionof C•l along the N30ø to 50øE from N to $. direction. These dykes were thus not emplaced in an extensionalstress field but in a right-lateraltranspressive one. Further to the south, the Cenozoic deformation is Review of the TectonicData on the Western concentratedalong the CretaceousNihonkoku-Miomote shear Margin of the Japan Sea zone describedby Otsukiand Ehiro [1978]. Like other left- lateral Cretaceous shear zones (Tanakura tectonic line) it is Sillitoe [1977] quantified the amount of right-lateral reactivated as right-lateral during the Cenozoic [Koshiya, displacementalong the Yangsan-Tsushimafaults. These two 1986]. Along the coast, the basementof the Cenozoicdeposits faultsrun NNE-SSW alongthe easterncoast of Korea and in the (Cretaceous granites) is cut by fiat-lying normal faults Tsushima strait. S illitoe noticed that one can correlate compatiblewith the sametranstensional stress field as in Sado metallogenicbelts in the Korean peninsulaand SW Japanif Island. one considersa right-lateraloffset of 200 km (Figure 2). The displacementwas achievedduring Oligo-Miocene time after the Analysisof the BathymetricFeatures along formation of the youngestbelt at 42 Ma and before the recent the EasternMargin of the Japan Sea left-lateral motion along the Tsushima fault. This recent motion gives rise to left-lateral en 6chelon compressive Tamaki and Honza [1985], Okada et al. [1985], and structures affecting Tertiary sediments in Tsushima island. Lallemand et a/.[1985] showed that the Shiribeshi trough, Counterclockwiselocal rotations shown by palcomagnetism Okushiri basin, and Nishi-Tsugarubasin (Figures 3 and 5), [lshikawa and Torii, 1988] occurredafter the intrusionof a 14 presently active as compressionalstructures, are reactivated Ma pluton. extensional structures. Their en 6chelon relations, and the The right-lateral motion is also seen in South Korea by en tectonichistory of Sado Island favor a right-lateralshear zone 6chelonNE-SW trendingnormal faultsin the Miocene deposits along the easternmargin from Oligoceneto Middle Miocene. of the Pohangbasin. Basaltic lavas were extrudedduring Early A simple analysis of the bathymetric chart of the eastern Miocene along these NE trending tension gashes [Lee and margin [Maritime Safety Agency of Japan, 1980a;b] which is Pouclet, 1988]. These en 6chelon normal faults are located east reproducedin Figure 3 reveals that two directions of scarps to the NNE-SSW trendingYangsan fault. We can also interpret prevail: (1) the directionof the grabensand (2) the directionof this simple geometryas the result of a right-lateralshear along linear structurestrending NNW-SSE to N-S. Figure 5 gives an the Yangsanfault. interpretation of these structures. The NNW-SSE trending The age of the fight-lateralmotion is thus bracketedbetween structurescorrespond to right-lateraltransfer faults betweenthe 42 and 14 Ma. It may have stoppedslightly earlier than along grabens. We shall consider later the implication of the the easternmargin of the Japan Sea. The clockwise rotations progressive change in strike from north to south for the transfer faults. shownby McCabe et al. [1988] in Early Miocene depositseast of the Yangsanfault in South Korea also favor a right-lateral A Right-Lateral Shear Zone motion along this fault during Miocene time. Several workers [Murata, 1987; Fabbri et al., 1987] The first observationis the consistencyin the strike of attemptedto relate the obvious bending of the pre-Cenozoic C•Hmaxfrom northto south(N30øE as an average)based on the belts of SW Japan in Kyushu with the clockwise rotation analysisof fault setsas we have shownin a previoussection. deducedfrom palcomagneticmeasurements. They showedthat 4374 JOLIVE-'r ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

/ / / \\ / ß / . / . ' /

/.// / .6

Fig. 5. Tectonicinterpretation of Figure3. Dashedcurves, direction of Oilmax; small arrows;motion on individualfaults; large arrows,resultant motion on the major shearzone. JOLIVETET AL.: ARC DEFORMATIONAND MARGINAL BASIN OPENING 4375 the bendingalso affects the Eocenestructures. They described ExperimentalProcedure conicalfolds with steeplydipping axes, on the concaveside of the bending (southeasternKyushu), and Miocene grabens We used the small-scalemodeling techniqueof a brittle- ductile systemdescribed by Faug•re and Brun [1984], and radially disposedon the convexside (northernKyushu on the Vendeville et al. [1987]. The two-layer models are made of JapanSea side,Figure 2). This supportsthe idea of a Miocene Fontainebleausand representingthe upper brittle crust and age for the bending. silicone (Rhodorsil Rhone Poulenc) representingthe lower Conclusion ductilecrust. The layersare addedto a rectangularcontainer, the baseof whichis madeof a fixedplate and a mobileone (Figure The offshoreand onlandtectonic data along the easternand 7). The mobileplate, which is displacedat a constantvelocity, westernmargins of the JapanSea are compatiblewith a right- has a "dog leg" shapedboundary, which appliespure strike- lateral pull-apart opening(Figure 6) during Oligoeene-Middle slip and transtensional displacementsto the base of the Miocene time, followed in Late Miocene by the E-W silicone. compression,still active at presenttime. The opening was also accompaniedby a rotation of SW velocity discontinuity Japanpossibly about a pole locatedin the Tsushimastrait and / the consequentbending of the arc. Rotations about vertical axes associatedwith strike-slip faults are already well known basal [e.g. Ron eta/., 1984; Walcott, 1984; 1987; Hornafius et al., strike -slip .... mobile 1986; Jackson, 1987; Lamb, 1988]. boundary plate Is then the right-lateral pull-apart opening mechanically compatible with the clockwise rotation of SW Japan?The following is an attempt to answerthis questionusing small- scale experiments. transtensional SMALL-SCAI• ANALOGEXPERIMENTS We haveperformed the small-scale experiments presented in the following section to examine the kinematic and '•'--. freeboundaryboundary mechanicalcompatibility of clockwiseblock rotationswith ..,of the sand layer right-lateralpull-apart opening and to determinewhich type of -•- 50cm •.....- boundaryconditions favor this development. sand

Fig. 7. Schematicdescription of the small-scaleexperiments (1 showsvelocity discontinuity).

Models are scaledto fit the shearstrength profile of the continentalcrust [Goetze and Evans, 1979]. The sandyields by a Mohr-Coulombcriterion with an internalfrictional angle of 30ø, which may satisfactorilymodel the frictionalbehavior of the upper crust [Byerle,, 1978]. The siliconeputty has a viscosityof 104Pa.s at roomtemperature. Its shearstrength dependson displacementvelocities. In the experiments presentedhere, sandand siliconelayers have a thicknessof 1 cm. For sucha model an appropriateshear strength for the siliconelayer is obtainedfor velocitiesvarying between 0.5 and10 cmfn(i.e., weakand strong viscous coupling). Before 12 Ma (i.e. before the beginningof the E-W compression), the Pacific subduction could be considered a stress-freeboundary (totally free of stressor partlyextensional if slab-pullis active)during the JapanSea opening. To testthe consequencesof this lateral boundarycondition, most of the models were constructed such that the sand did not reach the right wall of the container.Free spacebetween the model and the lateralwall was filled with silicone.The shapeof the free boundaryis built identicalto thatof thebasal mobile plate to maintain a constantdistance between them. During the buildingof the sandlayer, sandloading creates a flow of the materialtoward the free space.Even a smallamplitude flow nearthe free boundary is sufficientto creategrabens (see Figure Fig. 6. Simplifiedopening model with right-lateralpull-apart 10), so that,strictly speaking, the easternboundary is not free and dextral rotations in the transtensional domain and sinistral of stressbut subjectedto extension. rotations in the transpressionalone. Large solid arrows: Twelve experimentswith variousdisplacement velocities motionon the major shearzones; divergent arrows: direction andwith or withouta stress-freelateral boundary have been of extension;curved arrows: component of block rotation. conducted.Only a few of themare reportedhere; a more detailed 4376 JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

exp n ø'! 1' d .-.•=2,7 em

fast subsidence.. :•i

0.5 cm/h ,; ., •:.:•=: .:40-.o•

Fig.8. Experiments9 (a) and11 (b) showthe fault pattern with a smalldisplacement andlow velocity (0.5 cm/h)and ot = 40ø. Displacement areshown by the deformation of the whim grid, originaly equally spaced (2.5 cmat thecenter of themodel where most of thedeformation occurs, 5 cm elsewhere). At highervelocities the faultpattern is similarat the beginningof the displacement.Note the R andR' Riedelshears and the two parallelgrabens in thetranstensional area which define a quicklysubsiding and rotating block. JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING 4377 reportwill be publishedlater. In the followingwe describethe Block Rotations general characteristics, similarities and differences of the experiments. In the central area, blocks bounded by grabensundergo a rigid dextral rotationduring the displacement(Figure 8a). As the displacementincreases, the distancebetween Riedel shears Fault Pattern connectedto the grabensat the terminationof the central area The fault patternwhich developsat the onsetof deformation increases.R' type Riedel shears develop creating a system of combines strike-slip faults above the basal strike-slip rotating dominos (Figure 8b). This set of experiments boundariesand obliquenormal faults above the transtensional demonstrates that block rotations are enhanced by the boundary.This is well shown by experiments9 and 11 in existenceof a free lateral boundary. Figure 8 with an angleof 40ø, displacementvelocities of 0.5 Figure 9 illustratesthe progressiverotation of the central cm/h, and bulk displacements of 3.8 and 2.7 cm. The block for displacementsup to 7 cm. The bulk pattern may be transtensionaldomain is characterizedby a large amplitude compared with the Japan Sea structure where roughly N-S depressionbounded by two parallel grabans.The set of strike- trendingstrike-slip faults are connectedwith NE-SW trending slip faults exhibits well-developeden 6chelon shear, slightly grabens.The central block can be comparedwith the Yamato obliqueto the main fault direction(R type Riedel shear).In the bank which separatesthe two main basins(Japan and Yamato proximity of the transtensionalarea the faults curve and Basins)and suffereda strongsubsidence and clockwise rotation antitheticR' type Riedel shearsdevelop. This defines domino- (seebelow) during the opening.We show in a later sectionthat type systems. This general pattern is similar to the one geometricalconsiderations require a clockwiserotation of the obtainedin Mandl's [1987] experiments.At the junction of the Yamato Bank during the opening. In any case, the rotation transtensionaland strike-slip domains the Riedel shears tend must have occurredbefore 14 Ma [Sayanagiet al., 1987]. The to connectwith the grabans. 30ø of dextral rotation of SW Japan also fits this model well. Note that the amount of block rotation in the transtensional The angle of grabarison the strike-slip boundary increases with o[ (seeFigure 9a whereo[ = 90ø).Note alsothat in thiscase domain increases with o•. Riedel shears pass progressively to grabens toward the Figure 10 illustrates the development of a domino system transtensional area. (also active in Figure 9, but less clear). From the beginningof exp,nø7d=2,9cm c•=90 ø V=3cm/h .'...... d 56cm

':? ::•"':.,7-:"::;t::•.:-.:" ..... '" .::.:":.::.7F:F?•.,:..•?:.::::•L..:.:•.;,5:.::/:•::• • ,• _

: .:•::-.--...... %.•:.•..... •:::•'"';•:'•'• •..- rotation •:•..::..::....

Fig. 9. (a) Experiment7 with o[= 90ø, andvelocity = 3 cm/h.Center of themodel after (b) 5.6 and(c) 7 cm of displacement.Note the dominosystem and the clockwiserotation of the centralblock. A simplifiedsketch of the structureof the JapanSea illustrates the comparisonwhich is madein the text. 4378 JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

'c•= 35 'o V :,=3 cm/h exp nø3 d = 2..7 cm.:

d.ominos with ri.ght--:l:at.e'[al rot. al. io:n

Fig. 10. Experiment3 with oc= 35ø andvelocity = 3 cm/h.It showsvery clearlythe developmentof the dominosystem. Note that the rotationof dominosis accomodatedby letft-lateralmotion along R' shearsand thatduring the rotationthe widthof the dominosystem increases. Note alsothe grabenswhich develop along the free boundary.Displacement (a) 2.7 cm, (b) 5.3 cm, (c) 7.5 cm is shown.

the deformation,R Riedelshears of the strike-slipdomain pass distancebetween the two boundaryfaults increasesduring the to obliquenormal faults in the transtensionaldomain (Figure rotation. This is allowed by the extensional component 10a). As the displacementincreases, they separateand R' left- introducedby the free boundary. lateral shearsdevelop between them (Figures 10a and 10b). Althoughthe comparisonof the modelswith the openingof Subsequentdeformation rotates the fault-boundedblocks like the Japan Sea (Figure 9, insert) is interesting, the 20 ø of dominos. Figures 10b and 10c show that block rotation sinistral rotation in northern Japan deduced from decreasesand disappearswhere R' shearsare not present.A paleomagnetic data and also from our reconstructions(see direct comparisonbeween these models and the Japanarea below) still has to be explained. One solution is that the suggeststhe occurrenceof dextral rotatingdomino systems reverse sense of rotation here is xelated to the dextral along the easternmargin of the JapanSea. The "en 6chelon" transpressivemotion which characterizedthe northernpart of grabenswhich exist in this regioncould be interpretedin this the basin. way. RECONSTRUCTIONS This deformationpattern with rotation of dominoshas been studied by Nut et al. [1986] and Mandl [1987]. Martell et The reconstructionsare made using constraintsfrom the a/.[1988] and Ron et al. [1984] describedfield examplesof structuraland paleomagnetic data and the asymmetricpull-apart such structures.Martell et al. [1988] explained that the mechanism (with a "free boundary") just discussed.The rotationsense of blocksbounded by crossfaults betweentwo rotations of the different blocks and the right-lateral sinistralboundary faults is determinedby the componentof displacementsare constrainedby the presentstructure of the movement across these boundary faults: if an extensional basin. component is present the sense of rotation is Figure 11 showsthe right-lateraltransfer faults betweenthe counterclockwise,if a componentof compressionis instead en 6chelongraberis along the easternmargin of the JapanSea. present the sense of rotation is clockwise. In Nur et al. [1986] The strikesof thesefaults progressivelychange from NW-SE the authorsconcluded to a rotation in the oppositesense but in the north to N-S in the south.Figures 11 and 12a show that predeterminedthe crossfaultsin a position which does not thesetrends fit small circlesabout a pole locatedat 35.8øN and correspondto what we observehere. In our experimentsthe 122øE which also fits the Tsushima fault. Thus the deformation JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING 4379

460 pole is rather remote. We finally end with three different rotation poles to move the various blocks. The more external the block the more remotethe pole. The rotationpole of the block located between the two master strike-slip faults (southwesternJapan) is closerthan than the pole of the outside block(northeastern Japan). The crustalstructure recognized by Tamak/ [1986] (Figure 2] in the JapanSea constrainsthe total amountof rotationwe can use. Figures12b and 12c show two reconstructions.Figure 12b correspondsto the hypothesisof a single stageopening. The analogexperiments show that the 440 simultaneousopening of two depressions(Japan basin and Yamato Basin) with a rotating unstrained block between (Yamato bank) is possible. Figure 12c is a possible preopeningreconstruction. The total clockwiserotation of SW Japanis 20ø. The total counterclockwiserotation of NE Japan is also 20 ø. The history of the openingcan be summarizedas follows (Figure 12d) from Late Oligoceneto Middle Miocene (further dataaxe needed to constrainmore precisely the timing):a right- 420 lateral shear was distributed over the whole Japan Arc and Sakhalinregion with a transtensionalcontext in the south and a transpressionalone in the north. This led to the formation of NE-SW pull-apartbasins in the JapanSea region. The two main grabensevolved and becameoceanic basins giving birth to the Yamato and JapanBasins. Within this wide dextral shearzone, tectonic blocks rotated about vertical axes during the basins opening. During this processall the earlier major Mesozoic faults were reactivated.The strike-slip motion stoppedabout 12 Ma ago and was replacedby an E-W shorteningwhich is 400 now localized along the easternmargin as an active zone of underthrusting.

CONCLUSION We model the JapanSea openingas a distributeddextral shear over a width of about 500 km from E to W. The deformationproceeds along two N-S trendingboundary shear zones with rotations of blocks between them. The finite 380 displacementof the easternmargin relative to the westernone is comparableto the overallsize of the basin(500 km). The deformationof the islandarc has beencomplex. The marginal basin opened in a dextral shear context with a componentof extensionalong the Pacific margin.Only the northern part of the system (Hokkaido Central Belt and Sakhalin)has been subjectedto a transpressivestress field leading to counterclockwise rotations there. The reconstructionsshown here are describedin terms of rigid 360 blocks rotations although a more complex model with 138 o 1400 142 o distributeddeformation is morelikely. Distributeddeformation across the 80-km-widearea along Fig. 11. The transferfaults recognized in Figure3 and5 (thick the New Zealand Alpine fault has been demonstratedby lines) fit small circles about a pole located at 35.8øN and 122øE. Walcott [1984;,1987], Jackson [1987], Lamb [1988] on the basisof a geodeticand paleomagnetic survey. Hamburger and Isacks [1988] have intepretedthe openingof the North Fiji basin as a megapull-apart along the Pacific-Australiaplate is distributed over the whole width of the dextral shear zone boundary.West of the JapanSea, the Bohai intracontinental with a single rotation pole. basinalso opened in a dextralshear zone [Nabeleket al., 1987; We reconstructthe positionof the Yamato Bank before the Chen and Nabelek,1988]. The openingof suchbasins can no opening taking into account the shape similarity between its longerbe consideredto be only causedby trenchretreat which northern margin and the correspondingarea of the Eurasian may provide the overall transtensionalstress field but not the maxgin.A fit of thesetwo featuresgives a pole locatedat 34øN strike-slip component.The continentalcollision of India and and 129øEwhich is very close to the pole of the rotationof SW Asia is the mostlikely candidateas a drivingphenomenon for Japanproposed by Otofuji and Matsuda [1983]. thesemajor N-S dextralfaults. Tapponnier et al .[1982, 1986] To accommodate the deformation observed along the emphasizedthe role played by left-lateralmotion along the Hokkaido Central Belt, one must consider a certain mount of Red River Fault in the SouthChina Sea openingin the Late shortening during the strike-slip movement. Given the Tertiary. Kimura and Tamaki [1986] first proposedthat the direction of the belt, one needsa pole that makes the Okhotsk right-lateral pull-apart opening of the Japan Sea is a block move along a roughly north-southdirection. Thus the consequenceof the India-Asia collision. Based on small-scale I 40 ø

©B

I I I 130 ø 140 ø

40 ø

130 ø 140 ø Fig. 12. Kinematicmodel andreeonsla'uctions of the JapanSea opening. (a) Presentstage and the smallcircles aboutthe polesused in the reconstructions(A: 34øN, 129øE,which is usedfor the rotationof SW Japanand the Yamam Bank inside the pull apartsystem, B: 35.8, 122øE,used for the southwarddrift due to the right-lateral pull-apart).(b) Reconstructionduring the openingassuming that the Japanbasin and Yamam Basinsopened almostsynchroneously. The arrowsrepresent the displacementvectors necessary to reach the presentday position,the dashedcurves stand for small circles.Key shows(1) oceaniccrust, (2) thinnedcontinental crust, (3) small circles,(4) main strike-slipfaults, (5) spreadingcenters. (c) Possibleconfiguration at the beginning of the opening accordingto our model. (d) Synthetic figure showing the north-south translation and progressivebending of the are. Dark shadingrepresents the present-daysituation, light shadingis the pre- openingsituation, and the intermediateshading represents a stageduring the opening,possibly around 20 Ma. JOI.IVE-TET AL.: ARC DEFORMATIONAND MARGINAL BASIN OPENING 4381

Fig. 12. (continued)

experimentsof continentalcollision by Cobbold and Davy understand the respective parts played by the Pacific [1988] and Davy and Cobbold[1988], we proposed[Jolivet et subduction and the extrusion tectonics in terms of kinematics al., 1990] that the right-lateral faults in the Japan Sea and and, if possible, of forces. One important question is, for Bohai Gulf regions are N-S trendingconjugates of the major instance, the reason of the drastic change from extension to NE-SW trendingleft-lateral faults which guide the motion of compressionin Late Miocene and the resulting inception of a continentalblocks eastward.This is, however,possible, only new subductionzone along the easternmargin of the JapanSea if the Pacific subducitonprovides the low stresscondition in the Quaternary[Nakamura, 1983; Tamaki, 1988]. along the easternboundary of the deforming system.The descriptionof the geometry and mechanicsof the opening Acknowledgments. The authors are indebted to their made in the presentpaper will be used in our future works to colleagues in Japan, particularly K. Tamaki for fruitful 4382 JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING discussion.K. Otsuki and T. Ohguchiand S. Miyashitaand all Hayashida,A., Timing of rotationhalmotion of Southwest themembers of theHokkaido Tectonics Research group were Of Japan inferred from palcomagnetismof the Setouchi considerablehelp for the survey in northern Honshu and Miocene series, J. Geomagn. Geoelectr.,38, 295-310, Hokkaido. Thanks are due to R. McCabe and the second 1986. anonymousreviewer, as well as to R. Davies for useful Hilde, T. W. C., S. Uyeda, and L. Kroenke,Evolution of the commentsand improvementof the English.Part of this work western Pacific and its margins,Tectonophysics, 38, 145- wasf'manced by the FrenchMinistry of ForeignAffairs. 165, 1977. Hilde, T. W. C and J. M. Wageman,Structure and origin of the REFERENCES Japansea, in, The WesternPacific, Island Arcs, Marginal Seas, Geochemistry,edited by P. J. Coleman, Universityof Angelier, J., Tectonic analysis of faultslip data sets, J. WesternAustralia Press,Perth, pp. 123-137, 1975. Geophys. Res., 89, 5835-5848, 1984. Hirooka, K., H. Sakai, T. Takahashi, H. Kinoto, and A. Antipov, M.P., Kovylin, V. M. and V. P. Filat'yev, Takeuchi, Tertiary tectonic movement of central Japan Sedimentarycover of the deep water basinsof Tatar strait inferred from palcomagnetic studies, J. Geomagn. and the northernpart of the sea of Japan.lnt. Geol. Rev., Geoelectr., 38, 311-323, 1986. 22, 1327-1334, 1980. Honza, E. (Ed.), Geological investigationof the Japan sea, Byerlee, J. D., Friction of rocks, Pure Appl. Geophys.,116, April-June 1978 (GH-78-2 cruise),Geol. Surv.Jpn, Cruise 615-626, 1978. Rep.. 13, 99 pp., GeologicalSurvey of Japan,Tsukuba, CCOP-IOC, SouthwestJapan, Korea peninsula,International 1979. decadeof oceanexploration ]IDOE) and studiesin eastasian Hornafius, J. S., B. P. Luyendik, R. R. Terres, and M.J. tectonics and resources (SEATAR), U. N. ESCAP, CCOP, Kamerling,Timing and extentof Neogenetectonic rotation Tech. Bull., 123-137, 1980. in the WesternTransverse Ranges,California, Geol. Soc. Celaya, M. and R. McCabe, Kinematicmodel for the opening Am. Bull., 97, 1476-1487, 1986. of the Sea of Japanand the bendingof the Japaneseislands, Ichikawa, K., Geohistory of the Median Tectonic Line of Geology, 15, 53-57, 1987. Southwest Japan, Mere. Geol. Soc. Jpn, 18, 187-212, Chase, C. G., Extension behind island arcs and motion relative 1980. to hot-spots,J. Geophys.Res., 83, 5385-5387, 1978. Iijima, A., R. Tada,and Y. Watanabe,Development of Neogene Chen, W. P. and J. Nabelek, Seismogenicstrike-slip faulting sedimentarybasins in the northeasternHonshu arc with and the developmentof the North China basin,Tectonics, 7, emphasison Miocene siliceousdeposits, J. Fac. Sci.Univ. 975-989, 1988. Tokyo, 21, 417-446, 1988. Cobbold, P. R. and P. Davy, Indentationtectonics in nature Ingle, J.C., Summary of Late PaleogeneNeogene insular and experiments,2, Central Asia, Bull. Geol. lnst.Upsalla, stratigraphy, and correlations,Philippine Sea and sea of 14, 143-162, 1988. Japan region, in Initial Report of DSDP, vol. 31, US Davy, P. and P. R. Cobbold. Indentation tectonics in nature Government Printing Office, pp. 837-855, Washington and experiments.Experiments scaled for gravity,Bull. Geol. D.C., 1975. Inst. Upsalla, 14, 129-141, 1988. Isezaki, N., A magneticanomaly map of the Japan sea, J. Fabbri, O., J. Charvet, and M. Faure, Sur la d6formation Geomagn.Geoelectr., 38, 403-410, 1986. associ6e•t la rotation de Kyushu (Japon SW) au Miocene Isezaki,N. and S. Uyeda,S., Geomagneticanomaly pattern of Moyen, C. R. Acad Sci. Paris, 304, 1207-1212, 1987. the Japansea, Mar. Geophys.Res., 2, 51-59, 1973. Faug•re, E. and J.P. Brun, Mod•lisation exp6rimentale de la Ishikawa, N., and M. Torii, Counterclockwizerotations of the distension continentale, C. R. Acad. Sci. Paris, 299, 365- Tsushima and Goto islands, Southwestmargin of the Japan 370, 1984. Sea, evidences from palcomagnetism,paper presented at Faure, M. and F. Lalev•e Bent structural trends of Japan, 1988 DELP Tokyo International Symposium, Tokyo flexural slip folding related to the Neogeneopening of the Intsitute of Technology,Tokyo, Dec. 13-16, 1988. sea of Japan,Geology, 15, 49-52, 1987. Ishizuka, H., Greenstonesfrom the Idonnappuformation along Fukao, Y. and M. Furumoto,Mechanisms of large earthquakes the river Oku-Niikappu in the axial zone of Hokkaido, along the easternmargin of the Japansea, Tectonophysics, Japan,Mere. Fac. Sci., Kochi Univ.,2, 1-22, 1981. 25, 247-266, 1975. Ishizuka H., M. Imaizumi, N. Gouchi, and S. Banno, The Ganzawa,Y., "Green Tuffs" movementdefined by fission-track Kamuikotanzone in Hokkaido, Japan,tectonic mixing of ages of igneous rocks, 1, Sado Island, J. Geol. Soc. Jpn., high pressureand low pressuremetamorphic rocks, J. 88, 943-956, 1982. Metamorph. Geol., 1, 263-265, 1983. Ganzawa,Y., "GreenTuffs" movementdefined by fission-track Itoh, Y., Differential rotationof norteasternpart of southwest ages of igneous rocks, 2, Futimoyama area of Toyama Japan,palcomagnetism of Early to Late Miocenerocks from prefecture,central Japan,J. Geol. Soc. Jpn.., 89, 271-286, Yatsuoarea in Chubudistrict, J. Geomagn.Geoelectr., 38, 1983. 325-334, 1986. Ganzawa, Y., Fission tracks age of volcanic rocks from Itoh, Y., Differential rotation of the easternpart of southwest Cretaceousto Tertiary in the inner belt of northeastJapan, Japan inferred from palcomagnetismof Cretaceousand Okushiri island, Oga peninsula and Asahi mountains, J. Neogenerocks, J. Geophys.Res., 93, 3401-3411, 1988. Geol. Soc. Jpn., 93, 387-401, 1987. Itoh, Y., and H. Ito, Tertiary rotational movement of the Goetze,C. andB. Evans,Stress and temperaturein the bending eastern part of southwest Japan, palcomagnetismand lithophere as constrainedby experimentalrocks mechanics, fission-trackdating of the FutomiyamaGroup in the Uozu Geophys.J.R. Astron. Soc.,59, 463-478, 1979. area,Toyama prefecture, Japan, J. Geol.Soc. Jpn., 94, 11- Hamburger, M.W. and B. L. Isacks, Diffuse back-arc 18, 1988. deformation in the southwestern Pacific, Nature, 332,599- Jackson,J.A., Active continental deformation and regional 604, 1988. metamorphism,Philos. Trans. R. Soc.London., 321, 47- Harding, T. P., Seismic characteristicsand identification of 66, 1987. negative flower structures, positive flower structuresand Jolivet, L., The Hokkaido Central Belt, the successionof positive structural inversions, Am. Ass. Petr. Geol. Bull., tectonicstages, Bull. Soc. Gdol. France,2, 311-327., 69, 582-600, 1985. 1986a. JOLIVET ET AL.' ARC DEFORMATION AND MARGINAL BASIN OPENING 4383

Jolivet,L., America-Eurasiaplate boundaryin easternAsia and Ludwig, W. J., S. Murauchi, and R. E. Houtz, Sedimentsand the openingof marginalbasins, Earth Planet.Sci. Lett., 81, structureof the Japansea. Geol. Soc. Am. Bull., 86, 651- 282-288, 1986b. 664, 1975. Jolivet, L., P.Davy, and P. R. Cobbold, Right-lateral shear Mandl, G., Tectonic deformation by rotating parallel faults, along the northwestPacific margin and the India-Eurasia the "bookshelf" mechanism, Tectonophysics, 141, 277- collision, Tectonics, 9, 1409-1419, 1990. 316, 1987. Jolivet, L. and P. Huchon, Crustal scale strike-slipshear zone Maritime Safety Agency of Japan, Bathymetric map of in Hokkaido, Northeast Japan, J. Struct. Geol., 1 I, 509- Hokkaido, 1:1 000 000, folio 6311, 1980a. 522, 1989. Maritime Safety Agency of Japan,Bathymetric map of north- JolivetL. andS. Miyashita,The HidakaShear Zone (Hokkaido, east Japan,1:1 000 000, folio 6312, 1980b. Japan), genesisduring a right-lateralstrike slip movement, Martell, S. J., D. D. Pollard, and P. Segall, Developmentof Tectonics, 4, 289-302, 1985. simple strike-slip fault zones, Mount Abbot quadrangle, Jolivet, L., P. Huchon, and C. Rangin, Tectonic setting of Sierra Nevada, California, Geol. Soc. Am. Bull., I00, 1451- westernPacific marginalbasins. Tectonophysics, 160, 23- 1465, 1988. 48, 1989. McCabe, R., K. D. Min, J. Han, H. C. Han, and D. Lee, Blocks Kang, P. C., Geologic evolution of Korea and structural rotationsin southeastand east Asia, paper presentedat 1988 analysisof SeatarKorean transect area. U. N. ESCAP,CCOP DELP Tokyo International Symposium,Tokyo Institute of Tech. Bull. 14, 31-51, 1981. Technology, Tokyo, Dec. 13-16, 1988. Kawai, N., T. Nakajima,and K. Hirooka,The evolutionof the McKenzie, D., Spinning continents,Nature, 344, 109-110, island are of Japan and the formation of granitesin the 1990. citeurn Pacific belt, J. Geomagn.Geoelectr., 23, 267-293, Matsuda,T. and S. Uyeda, On the Pacific type orogenyand its 1971. model, extensionof the paired belts conceptand possible Kimura,G., Obliquesubduction and collision,forearc tectonics origin of marginal seas,Tectonophysics, II, 5-27, 1971. of the Kuril are, Geology, 14, 404-407, 1986. Mitani K., Changing of the Tertiary sedimentarybasins in the Kimura, G. and K. Tamaki, Collision, rotation and back arc western flank of the axial belt of Hokkaido- Bearing a spreading, the case of the Okhotsk and Japan seas, significanceof the Sunagawa lowland to Umaoi Hilly belt Tectonics, 5, 389-401, 1986. (in Japanese),Assoc. Geol. Collab. Jpn. Monogr, 21, 127- Kimura, G., S. Miyasaka, Y. Kontani, S. Miyashita, K. 137, 1978. tauymuaS,, and ,, ,•L,,,,,•o•,Tectonic significance of the Miyashita, S., Reconstructionof the ophiolite successionin Kamishiyubetsutectonic line in the uplift of the Hidaka the Western Zone of the Hidaka metamorphic belt, metamorphicbelt, Bull. Tectonic Res.Group Jpn., 27, 167- Hokkaido, J. Geol. Soc. Jpn., 89, 69-87, 1983. 177, 1982. Moteau, M.G., V. Courtillot, and J. Besse, Evidence for a Kimura G., S. Miyashita and S. Miyasaka,Collision tectonics Miocene widespread remagnetization in northern Japan, in Hokkaido and Sakhalin, in Accretion Tectonics in the Earth Planet. Sci. Lett., 84, 321-338, 1987. Circum Pacific Region, edited by M. Hashimotoand S. Moriya, T., Seismo tectonicsof Hokkaido basedon seismicity Uyeda, pp.117-128, Terrapub,Tokyo, 1983. and focal mechanismstudies (in Japanese),Assoc Geol. Kobayashi,K. and N. Isezaki, Magneticanomalies in the sea Collab. Jpn. Monogr., 31, 475-486., 1986. of Japan and the Shikoku basin, possible tectonic Murata, A., Conical folds in the Hitoyoshi bending, South implications,in The Geophysicsof the PacificOcean and Kyushu formed by the clockwiserotation of the Southwest Its Margins,Geophys. Monogr. Ser., vol. 19, editedby G. Japanarc. J. Soc. Geol. Jpn., 93, 91-105, 1987. H. Sutton,M. H Manghnani,and R. Moberly, pp. 235-251, Nabelek, J., W. P. Chen, and H. Ye, The Tangshanearthquake AGU, Washington,D.C., 1976. sequenceand its implicationsfor the evolution of the North KomatsuM., S. Miyashita, J. I. Maeda, Y. Osanai and T. China basin, J. Geophys.Res., 92, 12615-12628, 1987. Toyoshima,Disclosing of a deepestsection of continental Nakamura, K., Possiblenascent trench along the easternJapan type crustupthrust as the final eventof collisionof arcsin sea as the convergentboundary between Eurasia and North Hokkaido, North Japan, in Accretion tectonics in the American plates (in Japanese),Bull. EarthquakeRes. Inst., Circum Pacific Regions, edited by M. Hashimotoand S. 58, 721-732, 1983. Uyeda, pp. 149-165, Terrapub,Tokyo, 1983. Nakano, N., Metamorphism of the greenstones in the Koshiya,S., TanakuraShear Zone, the deformationprocess of Kamuikotanzone and the Hidaka WesternMarginal tectonic fault rocks and its kinematics.J. Geol. Soc. Jpn., 92, 15- zone in the Shizunai-Mitsuishi district, Hokkaido, J. Geol. 29, 1986. Soc. Jpn., 87, 211-214, 1981. Lallemand, S. and Jo!ivet, L., Japan Sea, a pull apart basin, Nut, A., H. Ron, and O. Scotti, Fault mechanics and the Earth Planet. Sci. lett., 76, 375-389, 1985. kinematicsof block rotations,Geology, •4,746-749, 1986. Lallemand, S., H. Okada, K. Otsuka, and L. Labeyrie, Okada, H., S. Lallemand, K. Otsuka, and L. Labeyrie, Tectoniqueen compressionsur la marge Est de la Met du Submarine geologic structureof the easternmargin of the Japon, raise en 6vidence de chevauchements•t vergence sea of Japan with special reference to the nascent trench orientale, C. R. Acad. Sci. Paris, 301, 201-206, 1985. problem, Geosci. Rep. ShizuokaUniv., 11, 119-133, 1985. Lamb, S. H., Tectonicrotations about vertical axes during the Osanai, Y., K. Arita, and M. Bamba, P-T conditions of last 4 Ma in part of the New Zealandplate boundaryzone, J. granulite facies rocks from the Hidaka metamorphicrocks, Struct. Geol., 10, 875-893, 1988. Hokkaido, Japan,J. Geol. Soc. Jpn., 92, 793-808, 1986a. Lee, J. S., P6trologie et relations structuralesdes volcanites Osanai, Y., S. Miyashita, K. Arita, and M. Bamba, The Cr6tac6 •t C6nozoYquesde la Cor6e du Sud, implications metamorphism and thermal structure of the collisional g6odynamiquessur la marge est-eurasiatique,Th•se de terrain of a continental and oceanic crusts, a case of the Doctorat, Universit6 d'Or16ans, Orl6ans, 1989. Hidaka metamorphicbelt, Hokkaido, Japan,Assoc. Geol. Lee, J. S. and A. Poutlet, Le volcanismen6og•ne de Pohang Collabor. Jpn.Monogr.,31, 205-222, 1986b. (SE Cor6e), nouvelles contraintesg6ochronologiques pour Otofuji Y. and T. Matsuda, Paleomagneticevidence for the l'ouverturede la ruer du Japon,C.R. Acad. Sci. Paris, 307, clockwise rotation of southwestJapan, Earth Planet. Sci. 1405-1411, 1988. Lett., 62, 349-359, 1983. 4384 JOLIVET ET AL.: ARC DEFORMATION AND MARGINAL BASIN OPENING

Otofuji, Y. and Matsuda,T., Timing of rotationhalmotion of Tapponnier,P. andP. Molnar,Slip-line field theoryand large southwest Japan inferred from paleomagnetism.Earth scale continentaltectonics, Nature, 264, 319-324, 1976. Planet. Sci. Lett., 70, 373-382, 1984. Tapponnier,P., G. Peltzer,A. Y. Le Dain, R. Armijo,and P. Otofuji Y., T. Matsudaand S. Nohda, Paleomagneficevidences Cobbold, Propagatingextrusion tectonics in Asia, new for the Miocene counter clockwise rotation of northeast insights from simple experimentswith plasticine, Japan- rifting processof the Japanarc, Earth Planet. Sci. Geology, 10, 611-616, 1982. Lett., 75, 265-277, 1985. Tapponnier,P., G. Peltzer,and R. Armijo,On the mechanics Otsuki, K., Reconstructionof Neogene tectonicstress field of of the collisionbetween India andAsia, Geol. Soc.Spec. Northeast Honshu arc from metalliferous veins. Mern. Soc. Publ., 19, 115-157, 1986. Geol. Jpn., 32, 281-304, 1989. Taylor,B. andG. D. Karner,On the evolutionof marginal Otsuki K. and M. Ehiro M., Major strike slip faults and their basins, Rev. Geophys.,21, 1727-1741, 1983. bearingon the spreadingof the JapanSea, J. Phys. Earth, Tosha,T., Paleomagnetismof northeastJapan. Ph.D. thesis, suppl. , 26, 537-555, 1978. 77 pp., Univ. of Tokyo, 1983. Peltzer, G. and P. Tapponnier, Formation and evolution of Tosha,T. and Y. Hamano,Paleomagnetism of Tertiary rocks strike-slip faults, rifts and basins during the India-Asia fromthe Ogapeninsula and the rotation of northeastJapan, collision: An experimentalapproach, J. Geophys.Res., 93, Tectonics, 7, 653-662, 1988. 15085-15117, 1988. Uda, T., Deformation of granitic pebbles in "Utaro Reedman,A. J. and S. H. Um, The Geologyof Korea, 39 pp., conglomerate"at cape Erimo, Hokkaido,Japan, J. Geol. GeologicalMining Institute, Korea, 1975. Soc. Jpn., 79, 391-398, 1973. Ron, H., R. Freund, Z. Garfunkel, and A. Nut, Block rotation Uda,T., Polyphasedeformation of the CapeErimo area by by strike slip faulting, structural and paleomagnetic changeof tectonicstress field. J. Geol.Soc. Jpn., 82, 1-18, evidence,J. Geophys. Res., 89, 6256-6270, 1984. 1976. Rozhdestvenskiy, V. S., The role of wrench faults in the Uyeda,S., Facts,ideas and open problems on trench-arc-back structure of Sakhalin, Geotectonic& 16, 323-332, 1982. arc systems,in The Origin of Arcs, editedby F.C. Wezel, Rozhdestvenskiy, V. S., Evolution of the Sakhalin fold pp. 435-460, Elsevier Science,New York, 1986. system,Tectonophysics, 127, 331-339, 1986. UyedaS. andH. Kanamori,Backarc opening and the mode of Sasajima,S., PreNeogenepaleomagnetism of Japaneseislands subduction,J. Geophys.Res., 84, 1049-1061, 1979. (and vicinities), in Paleoreconstructionsof the Continents, Vendeville,B., P. R. Cobbold,P. Davy, J.P. Brun,and P. Geodynarn.Ser., vol. 2, editedby M. W. Mac Elhinny and Choukroune,Physical models of extensionaltectonics at D. A. Valencio, pp. 115-128, AGU, Washington,D.C., variousscales, Geol. Soc.London Spec. Publ., 28, 95-107, 1981. 1987. Sayanagi,K., N. Isezaki, and Y. Kitahara,Report on DELP Walcott,R. I., The kinematicsof the plateboundary zone 1985 cruisesin the Japan Sea, IV, Geomagneticanomalies throughNew Zealand,a comparisonof shortand long term over the seamountsin the Yamato Basin, Bull. Earthquake deformation, Geophys.J. R. Astron. Soc.,79, 613-633, Res Inst., Univ. Tokyo,62, 391-415, 1987. 1984. Schluter,H. U. and W. C. Chun, Seismicsurvey off the east Walcott,R. I., Geodeticstrain and the deformational history of coast of Korea, U. N. ESCAP, CCOP Tech. Bull., 8, 1-15, theNorth island of New Zealandduring the LateCainozoic, 1974. Philos. Trans. R. Soc. London,321, 163-181, 1987. Shibata K., K/At age determinations on granitic and Watanabe, Y., Deformation structure of the Uenshiri horst in metamorphicrocks in Japan,Rep. Geol. Surv. Jpn., 227, 1- the Hidaka belt, Central Hokkaido, J. Geol. Soc. Jpn., 94, 73, 1968. 527-533, 1988. Shibata K. and S. Ishihara, Rb/Sr whole rock and K/Ax mineral Watanabe Y. and K. Iwata, The age of the Miocene agesof graniticrocks in Japan,Geochern. J., 13, 113-120, Kamishiyubetsu formation in northern Hokkaido and the 1979. basins formed by tectonic movements,J. Geol. Soc. Jpn., Shibata, K., S. Uchiumi, K. Uto, and T. Nakagawa, K-At 91, 427-430, 1985. results, 2, New data from the geologicalsurvey of Japan, Yamagishi,H. and Y. Watanabe,Change of stressfield of Late Bull. Geol. Surv. Jpn., 35, 331-340, 1984. Cenozoic Southwest Hokkaido, Japan, investigation of Sillitoe, R. H., Metallogeny of an Andean type continental geologic faults, dykes, ore veins and active faults. Geol. margin in South Korea, implications for opening of the Collabor. Jpn. Monogr., 31, 321-332, 1986. JapanSea, in Island Arcs, Deep Sea Trenchesand BackArc Yamaji, A., Geology of the Atsumi area and Early Miocene Basins, Maurice Ewing Ser., vol. 1, edited by M. Talwani rifting in the Uetsu District, NortheastJapan, Mern. Geol. and W. C. Pitman Ill, pp. 303-310, AGU, Washington,D. Soc. Jpn., 32, 305-320, 1989. C., 1977. Tamaki, K., Two modes of back arc spreading,Geology, 13, 475-478, 1985. J.P. Bnm, Laboratoirede Tectonique,Universit6 de Rennes Tamaki, K., Age estimationof the Japan Sea on the basis of 1, Avenue du G6n6ral Leclerc, 35042, Rennes Cedex, France. stratigraphy, basement depth and heat How data, J. N. Chamot-Rooke, P. Huchon, L. Jolivet, X. Le Pichon and Geomagn. Geoelectr., 38, 427-446, 1986. J.C. Thomas, D6partement de G6ologie, Ecole Normale Tamaki, K., Geological structure of the Japan Sea and its Sup6rieure,24 rue Lhomond,75231 ParisCedex, France. tectonic implications,Bull. Geol. Surv. Jpn., 39, 269-365, 1988. Tamaki, K. and E. Honza, Incipient subductionand obduction (ReceivedApril 17, 1989; along the easternmargin of the JapanSea, Tectonophysics, revisedApril 16, 1990; 119, 381-406, 1985. acceptedJuly 5, 1990.)