BasinResearch ...burbank.faculty.geol.ucsb.edu/Site/Publications_files/Lu...fold-and-faultzoneoftheforeland(Fig.1).Ourpaleomag-netic investigations provide a robust temporal framework

Late Cenozoic structural and stratigraphic evolutionof the northern ChineseTian Shan forelandHonghua Lu,n,w Douglas W. Burbank,w Youli Lin and Yunming LiuznCollege of Urban and Environmental Sciences, PekingUniversity, Beijing, PR ChinawInstitute for Crustal Studies, University of California at Santa Barbara, California, USAzSchool of Geographical Sciences, Southwest University, Chongqing, PR China

ABSTRACT

Three successive zones of fault-related folds disrupt the proximal part of the northernTianShan forelandinNWChina.A newmagnetostratigraphy of theTaxiHe section on the north limb of theTugulu anticlinein the middle deformed zone clari¢es the chronology of both tectonic deformation and depositionalevolution of this collisional mountain belt. Our �1200-m-thick section encompasses the upperCenozoic terrigenous sequence withinwhich �300 sampling horizons yield an age span of �8^2Ma.Although the basal age in theTaxiHe section of theXiyu conglomerate (often cited as an indicator of initialdeformation) is �2.1Ma,much earlier growth of theTugulu anticline is inferred fromgrowth strata datedat �6.0Ma.Folding ofNeogene strata and angular unconformities in anticlines in themore proximal anddistal deformed zones indicate deformation duringMiocene and Early Pleistocene times, respectively. IntheTaxiHe area, sediment-accumulation rates signi¢cantly accelerate at �4Ma, apparently in responseto encroaching thrust loads.Together, growth strata, angular unconformities, and sediment-accumulation rates document the northward migration of tectonic deformation into the northernTianShan foreland basin during the lateCenozoic.A progradational alluvial̂ lacustrine system associatedwiththis northward progression is subdivided into two facies associations atTugulu: a shallow lacustrineenvironment before �5.9Ma and an alluvial fan environment subsequently.The lithofacies progradationencompasses the time-transgressive Xiyu conglomerate deposits, which should only be recognized as alithostratigraphic unit. Along the length of the foreland, the locus of maximum shortening shifts betweenthe medial and proximal zones of folding, whereas the total shortening across the foreland remains quitehomogeneous along strike, suggesting spatially steady tectonic forcing since lateMiocene times.

INTRODUCTION

TheCenozoic India-Asia collision (Molnar &Tapponnier,1975; Gansser, 1981; Molnar et al., 1993; Tapponnier et al.,2001) has signi¢cantly impacted the tectonic and climaticenvironment of high Asia. Exceptional topographic reliefand the thick Cenozoic terrigenous sequences that havebeen shed into foreland basins provide an opportunity toprobe the history of tectonic deformation related to thiscollision and to understand the tectonic and/or climaticcontrols on syndepositional systems.

Separating theTarim Basin to the south from the Jung-gar Basin to the north, the east^west trending Tian Shanlies about 2000 km north of the initial India^Asia collisionfront (Deng et al., 2000; Charreau et al., 2005).Thick upperCenozoic terrigenous sequences were deposited in theintramontane and foreland basins of the Tian Shan (e.g.Burbank et al., 1999; Bullen et al., 2001; Sun et al., 2004;

Charreau et al., 2005, 2006; Chen et al., 2007; Heermanceet al., 2007).Whereas the scarcity of faunal or £oral fossilshas provided only limited biostratigraphic age constraintson foreland strata, magnetostratigraphy has been success-fully used to calibrate the terrestrial depositional record inthis region (e.g. Zheng et al., 2000; Chen et al., 2002, 2007;Sun et al., 2004, 2005a, b, 2007; Charreau et al., 2005, 2006,2008b; Huang et al., 2006; Heermance et al., 2007), despitesome inherent uncertaintieswhenever dating sedimentarystrata (Talling & Burbank, 1993). Many previous studieshave focused on the timing of tectonic growth of theTianShan using combinations of geologic and geodetic data(e.g. Hendrix et al., 1992; Avouac & Tapponnier, 1993;Avouac et al., 1993; Abdrakhmatov et al., 1996; Me¤ tivier &Gaudemer, 1997; Yang & Liu, 2002; Fu et al., 2003), ther-mochronology (e.g. Hendrix et al., 1994; Dumitru et al.,2001;Bullen etal., 2003; Sobel etal., 2006), or magnetostra-tigraphy (e.g. Sun et al., 2004; Charreau et al., 2005, 2006;Huang et al., 2006). In contrast, relatively few studies (Bul-len et al., 2001; Chen et al., 2002, 2007; Heermance et al.,2007) have focused on sedimentary facies in a robust tem-poral framework in order to delineate the depositional

Correspondence: Honghua Lu, Rm. 516, Ocean Building,TongjiUniversity, Shanghai 200092, PR China. E-mail: [email protected]

BasinResearch (2009) doi: 10.1111/j.1365-2117.2009.00412.x

r 2009 The AuthorsJournal Compilationr 2009 Blackwell Publishing Ltd, European Association of Geoscientists & Engineers and International Association of Sedimentologists 1

evolution of the upper Cenozoic coarsening-upward terri-genous sequences in theTian Shan.

Based on a magnetostratigraphic study, Sun et al. (2004)concluded that thick conglomerate deposits were direct in-dicators of tectonic uplift in theTian Shan. Others contendthat, although changes in lithofacies and accumulation ratescould indicate nearby tectonic deformation (Zheng et al.,2000; Charreau et al., 2005; Sun et al., 2005b; Huang et al.,2006), analogous changes could be driven by nontectonicforcing, including variations in climate (Molnar, 2001,2004;Zhangetal., 2001;Dupont-Nivet etal., 2007), lithologicresistance in the source area, or transport distance (DeCelles& Giles, 1996; DeCelles et al., 1998; Chen et al., 2002; Char-reau etal., 2006;Heermance etal., 2007)with the last variablebeing especially important in settings where tectonicdeformation migrated in steps toward the foreland basin.Furthermore, lithofacies changes, such as the onset of con-glomeratic deposition at any single location within a basin,commonly include a time lag from causative events in thehinterland (Jordan et al., 1988; Jones et al., 2004) or may onlyindicate local structural in£uences, rather than range-widechanges (Heermance et al., 2007).Thus, tectonic interpreta-tions of conglomeratic facies are best made in a reliable,three-dimensional, temporally controlled, depositional fra-mework (Chen et al., 2002, 2007; Heermance et al., 2007).Such contexts illuminate patterns of time-transgressivefaciesmigration andhelp restrain the tendency to treat faciesin anygiven section as chronostratigraphic units (e.g.Li etal.,1979; Huang &Cheng,1981; Liu et al., 1996).

Consequently, in order to improve understanding of thedeformational history and of possible controls on deposi-

tion in the northern Tian Shan foreland, we completed amagnetostratigraphic study of a �1200-m-thick, upperCenozoic terrigenous sequence of foreland strata onthe north limb of the Tugulu anticline in the middlefold-and-fault zone of the foreland (Fig.1).Our paleomag-netic investigations provide a robust temporal frameworkspanning from �8 to �2Ma within which we analyzesedimentary facies and discuss the controls of tectonicdeformation on depositional evolution along the northernpiedmont of the ChineseTian Shan.

GEOLOGICAL SETTING ANDSTRATIGRAPHY

The ancestral Tian Shan was formed during the Permianafter experiencing two Paleozoic accretion events withinpaleo-Asia (Windley et al., 1990; Allen et al., 1993; Carrollet al., 1995; Sobel et al., 2006): one along the southern mar-gin of the range (Late DevonianÊarly Carboniferous),which resulted in the accretion of the Tarim block ontothe central Tian Shan; and the other along the northernmargin (Late CarboniferousÊarly Permian), which led tothe amalgamation of the northern Tian Shan with thecombined Tarim-central Tian Shan continental block.Thesubsequent Mesozoic deformation of the Tian Shan ischaracterized by three distinguishable episodes in responseto successive accretion onto the south Asian margin of theQiangtang block in the Late Triassic, the Lhasa block inthe Latest JurassicÊarly Cretaceous, and the Kohistan^Dras arc complex in the Late Cretaceous periods (Hendrix

Fig.1. (a) Overall geological setting in Central Asia (after Hendrix et al., 1994). (b) Geological map of fault-and-fold zones IÎII in thenorth margin of theTian Shan (modi¢ed from Zhang et al., 1994; Deng et al., 2000), showing the tectonic setting and the Phanerozoicstrata. Solid star indicates the paleomagnetic sampling site in this study at Taxi He. Open circle, triangle and star show the locations ofthe KuitunHe (west) section (Sun et al., 2007), the KuitunHe (east) section (Charreau et al., 2005) and the Dushanzi section (Sun et al.,2004), respectively.

r 2009 The AuthorsJournal Compilationr 2009 Blackwell Publishing Ltd, European Association of Geoscientists & Engineers and International Association of Sedimentologists2

H. Luet al.

et al., 1992; Dumitru et al., 2001).The Late Jurassic event re-sulted in a widespread unconformity beneath the overlyingCretaceous strata that is recognized throughout much oftheTian Shan. During the Paleogene, the relative tectonicstability that prevailed in theTian Shan (Bullen et al., 2001,2003; Sobel et al., 2006) resulted in the beveling of topogra-phy and created severalwidespread regional unconformities(Burtman, 1975; Bullen et al., 2001) that are documented inboth the northern ChineseTian Shan (Windley et al., 1990;Allen et al., 1993; Avouac et al., 1993) and the Kyrgyz TianShan (Bullen et al., 2001, 2003; Oskin & Burbank, 2005,2007; Sobel et al., 2006).

The Mesozoic strata deposited in theTian Shan rangeare characterized bybraided/meandering £uvial sedimentsand alluvial conglomerate (Hendrix et al., 1992). Althoughthese strata are dominated by subaerial deposits, Juras-sic^Cretaceous organic or carbonate-rich strata provideevidence for local lacustrine environments at this time(Hendrix et al., 1992; Deng et al., 2000). Thick lowerTertiary ¢ne-grained deposits (composed mainly of mud-stone and siltstone) reveal that lacustrine environmentsexisted intermittently in the northern ChineseTian Shanforeland during the Early Tertiary. This interpretation isfurther reinforced by the presence of the gray, calcareous,Oligo-Miocene strata of the Huoerguosi anticline in themiddle deformed zone (Fig.1b, Deng et al., 2000).

Our work is focused on the north £ank of the ChineseTian Shan, where thick Cenozoic deposits were shed intothe Junggar foreland basin from the south.The present to-pography and high elevation of theTian Shan are the resultof north^south convergence driven by the India^Asia colli-sion during Cenozoic times (Tapponnier & Molnar, 1979;Patriat & Achache, 1984; Avouac et al., 1993) (Fig.1a).Withinthe northern piedmont depression zone of theTian Shan,this north^south compression produced three sub-parallelridges of folds, known as fault-and-fold zones I^III, num-bered sequentially outward from the mountains (Avouac etal., 1993; Burch¢el et al., 1999; Deng et al., 2000; Fu et al.,2003; Molnar et al., 1994) (Fig. 1b). North- £owing rivers in-cise the anticlines perpendicular to strikewhere they exposethick successions of Cenozoic deposits.TheTaxi He region(He is Mandarin for ‘river’) is a typical example, where thenorth- £owingTaxi He has sliced through theTugulu anti-cline (Figs 1b and 2), thereby revealing well-exposed out-crops on the northern limb of this fold (Fig. 2).

The outcrop width of the Taxi He section approaches2500m (Figs 2 and 3a). In general, the dips of the exposedstrata in the section gradually decrease northward fromvertical beds near the core of the fold to �451 at the top(Fig. 3a). A change of4101 in dip occurs at �350m abovethe base of the section (Fig. 3a) and is interpreted as repre-senting initial deposition of growth strata as discussedlater. No signi¢cant faulting or major erosional discor-dance was found in the measured section, suggestingcontinuous deposition without signi¢cant breaks orduplication. The measured Taxi He section comprisesthree formations from base to top: theTaxihe, Dushanzi,and Xiyu Formations (Bureau of Geological and Mineral

Resources of the Xinjiang Uygur Autonomous Region,1993) with thicknesses of �500, �660, and �40m,respectively (Fig. 3a). Several hundred meters of poorlyexposedXiyu strata overlie the studied section.TheTaxiheFormation consists dominantly of light brownish, ¢ne-textured muddy siltstone and mudstone beds interbeddedwith coarse-grained sandstone or conglomerate deposits.The age of this formation in the northern Tian Shanforeland is inferred to be Miocene based mainly on thepresence of Ostracoda fossils (Editing Committee of theStratigraphy of China, 1999). The Dushanzi Formationprimarily comprises coarse conglomerate, sandstone, andsiltstone beds. According to the presence of the Ostracodafossil group of Ilyocypris^Cypideris^Candona and mamma-lian fossil ofHipparion, the age of the Dushanzi Formationis traditionally inferred to bePliocene (EditingCommitteeof the Stratigraphy of China, 1999).The Xiyu Formation,also known as the Xiyu conglomerate, is widespread inthe piedmonts in northwest China (Liu et al., 1996), whereit typically consists of a thick sequence of gray or gray^black gravels in which some interbedded sands or sandylenses are found, especially in the lowermost part. In mostplaces, the dark Xiyu conglomerate clearly contrasts withthe lighter coloredDushanzi conglomerates wherever theyare juxtaposed.On the basis of the mammalian fossilEquussanmeniensis found in a section located within the Anjihaianticline (Peng,1975) to the west of Taxi He, the age of thisformation is commonly inferred to be latest Pliocene toearly Pleistocene.

MAGNETOSTRATIGRAPHY

Paleomagnetic sampling andmethods

Paleomagnetic sampling was conducted along both thewest and east bank of Taxi He (Figs 2 and 3a).Two or threesamples were collected from each of 303 horizons, yieldinga total of 676 samples for theTaxi He section.The verticalsampling interval varied between 0.75 and 13.4m with anaverage interval of �4m. All the samples were collectedwith a portable drill and orientedwith amagnetic compassthat was corrected for the local magnetic declinationanomaly (3.51E in the Taxi He area). Most samples werecollected from mudstone or siltstone beds. Where thedeposits are predominantly conglomeratic, especially inthe uppermost part of the sampled section, samples werecollected in thin lenticular beds of siltstone or ¢ne-grained sandstone. All the samples were cut into 2.0-cmlong cylinders that were 2.5-cm in diameter.

Remnant magnetization was measured with a 2Gcryogenic magnetometer at the Paleomagnetic Laboratoryof the Institute of Geomechanics, Chinese Academyof Geological Sciences in Beijing. The magnetometer islocated inside a set of large Helmholtz coils that reducedthe ambient geomagnetic ¢eld to around 300nT. Onesample per horizon (a total of 303 pilot samples) was sub-jected to stepwise thermal demagnetization in an ASC


NorthernTian Shan foreland evolution

TD-48 oven with an internal residual ¢eld o10nT. Forthese samples, 13^17 steps were applied, commonly withthe following stepwise heating routine: (i) 601C steps upto 3001C; (ii) 401C or 501C steps up to 5801C; (iii) 301Csteps up to 6401C; and (iv) 201C steps up to 6801C.

Demagnetization and statistical analysis

The intensities of the natural remanent magnetization(NRM) of the specimens were between 0.14 � 10� 2 and6.7 � 10� 2A/m with an average of 1.5 � 1.0 � 10� 2A/m.

Progressive demagnetization successfully isolated domi-nant magnetic components for the majority of the speci-mens after removing a viscous component ofmagnetization. According to their demagnetization beha-vior, all the samples are interpreted to display three majorcomponents (Fig. 4). A low-temperature component be-tween 01C and180^2401C represents �10^50% of the to-tal remanence and is either normal polarity or overlapswith the next reversed component resulting in an unde-¢ned polarity direction.One or two intermediate tempera-ture components (ITC), representing between 35 and 60%

Fig. 2. (a) Aerial photo view of the paleomagnetic sampling section in this study. (b) Locations of 303magnetic sampling sites alongTaxiRiver (all sampleswere collected in 2006).Normal (reversed) sites depicted as closed (open) circles.‘No.1’,‘R1’, and‘N1’show the sampling sitenumber andmagnetozones of reversed or normal polarity, respectively.The site(s) contained in each magnetozone are listed in the lower left.


H. Luet al.

of the total remanence, are observed between 180^2401Cand 540^580 1C.The ITC component is likely carried bymagnetite or titanomagnetite, as suggested by the drop inremanence that occurs above 540^580 1C. Representing10^40% of the total remanence, the high-temperaturecomponent (HTC) is manifested in a temperature rangeextending from �540^5801C to 610^6801C and com-monly lies in the same direction as the ITC.This compo-nent is most likely carried by hematite or maghemite(Butler, 1992).

About 17 samples (out of 303 pilot samples) showedonly unstable high-temperature demagnetization trajec-tories. For these stratigraphic horizons, another round ofprogressive thermal demagnetization was performed on asecond sample. The characteristic directions of magneti-zation were determined by principal component analysis(Kirschvink, 1980). Data from a minimum of three, butmore typically six to nine temperature steps, which werechosen above 4501C, were used for least-squares ¢ts.About 10% of the samples displayed relatively scattereddemagnetization directions. For these, the origin of thevector-component diagram was included as an additionaldatum when determining the characteristic directions ofmagnetization of these samples. For all samples, typicalmaximum angular deviation was between 1.21 and 151with an average of 5.5 � 2.91. Based on the statistics ofthe best- ¢t lines, we eliminated 29 samples with poorlyde¢ned magnetic directions. The magnetic declinationand inclination of the samples were then used to calculatevirtual geomagnetic pole (VGP) latitudes. Samples withVGPso451were not used for de¢ning the magnetic polar-ity stratigraphy (MPS) for theTaxi He section exceptwhentheywere adjacent to sites of the same polarity that showedVGPs 4451. This ¢ltering eliminated 28 sampled hori-zons. In total, 246 samples (one specimen per acceptedhorizon) were accepted to construct the MPS of theTaxiHe section.

Reversal and fold tests were employed in order to dis-cern the timing of characteristic remanent magnetization(ChRM)acquisition.Tilt correction signi¢cantly improvesthe grouping of ChRM poles (Fig. 5).The Fisherian site-averaged normal polarity and reversed polarity poles oftheTaxi He section have an observed angular di¡erence of3.11 and a calculated critical angle of 9.81 [95% con¢dencelevel; assuming identical dispersion (K) values].This indi-cates that the section passes the reversal testwith a‘B’qual-ity classi¢cation (McFadden & McElhinny, 1990). Becauseno samples were collected from the south limb of theTu-gulu anticline, the simple fold test of McElhinny (1964)was used with Kcrit de¢ned as the statistical f distributionwith 2(N�1) degrees of freedom, whereN is the numberof samples of a speci¢ed polarity. Kcrit is compared withthe ratio ofK2/K1, whereK1 is the dispersion in geographiccoordinates, and K2 is the dispersion in tilt-corrected co-ordinates. A positive fold test for this section is indicatedby the K2/K1 values of 1.53 (normal polarity) and 1.35 (re-versed polarity), which are, respectively, larger than theKcrit values of 1.24 (normal) and 1.22 (reversed) at the 95%con¢dence level.Thus, the positive reversal and fold testsare interpreted to indicate that the ChRM directions wereacquired by the sediments during or soon after deposition.

MPS

When constructing the MPS for theTaxi He section withVGP data, two samples with VGP latitudeso451 were re-tained because they reinforce an adjacent, single, high-VGP-latitude site, and together these sites serve to de¢nepolarity zone N10 (Fig. 6d). Six magnetozones (n1^n5, r1)with only one sample are represented by a short half bar(Fig. 6d), and were not used to de¢ne the MPS of theTaxiHe section or its correlationwith the geomagnetic polaritytime scale (GPTS). Overall, 11normal (N1^N11) and11 re-versed (R1^R11)magnetozones are clearly identi¢ed in the

Fig. 3. (a)Measured geological section atTaxi He. Stratigraphic dips steepensouthward from the lower part of theLower Pleistocene Xiyu Formation( �451) to the upperMioceneTaxiheFormation (881).‘T’and ‘B’ show the topand bottom of theTaxi He magneto-stratigraphic section, respectively. Solidnumbered circles denote three abruptchanges in bedding dip. (b) Geologiccross-section of theTugulu anticline alongtheTaxi He valley (modi¢ed fromBurch¢el et al., 1999). Based on themeasured Taxi He section (a), the ¢rstchange in dip is interpreted to representthe initial deposition of growth strata.



Fig.4. Representative thermal demagnetization plots (remanent magnetic intensity graphs (left) and both Zijderveld diagrams (center)and equal-area projections (right) in tilt-corrected coordinates) typically showing three magnetic components. Examples of the low-temperature component (LTC), the intermediate temperature components (ITC), and the high-temperature component (HTC) of thecharacteristic remanence magnetization (ChRM) are depicted in (a). 3005 300 1C.‘No. 2 (2.8m)’ indicates the sample number with thecorresponding stratigraphic level in the studied section.


H. Luet al.

Taxi He column (Fig. 6d). Each magnetozone is de¢ned byat least three paleomagnetic samples from di¡erentsampled horizons. Our magnetostratigraphy was corre-latedwith theGPTS of Lourens etal. (2004) (Fig. 6e) usingthe following characteristics: (1) three intervals with dis-tinctive patterns of local magnetic polarity (N1^N3, N4^N7, and N8^N9), (2) four relatively long reversed magne-tozones (R1, R4, R8, and R10), and one relatively long nor-mal magnetozone (N11), (3) the absence of signi¢cant

depositional interruption or faults throughout the section,(4) the presence of fauna/£ora fossils (Peng, 1975; EditingCommittee of the Stratigraphy of China, 1999) suggestingthe probability that theTaxi He section has been sampledwithinNeogene to Lower Pleistocene strata, and (5) an as-sumption that sediment-accumulation rates vary system-atically and relatively slowly.

Overall, our preferred correlation (Fig. 6d and e) suc-cessfully matches the major reversal patterns revealed in

Fig. 5. Stereonet plots of palaeomagneticdata from theTaxi He section. Small solid(inclined down) andwhite (inclined up)circles indicate individual measurementswith the larger circles showing thea95 error around the Fisher mean. Fisher’smean data are listed adjacent to each plot.

Fig. 6. Magnetic polarity stratigraphy(MPS) of theTaxiHe section. (a) Lithologylog with the base of growth-stratadepicted at �840m depth ( �6.1Ma). (b)Magnetic declination in tilt-correctedcoordinates (shaded zone equals 1801^3601). (c) Virtual geomagnetic pole (VGP)latitude. Open circles represent sampleswith absolute values of VGPso451, whichwere not used to de¢ne theMPS of theTaxi He section except for the localmagnetozone N10. Shaded zone equals� 451. (d)MPS de¢ned commonly bysamples withVGPs � 451. Black andwhite zones indicate normal and reversedpolarity, respectively. Boundaries betweenmagnetozones are de¢ned by themidpoint between two samples withopposite polarity.Magnetozones de¢nedonly by one sample are displayed by short-half bars. Ages of stratigraphic andlithofacies boundaries are shown adjacentto theMPS. (e) Reference geomagneticpolarity time scale (GPTS) of Lourenset al. (2004).



theLateMiocene toPlioceneGPTS(Lourens etal., 2004).Ofcourse, some ambiguities remainwith this correlation, for ex-ample, between normal magnetozones N10 andN11, where afew short-duration polarity events are missing in theTaxiHecolumn.We matched N11 with the subchron C4n.2n basedmainly on its relatively long span and on the correlation ofthe reversed magnetozone R10 with the subchron C3Ar. De-spite inevitable uncertainties, the correlation, especially be-tween N1-R10 and C2An.1n-C3Ar of the GPTS (Lourens etal., 2004) (Fig. 6d and e), appears reasonable.The ages of thetop and the base of theTaxi He section were calculated via alinear extrapolation of the sediment-accumulation rate in re-spective magnetozonesN1and R10, thereby constraining thesampled section to span from �8.1to �2Ma (Fig. 6). Simi-larly, the basal age of theXiyuFormation is dated at �2.1Ma.According to our preferred correlation, the relationship be-tween the stratigraphic thickness and the magnetostrati-graphic ages shows a signi¢cant increase in sediment-accumulation rate from �160 to �250mMyr�1 at �4Ma(Fig. 7), with an average rate of �200mMyr�1 over the en-tire 6-Myr interval of the sampled section.

SEDIMENTOLOGIC ANALYSIS

Depositional evolution of the forelandsince 8Ma

Based on the magnetostratigraphic age controls, the ¢rstoccurrence of thick conglomerate deposits in theTaxi Hesection occurs at �5.9Ma (Fig. 6a). The two superposed

parts of theTaxi He section separated by this age displaydistinct sedimentary characteristics.The strata depositedbefore �5.9Ma comprise largely ¢ne-grained mudstoneand siltstone. In contrast, the strata accumulated after�5.9Ma are dominated by coarse-grained conglomeratesor pebbly sandstone.These two distinguishing lithofaciesassociations de¢ne the evolution of an alluvial^lacustrinedepositional system in theTaxi He area.

Shallow lacustrine environment

The lowest part of the Taxi He stratigraphic section be-tween �1200 and �800m spans from �8.0 to �5.9Maand is dominated by thick (30^80m), laminated or tabular,light brown mudstones and siltstones (Figs 6a and 8a). Re-latively thin pebbly sandstone ( �0.5^3-m thick) and con-glomerate beds (up to 5m thick) are intercalated withinthese thick ¢ne-textured deposits, but represent o10%of the accumulated thickness (Fig. 8a). Erosional scoursup to �1m deep occur at the base of some of the thickerconglomerate beds. Sparse gypsum veins are also presentwithin the lower part of theTaxiHe section. Conglomerateinterlayers are characterized by poorly sorted, subangularto subrounded clasts in a sandy matrix and by massivebedding or crude strati¢cation.

The presence of horizontal bedding and gypsum, thewell-preserved ¢ne laminations, thick ¢ne-textureddeposits, and the absence of paleosols and mud cracks areinterpreted as recording a shallow lacustrine environmentin an arid setting (Ren & Wang, 1985; Magee et al.,

Fig.7. Age versus thickness plot of theTaxiHe section based on the correlation ofFig. 6d and e showing a450%acceleration in sediment-accumulationrate at �4.0Ma.The section has not beendecompacted, and thus calculatedsediment-accumulation rates representminima. Shaded box depicts the range oflikely ages for the acceleration in rates,based on a projection of the uncertaintieson each regression (gray shadedenvelopes).


H. Luet al.

1995; Song et al., 2001; Ort|¤ et al., 2003; Sun et al., 2004,2005a; Charreau et al., 2005; Chen et al., 2007; Heermanceet al., 2007). The lacustrine character of the ¢ne-textureddeposits that enclose the sand^matrix-supported, poorlysorted conglomerates suggests that these conglomeraticunits could represent subaqueous gravity £ows into a shal-low lacustrine environment (Nemec & Steel, 1984; Ren &Wang, 1985).

Alluvial fan facies

In contrast to the ¢ne-textured shallow lacustrine sedi-ments, the deposits younger than 5.9Ma are dominatedby thick cobbly conglomerate beds interlayeredwith eitherthin beds of siltstone and muddy siltstone or lens/beds ofsandstone. The overall coarse-grained deposition, how-ever, is strikingly interrupted between �400 and �580m(�3.4^4.3Ma; Fig. 6) by continuous thick beds of siltstoneand ¢ne-grained sandstone.

The most common lithofacies assemblage typicallyconsists of a composite unit of conglomerate, usually afew centimeters to several meters in thickness, commonlycapped by siltstone or ¢ne-grained sandstone and show-ing a crude ¢ning-upward signature. The conglomeratedeposits typically comprise clast supported, poorly sorted,subrounded to rounded clasts. Typically, the 10-m-thickconglomerate units are composed of several storeys thatare each �1^3m thick. Individual beds commonly displaymassive bedding with subrounded, crudely imbricatedclasts, and indistinct strati¢cation.These coarse, amalga-mated conglomerates are commonly capped by the 2- to3-m-thick ¢ne-grained sediments (Fig.9b and c).Thinnerconglomerate strata (few centimeters to decimeters inthickness) typify the stratigraphic section between �400and 580m in theTaxi He section (Fig. 6a).

Evidence of scouring or channelized £ow conditions,vague to less commonly distinct strati¢cation of conglom-erate deposits, and frequent alternation of thick beds ofpebble or cobble conglomerates with relatively thin ¢ne-grained deposits support the interpretation of an alluvialfan depositional environment since �5.9Ma in the TaxiHe section for these terrigenous deposits (Miall, 1978;Nemec&Steel,1984;DeCelles etal., 1991; Blair &McPher-son, 1994; Song et al., 2001; Sun et al., 2004, 2005a;Charreau et al., 2005; Heermance et al., 2007). These fanstrata can be further divided into proximal, medial, anddistal alluviumbased on grain size, rounding, sedimentarystructure, facies organization, and the nature of thecontacts between di¡erent lithologic layers (Rust, 1978;Nemec & Steel, 1984; Ren &Wang, 1985).The proximal al-luvial facies (Fig. 9a), equivalent to the deposits youngerthan �2.1Ma, is dominated by widespread cobble con-glomerate and uncommonly contains interlayers or lenti-cular beds (usually a few centimeters to decimeters inthickness) of siltstone and ¢ne-grained sandstone (Rust,1978;Ren&Wang,1985).This proximal alluvium is typi¢edby the coarsest grain size, subangular clasts, and massivebedding, despite the uncommon presence of crude or dis-tinct strati¢cation. The medial alluvial facies (Fig. 9b)displays a typical depositional rhythm of high-energybraided channels consisting of thick beds of cobbleconglomerate or coarse-grained sandstone interbeddedwith thin ¢ne-grained beds (Fig. 8c). The metric-scale,¢ning upward trends are interpreted to result fromwaningdischarge, typical of the deposits of many gravelly braidedrivers (Miall, 1978; Nemec & Steel, 1984; DeCelles, 1994).The multi- storey conglomerates are interpreted to repre-sent primary fan channels or moderately channelized£ood events related to nearby channels (Nemec & Steel,1984; DeCelles et al., 1991; Song et al., 2001; Heermance

Fig. 8. Typical depositional rhythms intheTaxi He section. (a) Thick beds of ¢ne-textured deposits intercalatedwithrelatively thin beds of coarse-graineddeposits in shallow lacustrineenvironment with distributary channels.(b) Alternation of thick siltstone andrelatively thin sandstone andconglomerate interpreted as distal alluvialdeposits. (c) Thick conglomerate bedsinterbeddedwith thin ¢ne-grained bedsin medial alluvial fan setting.



et al., 2007). The distal fan deposits (representing less-well channelized £oodplain deposits) merge with themarginal lacustrine facies and dominantly comprise¢ne-textured sediments intercalated with thin beds(20^50 cm) of coarse-grained sandstone and pebblyconglomerate (Ren&Wang,1985) (Figs 8b, 9d and e).Gyp-sum veins are present but uncommon in the distal fanfacies.

Overall, the depositional evolution of theTaxi He sec-tion since 8Ma can be divided into two distinct stages.Stage I ranges from �8.0 to �5.9Ma, includes the middleand upper Taxihe Formation, and comprises predomi-nantly ¢ne-textured beds of mudstone and muddy silt-stone that suggest a relatively low-energy environmentinterpreted as shallow lacustrine. Stage II ranges from�5.9 to �2.0Ma and is equivalent to the uppermostTaxi-he Formation, the Dushanzi Formation, and the lowestXiyu Formation.These sedimentary facies are interpretedto be typical of a prograding and upward coarsening

alluvial fan system, thus marking a signi¢cant shift fromthe previous lacustrine environment to a subsequentsubaerial environment. Overall, this succession of stagesresults in an overall coarsening-upward sequence in theTaxi He area that implies a progradational depositionalsystem in the northern Tian Shan foreland since at leastlateMiocene times.

Hierarchical sequencearrangement in theTaxiHe section

The method of Lˇ et al. (2006) was used to probe the hier-archical sequence of Taxi He lithostratigraphy. First,lithologies ofmudstone, siltstone, ¢ne sandstone,mediumsandstone, coarse sandstone, and conglomerate wereassigned values of 5, 10, 20, 25, 30, and 45, respectively.Second, in order to give the same weight to each strati-graphic increment, every measured bed was representedas multiple 10-cm-thick beds of the same lithology as the

Fig.9. Sedimentary characteristics in theTaxi He sectionwith paleomagnetic sampling sites (solid circles) and stratigraphic level. (a)Continuous conglomerates lying stratigraphically above �40m depth in the section uncommonly contain interlayers or lens of ¢ne-grained deposits. Black dashed lines denote the bedding of strata. Field view is �30m. (b) Strata in the upper Dushanzi Formationinterpreted to represent a medial alluvial fan environment. 3-m-thick siltstone lies stratigraphically above �10-m-thick conglomeratedeposits. (c) Clast- supported, subangular-subrounded, crudely strati¢ed and imbricated conglomerates in the medial alluvial fanenvironment at �635m depth. (d and e) Lower Dushanzi Formation interpreted as distal alluvial fan deposits, characterized by few-centimetre-to-decimeter-thick tabular and lenticular beds of coarse-grained deposits (usually sandy or pebbly conglomerate shownwith dashed line) intercalatedwithin thick siltstone deposits.


H. Luet al.

original bed. When plotted against stratigraphic depth,an arithmetic, 81-point moving average of these 10-cmincrements (equivalent to 8m of sediment) clearlyindicates the cyclic variations in lithology in theTaxi Hesection at several di¡erent temporal scales (Fig.10).

A ¢rst-order upward coarsening trend is displayed bythe vertical sequence of the whole depositional record intheTaxi He section (Fig.10), corresponding to the verticalsuperposition of shallow lacustrine facies followed by allu-vial deposits. Our magnetostratigraphic study assigns the¢rst-order sequence a minimum span of �6Myr (Fig.10).Similar to some previous studies (e.g. Burbank et al., 1988;DeCelles & Giles, 1996; DeCelles et al., 1998; Chen et al.,2002, 2007; Heermance et al., 2007), we interpret thislarge-scale coarsening-upward sequence at Taxi He to re-present a progradational depositional system controlledby progressive tectonic deformation that encroached on aforeland and drove lateral facies migration. In contrast,

Sun et al. (2004) interpreted the ¢rst occurrence of thickconglomerates as the direct indicator of initial, coevaltectonic uplift of Tian Shan. Such an interpretation seemsunlikely to be correct, however, because conglomeraticfacies progradation is time transgressive (e.g. Burbank etal., 1988; DeCelles & Giles, 1996; Chen et al., 2002, 2007;Charreau etal., 2005;Heermance etal., 2007), and time lagsexist among di¡erent types of sedimentary indicatorsof tectonism in alluvial foreland basin settings (Burbanket al., 1988; Jordan et al., 1988; Jones et al., 2004; Heermanceet al., 2007).

Encompassed within this overall trend, three second-order coarsening-upward sequences are interpreted to re-present meso-scale progradational processes (Fig. 10).Each second-order sequence spans a period of �1^3Myrbased on the magnetostratigraphy and ranges from �160to 560m in thickness (Fig. 10). Third-order intervals aremainly represented by the alternation of ¢ne-grained

Fig.10. Lithologic cycles of theTaxi Hesection. An arithmetic, 81-point movingaverage of grain size was used to de¢nelithologic cycles. See text for the detaileddescription. Ages shown are based onmagnestostratigraphic correlations inFig. 6. See Fig. 6 for lithologic symbols.



assemblages (usually mudstone and siltstone) and coarse-grained conglomerate packets.We identify 12 third-orderintervals in theTaxi He succession (Fig.10), each of whichhas a duration of �100^500 kyr.Generally, each conglom-erate packet is thicker than the associated ¢ne-grainedassemblage, excepting third-order sequences 1, 3, and 5(Fig. 10).The third-order cycles 3 and 5 are interpreted asrepresenting distal fan (£oodplain) deposits (Fig.10).

Beyond the above hierarchical sequence arrangement,smaller- scale lithologic variation is recorded by alterna-tion of thick conglomerates with thin beds of ¢ne-grainedsandstone or siltstone within coarse conglomerate packets(Fig. 10). This pattern is interpreted as the result of theabandonment and migration of braided channels (Nemec& Steel, 1984; DeCelles et al., 1991; Song et al., 2001; Heer-mance et al., 2007).

As described by Charreau et al. (2005) and Sun et al.(2004), the Dushanzi section (see Fig. 5 of Sun et al., 2004)and the Kuitun He (east) section (see Fig. 2 of Charreauet al., 2005) also show a large-scale coarsening upwardsequence during Late Cenozoic times that is similar tothat of theTaxi He section. Although separated by up to�130 km along strike in the Junggar foreland, all thesesections record similar stages of depositional evolution: aprogradational depositional succession that evolves fromolder lacustrine facies to a younger alluvial environment(Sun et al., 2004; Charreau et al., 2005; this study). Thus,the establishment of the ¢rst-order upward coarsening se-quence and its correlation in the northern ChineseTian Shan foreland can be easily recognized. In contrast,because of the transverse variation in depositional systemsalong the north piedmont of the Tian Shan that resultsfrom both the spacing of individual river systems and theactivity on numerous discrete structures within the fore-land, second- and third-order sequences are unlikely tocorrelate across the entire foreland.

DISCUSSION

Correlations among northTian Shanmagnetostratigraphic sections

Over the past several years, Charreau et al. (2005) and Sunet al. (2004, 2007) published new magnetostratigraphiesfrom three closely spaced sections near Dushanzi (Fig.1b): one is located on the eastern end of theDushanzi anti-cline (Sun et al., 2004), the other two lie �5 km fartherwest along the east and west bank of Kuitun He where it£ows across this fold. These three sections lie �130 kmwest of theTaxi He section (Fig. 1b).We refer to them asthe Kuitun He (west) section (Sun et al., 2007), the KuitunHe (east) section (Charreau et al., 2005), and the Dushanzisection (Sun et al., 2004), respectively (Fig. 11). Becausethey each span overlapping parts of the upper Cenozoicsuccession, we have correlated their reversal patterns witheach other and to theTaxi He section based on lithostrati-graphy and reversal patterns (Fig.11).Here, we support there-correlations of the Dushanzi section (Sun et al., 2004)

and the Kuitun He (west) section (Sun et al., 2007) as pro-posed by Charreau et al. (2005, Fig. 6; 2008a, Fig. 2) for thefollowing reasons. According to the lithologic logs (Fig. 2,Charreau et al., 2005; Fig. 5, Sun et al., 2004; Fig. 8, Sun etal., 2007) and the description on paleomagnetic sampling,we infer that they both identify the base of the Xiyu con-glomerate at about the same stratigraphic level within theDushanzi anticline with respect to the onset of very con-tinuous conglomeratic facies. Accordingly, the basal agesof the Xiyu conglomerate in these three nearby sectionsshould approximate each other. As originally published,however, these ages di¡ered by 42Myr: 2.6Ma in theDushanzi and Kuitun He (west) sections vs. 4.8Ma in theKuitunHe (east) section.We argue that the Xiyu conglom-erate is unlikely to be so starkly diachronous across a shortdistance parallel to strike. We consider Charreau et al.’s(2005) overall correlation to be more convincing due tothe longer time spanned by their section and the need forfewer abrupt, large changes in accumulation rates in com-parison with Sun et al.’s (2004, 2007) correlations. Notsurprisingly perhaps, this recorrelation yields a more con-sistent basal age for the Xiyu conglomerate in thesenearby sections. Because correlations to theGilbert chron(C2Ar-C3r: Fig. 11) are somewhat ambiguous in thesemagnetostratigraphies, the Xiyu conglomerate’s basal agestill retains some uncertainty. Despite this, the onset ofXiyu deposition (Fig.11) is likely to be �0.5Myr youngerat Dushanzi ( �4.2� 0.2Ma) than in the Kuitun Hesections ( �4.7 � 0.2Ma) that are �5 km away.

Timing, magnitude, and patterns of forelanddeformation

North of the Junggar Frontal Thrust, three elongate zonesof sub-parallel folds (Fig. 1b) record the progressive en-croachment of deformation into the foreland during theNeogene. Growth strata, limb rotation, and angular un-conformities in the anticlines of the three east^west trend-ing fault-and-fold zones (Fig. 12) provide evidence forprolonged, but episodic deformationwithin theTianShan.The Qigu anticline is located in fault-and-fold zone I(Fig. 1b) and comprises Jurassic, Cretaceous, and Tertiarystrata (Fig. 12c). The unconformable contact between theupper Jurassic and the lower Cretaceous strata indicatessigni¢cant deformation across the region during LateMe-sozoic times (Hendrix et al., 1992; Burch¢el et al., 1999;Deng et al., 2000). The generally parallel dips and con-formable contacts of the Cretaceous through Neogenestrata on the northern limb of this fold (Fig.12c), however,suggest that, for more than100Myr before the lateTertiary,no signi¢cant deformation disrupted deposition (Bullen etal., 2001, 2003; Sobel etal., 2006). Folding ofNeogene stra-ta in Qigu’s northern limb indicates renewed deformationsometime in the last 20Myr, but no precise limits can beplaced on its initiation in this fold. Further west in zone I,on the northern limb of theTuostai anticline, geometricconstraints and seismic interpretations require abruptthinning of Pliocene strata onto the fold (Fig.12a), thereby


H. Luet al.

implying Pliocene fold growth. Here, it appears that fold-ing continued through deposition of the Xiyu conglomer-ate, which is also deformed on the northern limb of thefold.UndeformedLateQuaternary terraces ofHutubiRiv-er cutting the Qigu anticline (Deng et al., 2000) imply thatthis anticline is currently inactive.

The best-dated structure in the medial deformed fore-land zone (fault-and-fold zone II) is theTugulu anticline.The topographic expression of this fold strikes approxi-mately east^west, and is �50 km long and �8 km wide(Fig. 2a). As a north-vergent fault-propagation fold(Burch¢el et al., 1999), the asymmetric Tugulu anticlinedisplays steeper strata on its forelimb (Fig. 3), where themeasured Taxi He section reveals a systematic decrease indip from nearly vertical beds near the core of the fold to�451 at the top of the section (Fig. 3a). A unconformityoccurs at �350m above the base of the section (labeledwith the number ‘1’, Fig. 3a) where depositional onlapacross a101 discordance in dip is interpreted as represent-ing the initial deposition of growth strata (Fig. 3b) whenfolding of theTugulu anticline commenced. Over the next

�150m of strata, the dip progressively decreases by an-other 131: an observation consistent with growth strata,rather than with simple dip panels (Suppe et al., 1992). Incontrast, some other dip changes, for example, at 790mheight (labeled ‘3’ in Fig. 3a), discretely separate zones ofquite uniform dips, suggesting folding across axial sur-faces. Based on the position of the growth strata withinour magnetostratigraphy, folding of the Tugulu anticlinecommenced at �6.1Ma (Figs 3 and6).The youngest stratain our section date from �2Ma and dip at 451, indicatingthat over half of the limb rotation has occurred since 2Ma.Late Pleistocene terraces are deformed across theTuguluanticline (Burch¢el et al., 1999). Hence, the growth of theTugulu fold has spanned �6Myr. In the absence of seis-mic data to illuminate stratal geometries in the subsurface,we note that our interpretation of growth strata is based onthe bedding^dip relationships and depositional onlap de-scribed above.The lack of subsurface data for this fold pre-cludes a truly unambiguous interpretation.

The Dushanzi anticline in fault-and-fold zone III ex-poses folded Neogene^Pleistocene strata (Fig. 12b). Here,

Fig.11. Correlations amongMPSs of theKuitunHe (west) section (Sun et al., 2007), theKuitunHe (east) section (Charreau et al., 2005),theDushanzi section (Sun etal., 2004) and theTaxiHe section (this study).The basal ages of theXiyu conglomerate in these four sectionsalso are shown adjacent to the corresponding magnetic columns.



a middle Pleistocene unconformably truncates the lowerPleistocene Xiyu conglomeratic (Fig. 12b; Deng et al.,2000), indicating signi¢cant folding of the Dushanzi anti-cline sometime in Early to Middle Pleistocene times(Deng et al., 2000; Fu et al., 2003). As at Tugulu, deformed£ights of terraces indicate that folding has continued untilpresent (Molnar et al., 1994).To the north of the tradition-ally delineated fault-and-fold zone III, Pleistocene growthstrata that are revealed by seismic lines (Xu etal.,1992) helpde¢ne the presently active Xihu anticline (Fig. 12a): theoutermost fold currently developing in the north pied-mont ofChineseTianShan (Fig.1b). Because foreland sub-sidence and sediment-accumulation rates have outpacedthe rate of vertical anticlinal growth, the Xihu anticlinehas little to no surface expression (only 10m topographicrelief). Nonetheless, �500m of structural relief has de-veloped in the subsurface according to the seismic section(Fig.12a; Xu et al., 1992).

The post-Triassic detrital strata in the Junggar forelandbasin taper progressively northward from a thickness of6^9 km near the Junggar frontal thrust to 4^5 km atthe northern limit of deformation (Fig. 13). At the base ofthe Jurassic, a detachment horizon is interpreted to haveaccommodated most of the fault-localized slip as defor-mation stepped out into the foreland (Fig.13). In zone I, alarge fault-bend fold (Figs 12 and 13) accounts for most ofthe shortening in the innermost fold, although severalthrusts cause localized folding above the basal detach-

ment. Farther north, thrust faults are interpreted to rampfrom the basal Jurassic detachment toward the surfacewhere they modulate the deformation of the well- studiedsuite of fault-propagation and fault-tip folds in zones IIand III (Deng etal.,1991, 2000;Avouac etal.,1993; Burch¢elet al., 1999; Dae« ron et al., 2007).

Shortening estimates (Table 1) have previously beencalculated for most of the anticlines in the northernTianShan foreland. For the Nanmanasi anticline (whichwe didnot map) in the east part of zone I, we tentatively estimate ashortening of �4^6 km (Table1), on the basis of the short-ening magnitude of other anticlines in fault-and-fold zoneI. If this estimate is acceptable, then when combined withprevious shortening estimates (Deng et al., 1991, 2000;Molnar et al., 1994; Yang et al., 1996; Burch¢el et al., 1999;Dae« ron et al., 2007; Charreau et al., 2008b), these data(Table 1) permit us to examine the regional patterns ofshortening along and across strike in the foreland (Fig.14). Overall, the total magnitude of Neogene shorteningacross the foreland ranges from �8 to �15 km, butwithineach fault-and-fold zone, the magnitude of shorteningvaries along strike. In fault-and-fold zone II, for example,adjacent anticlines show �6^10 km of shortening in theeastern two-thirds of the study area, whereas the magni-tude of shortening markedly diminishes to the west (Figs12a and 14). In zone III, active fold growth in the westaccommodates �1^3 km of shortening, whereas no fold-ing is recorded in this zone farther to the east (Figs1b and

Fig.12. Structural cross sections depicting deformation and cross-cutting relationships in the northernTian Shan foreland. Forsection locations, see Fig.1b. (a) Line drawing of seismic pro¢le through theTuostai and Xihu anticlines (modi¢ed fromXu et al., 1992).Growth strata are identi¢ed in the Pliocene-Quaternary deposits of Xihu and Tuostai anticlines. J, K, E, N1, N2, Q are the Jurassic,Cretaceous, Paleogene,Miocene, Pliocene and Pleistocene, respectively. (b) Geological section of the Dushanzi anticline along theKuitunHe valley (modi¢ed from Burch¢el et al., 1999). Middle Pleistocene strata (Q2) unconformably overlie lower Pleistocene strata(Q1) (Deng et al., 2000), indicating Early Pleistocene deformation of the Dushanzi anticline. (c) Geological section of the Qigu anticlinealong theHutubiHe (modi¢ed fromBurch¢el etal.,1999). An anticline-syncline pair lies north of a gentlyN-dipping section of Jurassicstrata.Unconformity between the upper Jurassic (J3) and lower Cretaceous (K1) strata impliesMesozoic deformation atQigu (Burch¢elet al., 1999; Deng et al., 2000). Uniform steep north dips (4601) and comformable contacts of Cretaceous (K) throughNeogene (E)strata, however, suggest tectonic quiescence until the lateTertiary. Folding postdates deposition of Neogene strata preserved in thenorthern limb of the anticline.


H. Luet al.

14). The locus of maximum shortening switches betweenzones I and II along strike, but never occurs in zone III,where the duration and intensity of folding are the least.The general consistency of total shortening across theforeland suggests that tectonic forcing was spatially uni-form during Late Cenozoic times along this �130-km-long swath of the northern Tian Shan foreland.Whetherthat shortening was absorbed primarily in the most proxi-mal foreland structures (zone I) or in more distal folds

(zone II in the east) probably depended on how shorteningwas fed into the detachment along strike andwhether pre-existing structures in theTriassic and older bedrock werereactivated during the Neogene.

Despite shortening estimates for numerous individualanticlines (Table 1), calculation of long-term shorteningrates for all these folds is di⁄cult, because the time of in-itiation of deformation is poorlyknown.Based on assump-tions that the Xiyu conglomerate could be treated as achronostratigraphic unit with a basal age of �2.5Mathroughout the Tian Shan and that deformation begansome time after initial Xiyu deposition, Burch¢el et al.(1999) calculated the shortening rates for several folds inboth the north and south £anks of theTian Shan. Thesesimplifying assumptions, however, are violated both bythe strongly diachronous base of the Xiyu conglomeratein this region (Fig. 11) and by more recent interpretationsplacing the start of deformation signi¢cantly before theonset of conglomeratic deposition for some folds. For theTugulu anticline, for example, we calculate a shorteningrate using a �6.0-Ma age for the initiation of foldingbased on our dated growth strata (Fig. 6).When combinedwith a total shortening of �5.5^6.0 km estimated byBurch¢el et al. (1999) and Yang et al. (1996), this age yieldsan average rate of �0.9^1.0mmyr�1. Within the datedsection at Dushanzi anticline (Charreau et al., 2005), thesteadiness of sediment accumulation and the absence ofrecognized growth strata suggest that folding there beganafter 3Ma, thereby providing a lower limit on the shorten-ing rate of � 0.7mmyr�1. Using a fault-tip fold growthmodel of the Anjihai anticline and magnetostratigraphicages (Charreau etal., 2005) extrapolated from theDushanzifold ( �45 km to the west), Dae« ron et al. (2007) estimatedthat, deformation began at Anjihai at �7Ma. Based on

Fig.13. Generalized geological cross-section of the northern ChineseTian Shan foreland showing major thrust faults, detachments,and fault-and-fold zones.The total stratigraphic thickness ofMesozoic and Cenozoic strata is �7^12 km (Cenozoic: �3^5 km thick),and fault ramps underlying the anticlines in each deformed zone dip �20^601 (Xu etal., 1992; Burch¢el etal., 1999;Deng etal., 2000). Amajor detachment horizon is interpreted as a lower Jurassic composite of shale, mudstone, and coal measures that attains �600m inthickness, is located at between �7 and 9 km depth (Deng et al., 2000), and converges southwardwith the Junggar Frontal Thrust.Shortening is estimated for anticlines in each fault-and-fold zone (Table1). Interpreted ages for initiation of folding within each zonesuggest progressive migration of tectonic deformation northward into the foreland.

Table1. Minimum shortening estimates of anticlines across thefault-and-fold zones of the northernTian Shan foreland

Fault-and-fold zoneShorteningestimates (km) References

ITuostai anticline �4.6^5.0 Burch¢el et al. (1999)Nananjihai anticline 44.2 Deng et al. (2000)Nanmanasi anticline �5 This studyQigu anticline 6.2 Burch¢el et al. (1999)

5.5 Deng et al. (2000)IIHuoerguosi anticline �10 Charreau et al. (2008b)Manasi anticline �6.5 Deng et al. (2000)Tugulu anticline 6.0 Yang et al. (1996)

5.5 Burch¢el et al. (1999)Molnar et al. (1994)

IIIDushanzi anticline 2.12^2.35 Burch¢el et al. (1999)

Molnar et al. (1994)2.9 Deng et al. (2000)

Anjihai anticline 1.5 Dae« ron et al. (2007)Deng et al. (1991)



observed geomorphic and structural constraints at thesurface, they further argue that, despite a mean long-termshortening rate of �0.2mmyr�1, rates have recently ac-celerated to �1mmyr�1 andwere only �0.1mmyr�1 formost of the Anjihai’s shortening history. The apparentdepth of growth strata in the Xihu seismic data (Fig. 12a),its seismic similarity to the Anjihai (Dae« ron et al., 2007),the absence of a signi¢cant topographic expression of theXihu fold, and its distal position with respect to theDushanzi fold where Mio-Pliocene sedimentation rateswere �0.2mmyr�1 argue for sustained slow growth ofthe Xihu fold, analogous to that at Anjihai anticline.

Whereas we now have substantial insight on shorteningrateswithin fault-and-fold zones II and III in the northernTian Shan foreland, zone I is only broadly dated. Betterunderstanding of the shortening history across the fore-land requires signi¢cantly improved chronological con-straints on deformation of these undated folds.

Structural controls on depositional evolutionand accumulation rates

Sedimentation patterns in foreland basins are sensitive tothe distance to the front of the orogenicwedge, to the char-acter of the rocks that are being raised above base levelwithin the foreland due to ongoing deformation, and to re-organization of sediment distribution systems by growingfolds and active faults (e.g. Colombo, 1994; Burbank et al.,1996; DeCelles & Giles, 1996; Tucker & Slingerland, 1996;DeCelles et al., 1998; Chen et al., 2002, 2007; Pare' s et al.,2003; Charreau et al., 2005; Heermance et al., 2007).Typically, proximal coarse-grained deposits are depositedcoevally with distal, ¢ner-grained deposits within asubsiding foreland (Heller et al., 1988; Charreau et al.,2005; Heermance et al., 2007). The progressive encroach-ment of deformation into the foreland drives a tendencyfor deposition at anygiven site to coarsen upwards throughtime, as the entire alluvial system progrades outward intothe foreland basin. During this process, older, more prox-imal strata become gradually deformed both at and behindthe leading edge of encroaching structural disruption,

whereas growing folds commonly de£ect £uvial channelsand cause lateral shifts in the loci of deposition (Heller &Paola, 1992; Burbank et al., 1996, 1999; DeCelles & Giles,1996). Structural encroachment into a foreland is alsoexpected to drive an overall acceleration in the rate of sedi-ment accumulation as the locus of tectonic loadingapproaches any given site. As seen in many forelands, how-ever, tectonic encroachment is not a steady, incrementalprocess, but rather consists of episodic forward steps asolder, more proximal faults and folds are abandoned andnew, more distal ones are formed. This step-wise en-croachment should generate coeval accelerations in sedi-ment-accumulation rates in front of newly developingtectonic loads.

In theTaxi He area, a ¢rst-order upward coarsening cy-cle is de¢ned by the lacustrine depositional environmentthat dominates until �6Ma and is succeeded by strata de-posited bybraided rivers and alluvial fans (Fig.10).Upwardcoarsening also characterizes the foreland sections in theDushanzi area (Sun et al., 2004; Charreau et al., 2005) andis consistent with the expectation of a clastic wedge pro-grading from the deformed hinterland. More detailed in-sight on tectonic controls emerges from analysis of thesediment-accumulation rate curve from theTaxi He sec-tion (Fig. 7). The magnitude of the acceleration in sedi-ment-accumulation rate at 4Ma is striking: a 450%increase from �160 to �250mMyr�1 (Fig. 7). Its timingis consistent withwidespread observations of Pliocene se-dimentation-rate increases that have been attributed toclimate change in which both the sediment £ux and themean grain size increase (Zhang et al., 2001; Molnar,2004). In contrast, at Taxi He a striking decrease in grainsize occurs at the base of the second-order upward-coar-sening cycle at �4Ma as accumulation rates increase(Fig. 10).We interpret this abrupt ¢ning to be a responseto accelerated subsidence and temporary retraction of thegravel front due to enhanced tectonic loading (Burbank,1992). We suggest that the absence of a signi¢cantly in-creased £ux of coarse sediment, despite an increased rateof sediment accumulation, is more consistent with tec-tonic, rather than climatic, forcing. Before 4Ma, our data

Fig.14. Shortening estimates ofanticlines in the northernTian Shanforeland.Total magnitude of shorteningacross the foreland ranges from �8 to15 km, whereas the magnitude ofshortening varies within each fault-and-fold zone along strike.The data are fromTable1.


H. Luet al.

indicate steady rates of sediment accumulation (Fig. 7).From this steadiness, we infer that, although folding hadbegun earlier in zone I (Qigu to Tuostai anticlines, Fig.1b), initial stages of deformation and rock uplift had mini-mal impact on crustal thickness and tectonic loads becausethe uplifted foreland strata would have been readilyeroded, such that little new topography or crustal thicken-ing would result (Burbank&Beck,1988;Molnar etal.,1993;Burbank etal.,1999; Sobel etal., 2006).Onlywhen the moreerosion-resistant, pre-Jurassic bedrock emerged at thesurface of the growing folds would signi¢cant thickeningbegin to drive enhanced foreland subsidence. Our datasuggest that this emergence occurred at �4Ma.

Asynchronous Xiyu conglomerate deposits innorthwest China

The conglomeratic Xiyu Formation is widely developedalongboth the northern and southern piedmonts of theChi-neseTian Shan (Liu etal., 1996; Fig.15). Further south it alsois found on the northern piedmonts of the Kunlun Moun-tains and the QilianMountains (Fig.15).This formation has

been widely interpreted as marking an important tectonicand/or climatic event (Li et al., 1979; Huang & Cheng, 1981;Liu et al., 1996; Zheng et al., 2000; Sun et al., 2004, 2005a)and has commonly been viewed as a chronostratigraphicunit (Li et al., 1979; Huang & Cheng, 1981). As demonstratedby several previous studies from theTian Shan (Chen et al.,2002; Charreau etal., 2005;Heermance etal., 2007), however,the onslaught of this facies is asynchronous.The magnetos-tratigraphic ages of the Xiyu conglomerate throughoutnorthwest China (Table 2 and Fig. 15) unambiguously indi-cate its time-transgressive nature. In the piedmont subsi-dence zones on both sides of theTian Shan, the basal Xiyuconglomerates range from middle Miocene (415Ma) tomiddle Pleistocene (o1Ma) in age (Chen et al., 2002, 2007;Sun et al., 2004; Charreau et al., 2005; Heermance et al.,2007; this study), whereas farther south, basal ages rangefrom �3.5 to 3.0Ma on the northern £ank of the KunlunMts. (Zheng et al., 2000; Sun & Liu, 2006). Such variationsin age clearly indicate that the Xiyu conglomerate in north-west China should only be considered as a lithostratigraphicunit that lacks temporal implications beyond the local areafor which its age has been independently de¢ned.

Fig.15. Distribution of the foreland Xiyuconglomerate deposits in northwest China(after Liu et al., 1996).Themagnetostratigraphic ages of the basalXiyu deposition indicate its diachronousnature.The age data are fromTable 2.

Table2. Magnetostratigraphic ages of the basal Xiyu conglomerates in northwest China

Studied area Studied sections Basal age of Xiyu Fm. (Ma) References

Northern ChineseTian Shan foreland KuitunHe (west) section 4.7 � 0.2 (reinterpreted) Sun et al. (2007)KuitunHe (east) section 4.7 � 0.2 Charreau et al. (2005)Dushanzi section 4.2 � 0.2 (reinterpreted) Sun et al. (2004)Taxi He section �2.1 This study

Southwestern ChineseTian Shan foreland Ganhangou section o1.0 Chen et al. (2002)Boguzihe section 1.9 Chen et al. (2002)Northern Kashi foreland 15.5 Heermance et al. (2007)Middle Kashi foreland 8.6 Heermance et al. (2007)Southern Kashi foreland 1.9, 1.4, 0.7 Chen et al. (2007)

North piedmont of the KunlunMts. Sanju section 3.0 Sun & Liu (2006)Yecheng section 3.5 Zheng et al., 2000



CONCLUSION

Our paleomagnetic investigation provides a new, detailedchronology for the upper Cenozoic terrigenous sequencein theTugulu anticline of the northernTianShan foreland.Our magnetostratigraphic correlation indicates that the�1200-m-thick Taxi He section spans from �8.1 to�2.0Ma. Although sediment-accumulation rates average�200mMyr�1over this 6-Myr interval, they abruptly in-crease by450% at �4Ma.We interpret this increase toresult from accelerated subsidence that was, in turn, dri-ven by enhanced rates of tectonic loading. Whereas slipon faults in the hinterland and within the proximal fore-land undoubtedly began before 4Ma, we speculate thatcontinuing deformation brought more resistant rocks tothe surface at this time, driving accelerated rates of surfaceuplift, crustal thickening, and tectonic loading. Initialgrowth of theTugulu fold is interpreted from growth stratadated at �6Ma.Over1km of conglomerates dipping up to451 overlie our dated section, suggesting that major topo-graphic emergence of this fold occurredwell after �2Ma.Based on this new magnetostratigraphy, the basal age ofthe Xiyu conglomerate in theTaxi He section is �2.1Ma:younger than the previous reported ages of 2.6 and 4.8Mafor foreland folds about 130 km to the west.Whereas someprevious studies interpreted the onset of massive con-glomeratic deposition at diverse sites within the forelandas indicative of tectonic uplift of theTian Shan at varioustimes between 7 and 1Ma, a synthesis of chronologic datafor the Xiyu conglomerate clearly demonstrates its time-transgressive nature and reduces its utility for predictingtime-speci¢c tectonic events in the hinterland based onthe site- speci¢c age of foreland conglomerates. Balancedcross sections and inferred stratigraphic ages indicate netshortening of 1.5^10 kmwithin the medial to distal north-ernTian Shan foreland at rates of 0.2^1mmyr�1 over thepast 6^7My. Although better age constraints are neededon structures in the proximal foreland, seismic re£ectionlines and previous work suggest that deformation beganearlier in these more proximal sites. The subsequentnorthward migration of deformation drove the prograda-tion of the coarsening-upward alluvial system that iswidely preserved across the northernTian Shan foreland.In the vicinity of theTaxiHe section, the overarching, ¢rst-order depositional sequence comprises an initial shallowlacustrine environment that was succeeded by alluvial fandeposition after �5.9Ma. Enhanced understanding of thedynamics of the northern Tian Shan’s depositional andtectonic evolution awaits more detailed and densely spacedchronostratigraphic studies.

ACKNOWLEDGEMENTS

HonghuaLugives special appreciation toChinaScholarshipCouncilwhich supported him to study atUniversity of Cali-fornia at Santa Barbara.We thank Dongjiang Sun, Supei Si,Hongzhuang Zhao andMangQian for their ¢eld assistance.We are especially grateful toZhengyuYang for the use of and

assistance in the Paleomagnetic Laboratory of the InstituteofGeomechanics, Chinese Academy ofGeological SciencesinBeijing, and toLinfengShi,XiaodiZhou, and Jianli Fu fortheir assistance in the lab.This paper bene¢ted from discus-sions with Zhengkai Xia, Mingda Ren, Jingchun Yang andJosephGoode.We especially thankRichardLease for his as-sistance in the analysis of the paleomagnetic data.Thought-ful and thorough reviews by G. Dupon-Nivet, J. Charreau,another anonymous reviewer, and editorP. van derBeekpro-foundly improved this manuscript. This study was ¢nan-cially supported by National Natural Science Foundation ofChina (Grant 40571013) and by the US National ScienceFoundation (EAR 0230403).

REFERENCES

Abdrakhmatov,K.Y.,Aldazhanov, S.A.,Hager, B.H.,Ham-burger, M.W., Herring,T.A., Kalabaev, K.B., Makarov,V.I., Molnar, P., Panasyuk, S.V., Prilepin, M.T., Reilin-ger, R.E., Sadybakasov, I.S., Souter, B.J., Trapeznikov,Y.A.,Tsurkov,V.Y. & Zubovich, A.V. (1996) Relatively recentconstruction of the Tien Shan inferred from GPS measure-ments of present-day crustal deformation rates. Nature, 384,450^453, doi: 10.1038/384450a0.

Allen, M.B.,Windley, B.F., Zhang, C. & Guo, J.H. (1993)Evolution of theTurfan Basin, Chinese Central Asia.Tectonics,12(4), 889^896.

Avouac, J.-P. & Tapponnier, P. (1993) Kinematic model of ac-tive deformation in central Asia. Geophys. Res. Lett., 20(10),895^898.

Avouac, J.-P., Tapponnier, P., Bai, P., You, M. & Wang, G.(1993) Active thrusting and folding along the northern TienShan and late Cenozoic rotation of theTarim relative toDzun-garia and Kazakhstan. J. Geophys. Res., 98(B4), 6755^6804.

Blair,T.C. & McPherson, J.G. (1994) Alluvial fans and theirnatural distinction from rivers based onmorphology, hydraulicprocesses, sedimentary processes and facies assemblages.J. Sedim. Res., 64(3a), 450^489.

Bullen,M.E., Burbank,D.W.,Garver, J. & Abdrakhmatov,K.Y. (2001) Late Cenozoic tectonic evolution of the northwes-ternTien Shan: new age estimates for the initiation of moun-tain building.Geol. Soc. Am. Bull., 113(12), 1544^1559.

Bullen,M.E.,Burbank,D.W.&Garver, J. (2003) Building thenorthernTien Shan: integrated thermal, structural, and topo-graphic constraints. J. Geol., 111, 149^165.

Burbank, D.W. (1992) Causes of recent Himalayan uplift de-duced from deposited patterns in the Ganges basin. Nature,357, 680^683, doi: 10.1038/357680a0.

Burbank, D.W. & Beck, R.A. (1988) Rapid, long-term rates ofdenudation.Geology, 19, 1169^1172.

Burbank,D.W., Beck, R.A., Raynolds, R.G.H.,Hobbs, R. &Tahirkheli, R.A.K. (1988) Thrusting and gravel prograda-tion in foreland basins: a test of post-thrusting gravel disper-sal.Geology, 16(2), 1143^1146.

Burbank,D.W.,McLean, J.K., Bullen,M., Abdrakhmatov,K.Y.&Miller,M.M. (1999)Partitioning of intermontane ba-sins by thrust-related folding, Tien Shan, Kyrgyzstan. BasinRes., 11(1), 75^92.

Burbank, D.W.,Meigs, A. & Brozovic, N. (1996) Interactionsof growing folds and coeval depositional systems.Basin Res., 8,199^223.


H. Luet al.

Burchfiel, B.C., Brown, E.T., Deng, Q., Feng, X., Li, J.,Molnar,P.,Shi, J.,Wu,Z.&You,H. (1999) Crustal shorten-ing on the margins of theTien Shan,Xinjiang, China. Int.Geol.Rev., 41(8), 665^700.

Bureau of Geological and Mineral Resources of theXinjiang Uygur Autonomous Region. (1993) RegionalGeology of Xinjiang Uygur Antonomous Region. Geology Press,Beijing, 841pp (in Chinese).

Burtman,V.S. (1975) Structural geology of the Variscan TienShan.Am. J. Sci., 280, 725^744.

Butler,R.L. (1992)Paleomagnetism. Blackwell Scienti¢cPublica-tions, Cambridge,MA.

Carroll, A.R., Graham, S.A., Hendrix, M.S., Ying, D. &Zhou, D. (1995) Late Paleozoic tectonic amalgamation ofnorthwestern China: sedimentary record of the northernTarim, northwestern Turpan, and southern Junggar basins.Geol. Soc. Am. Bull., 107, 571^594.

Charreau, J., Avouac, J.-P., Chen, Y., Dominguez, S. &Gilder, S. (2008b) Miocene to present kinematics offault-bend folding across the Huerguosi anticline, northernTianshan (China), derived from structural, seismic, and mag-netostratigraphic data.Geology, 36(11), 871^874.

Charreau, J., Chen,Y.,Gilder, S.,Dominguez, S., Avouac,J.-P., Sen, S., Sun,D.J., Li,Y.A. &Wang,W.M. (2005) Mag-netostratigraphy and rock magnetism of the Neogene KuitunHe section (northwest China): implications for Late Cenozoicuplift of theTianshan mountains. Earth Planet. Sci. Lett., 230,177^192.

Charreau, J.,Gilder, S., Chen,Y.,Dominguez, S., Avouac,J.-P., Sen, S., Jolivet, M., Li, Y.A. & Wang,W.M. (2006)Magnetostratigraphy of theYaha section,TarimBasin (China):11Ma acceleration in erosion and uplift of the Tian ShanMountains.Geology, 34(3), 181^184.

Charreau, J., Sun, J., Chen, Y., Gilder, S., Huang, B. &Wang, Q. (2008a) Addendum to ‘‘Late Cenozoic magneto-chronology and paleoenvironmental changes in the northernforeland basin of the Tian Shan Mountains’’ by Jimin Sun,Qinghai Xu, and Baochun Huang. J. Geophys. Res., 113,B06103, doi: 10.1029/2007JB005489.

Chen, J., Burbank,D.W., Scharer,K.M., Sobel, E., Jinhui,Y., Rubin, C. & Ruibin, Z. (2002) Magnetochronology of theUpper Cenozoic strata in the Southern Chinese Tian Shan:rates of Pleistocene folding and thrusting. Earth Planet. Sci.Lett., 195, 113^130.

Chen, J., Heermance, R., Burbank, D.W., Scharer, K.M.,Miao, J.J. &Wang, C.S. (2007) Quanti¢cation of growth andlateral propagation of the Kashi anticline, southwest ChineseTian Shan. J. Geophys. Res., 112, doi: 10.1029/2006JB004345.

Colombo, F. (1994) Normal and reverse unroo¢ng sequences insyntectonic conglomerates as evidence of progressive basin-ward deformation.Geology, 22(3), 235^238.

Dae« ron,M.,Avouac, J.-P.&Charreau, J. (2007)Modeling theshortening history of a fault tip fold using structural and geo-morphic records of deformation. J. Geophys. Res., 112, B03S13,doi: 10.1029/2006JB004460.

DeCelles, P.G. (1994) Late Cretaceous-Paleocene synorogenicsedimentation and kinematic history of the Sevier thrust belt,northeast Utah and southwest Wyoming. Geol. Soc. Am. Bull.,106, 32^56.

DeCelles, P.G.,Gehrels,G.E.,Quade, J. &OJHA, P. (1998)Eocene-early Miocene foreland basin development and thehistory of Himalayan thrusting, western and central Nepal.Tectonics, 17(5), 741^765.

DeCelles, P.G. & Giles, K.A. (1996) Foreland basin systems.Basin Res., 8, 105^123.

DeCelles, P.G.,Gray,M.B.,Ridgway,K.D.,Cole,R.B., Piv-nik, D.A., Pequera, N. & Srivastava, P. (1991) Controls onsynorogenic alluvial-fan architecture, Beartooth Conglomer-ate (Palaeocene), Wyoming and Montana. Sedimentology, 38,567^590.

Deng, Q.D., Feng, X.Y., You, H.C., Zhang, P.Z., Zhang, Y.,Li, J.,Wu, Z.M.,Xu,X.W.,Yang,X.P. & Zhang,H.W. (1991)The characteristic of deformation and mechanism ofDushan-zi- Anjihai active reverse fault-fold zone in the northern £ankof theTian ShanMountains. In:Research of Active Fault (I) (Ed.byEditing Committee of the Research of Active Fault)pp. 17^36. Seismology Press, Beijing (in Chinese, Englishabstract).

Deng, Q.D., Feng, X.Y., Zhang, P.Z., Xu, X.W., Yang, X.P.,Peng,S.Z.&Li, J. (2000)ActiveTectonics of theTian ShanMoun-tains. Seismology Press, Beijing, 399pp (in Chinese).

Dumitru, T.A., Zhou, D., Chang, E.Z., Graham, S.A.,Hendrix,M.S., Sobel, E.R.&Carroll,A.R. (2001) Uplift,exhumation, and deformation in the Chinese Tian Shan. In:Paleozoic and Mesozoic Tectonic Evolution of Central and EasternAsia: From Continental Assembly to Intracontinental Deformation(194) (Ed. by M.S. Hendrix & G.A. Davis), Geol. Soc. Am.Memoir, 71^99.

Dupont-Nivet,G.,Krijgsman,W.,Langereis,C.G.,Abels,H.A.,Dai, S. & Fang, X. (2007) Tibetan plateau aridi¢cationlinked to global cooling at the Eocene-Oligocene transition.Nature, 445, 635^638.

Editing Committee of the Stratigraphy of China. (1999)Stratigraphy in China (Teritary). Geology Press, Beijing, 163pp(in Chinese).

Fu, B., Lin, A., Kano, K.-I., Maruyama, T., Guo, J. & Ab-drakhmatov, K.Y. (2003) Quaternary folding of the easternTian Shan, northern China.Tectonophysics, 369, 79^101.

Gansser,A. (1981) The geodynamic history of theHimalaya. In:Zagros, Hindu Kush, Himalaya Geodynamic Evolution (3) (Ed. byH.K. Gupta & F.M. Delany), pp.111^121. American Geophysi-cal Union,Washington, DC.

Heermance, R.V., Chen, J., Burbank, D.W. & Wang, C.S.(2007) Chronology and tectonic controls of LateTertiary de-position in the southwesternTian Shan foreland, NW China.Basin Res., 19, 599^632.

Heller, P.L., Angevine, C.L.,Winslow, N.S. & Paola, C.(1988) Two-phase stratrgraphic model of foreland-basin se-quences.Geology, 16, 501^504.

Heller, P.L. & Paola, C. (1992) The large-scale dynamics ofgrain-size variation in alluvial basins, 2: application to syntec-tonic conglomerate.Basin Res., 4, 91^102.

Hendrix, M.S., Dumitru, T.A. & Graham, S.A. (1994) LateOligocene-earlyMiocene unroo¢ng in the ChineseTian shan:an early e¡ect of the India-Asia collision.Geology, 22, 487^490.

Hendrix, M.S., Graham, A.S., Carroll, A.R., Sobel, E.R.,McKnight, C.L., Schulein, B.J. & Wang, Z. (1992) Sedi-mentary record and climatic implications of deformation inthe Tian Shan: evidence from Mesozoic strata of the northTarim, south Junggar, and Turpan Basins, northwest China.Geol. Soc. Am. Bull., 104, 53^79.

Huang, B.C., Piper, J.D.A., Peng, S.T., Liu,T., Li, Z.,Wang,Q.C. & Zhu, R.X. (2006) Magnetostragraphic study of theKuche Depression, Tarim Basin, and Cenozic uplift of theTian Shan Range,Western China. Earth Planet. Sci. Lett., 251,346^364.



Huang,T.K. & Cheng, B.W. (1981) Pliocene-Quaternary mo-lasse deposits in the Tethys-Himalayan tectonic domain andits relation with the Indian Plate motion. In: Scienti¢c Paperson Geology for International Exchange (I) (Ed. by EditingGroup of Book and News in Ministry of Geology) pp.1^14. Geology Press, Beijing (in Chinese).

Jones,M.A.,Heller, P.L., Roca, E.,Garce¤ s,M. & Cabrera,L. (2004) Time lag of syntectonic sedimentation across an al-luvial basin: theory and example from the Ebro Basin, Spain.Basin Res., 16, 467^488.

Jordan, T.E., Flemings, P.B. & Beers, J.A. (1988) Datingthrust-fault activity by use of foreland-basin strata. In: NewPerspectives in BasinAnalysis (Ed. byK.L.Kleinspehn &C. Pao-la), pp. 307^330. Springer-Verlag, NewYork.

Kirschvink, J.L. (1980) The Least-Squares line and plane andanalysis of paleomagnetic data. Geophys. J. Int., 62(3), 699^718,doi: 10.1111/j.1365-246X.1980.tb02601.x.

Li, J.J.,Wen,S.X.,Zhang,Q.S.,Wang,F.B.,Zheng,B.X.&Li,B.Y. (1979) A discussion on the period, amplitude and type ofthe uplift of the Qinghai-Xizang Plateau. Sci. Sinica (Series A),22(11), 1314^1328 (in Chinese, English abstract).

Liu,T.S.,Ding,M.G.&Derbyshire,E. (1996)Gravel depositson the margins of theQinghai-Xizang plateau, and their envir-onmental signi¢cance. Palaeogeogr. Palaeoclimatol. Palaeoecol.,120, 159^170.

Lˇ,H.,Xia,Z., Jiang,B.,Ren,M.&Li,Y. (2006)Application ofmathematical geology methods to analysis of the sedimentaryfacies of sandstone reservoir: an example of the reservoir of theNeogene in Huatugou Oil¢eld, Qaidam Basin. Acta Sci. Nat.Univ. Pekinensis, 42(4), 462^469 (in Chinese, English abstract).

Lourens,L.,Hilgen,F.,Shackelton,N.J.,Laskar, J.&Wil-son, D. (2004) The Neogene Period. In: A GeologicTime Scale(Ed. by F.M. Gradstein, J.G. Ogg & A.G. Smith), pp. 409^440.Cambridge University Press, Cambridge.

Magee, J.W., Bowler, J.M., Miller, G.H. & Williams,D.L.G. (1995) Stratigraphy, sedimentology, chronology andpalaeohydrology ofQuaternary lacustrine deposits atMadiganGulf, Lake Eyre, south Australia.Palaeogeogr. Palaeoclimatol. Pa-laeoecol., 113, 3^42.

McElhinny,M.W. (1964) Statistical signi¢cance of the fold testin paleomagnetism.Geophys. J. Int., 8(3), 338^340, doi: 10.1111/j.1365-246X.1964.tb06300.x.

McFadden,P.L.&McElhinny,M.W. (1990) Classi¢cation of thereversal test in palaeomagnetism.Geophys. J. Int., 103, 725^729.

Me¤ tivier,F.&Gaudemer,Y. (1997)Mass transfer between east-ern Tien Shan and adjacent basins (central Asia): constraintson regional tectonics.Geophys. J. Int., 128, 1^17.

Miall,A.D. (1978) Lithofacies types and vertical pro¢le models inbraided river deposits: a summary. In: Fluvial Sedimentology 5(Ed. by A.D.Miall),Memoir, Canad. Soc. Petrol. Geol., 597^604.

Molnar, P. (2001) Climate change, £ooding in arid environ-ments, and erosion rates.Geology, 29(12), 1071^1074.

Molnar,P. (2004) Late Cenozoic increase in accumulation ratesof terrestrial sediment: how might climate change have af-fected erosion rates?Annu. Rev. Earth Planet. Sci., 32, 67^89.

Molnar, P., Brown, E.T., Burchfiel, B.C., Deng, Q.D.,Feng, .X.Y., Li, J., Raisbeck, G.M., Shi, J.B., Wu, Z.M.,Yiou, F. & You, H.C. (1994) Quaternary climate change andthe formation of river terraces across growing anticlines onthe north £ank of theTien Shan, China. J. Geol., 102, 583^602.

Molnar, P., England, P. & Martinod, J. (1993) Mantle dy-namics, uplift of theTibetan Plateau, and the Indian monsoon.Rev. Geophys., 31(4), 357^396.

Molnar,P.&Tapponnier,P. (1975)Cenozoic tectonics of Asia:e¡ects of a continental collision. Science, 189, 419^426.

Nemec,W. & Steel, R.J. (1984) Alluvial and coastal conglomer-ates: their signi¢cant features and some comments on gravellymass- £ow deposits. In:Sedimentology of Gravels and Conglomer-ates (10) (Ed. by E.H. Koster & R.J. Steel), pp. 1^32. CanadianSociety of PetroleumGeologists, Calgary.

Ort|¤ ,F.,Rosell,L.&Anado¤ n,P. (2003)Deep to shallow lacus-trine evaporites in the Libros gypsum (SouthernTeruel Basin,Miocene,NeSpain): an occurrence of PelletalGypsumRhyth-mites. Sedimentology, 50(2), 361^386.

Oskin,M. & Burbank,D.W. (2005) Alpine landscape evolutiondominated by cirque retreat.Geology, 33, 933^936.

Oskin,M.&Burbank,D.W. (2007)Transient landscape evolutionof basement-cored uplifts: example of the Kyrgyz Range, TianShan. J.Geophys. Res., 112, F03S03, doi: 10.1029/2006JF000563.

Pare' s, J.M.,Van DerVoo, R.,Downs,W.R.,Yan,M. & Fang,X. (2003) Northeastward growth of theTibetan Plateau: mag-netostratigraphic insights from the Guide basin. J. Geophys.Res., 108(B8), 1^11.

Patriat, P.A. & Achache, J. (1984) India-Eurasia collisionchronologyhas implications for crustal shortening and drivingmechanisms of plates.Nature, 311(5987), 615^621.

Peng, X.L. (1975) Cenozoic vertebrate localities and horizon ofJunggar Basin, Xinjiang.Verteb. PalAs., 13(3), 185^189 (in Chinese,English abstract).

Ren, M.D. & Wang, N.L. (1985) An Introduction to Modern Sedi-mentaryEnvironments. SciencePress, Beijing, 231pp (inChinese).

Rust, B.R. (1978) Depositional models fro braided alluvium. In:Fluvial Sedimentology (5) (Ed. by A.D.Miall),Canad. Soc. Petrol.Geol., Memoir 605^625.

Sobel, E.R., Oskin,M., Burbank, D.W. &Mikolaichuk, A.(2006) Exhumation of basement-cored uplifts: example oftheKyrgyzRange quanti¢edwith apatite ¢ssion track thermo-chronology.Tectonics, 25, doi: 10.1029/2005TC001809.

Song, C.H., Fang, X.M., Li, J.J., Gao, J.P., Zhao, Z.J. & Fan,M.J. (2001) Tectonic uplift and sedimentary evolution of JiuxiBasin in the northern margin of Tibetan Plateau since 13Ma.Sci. China (Series D), 44(Suppl), 192^202.

Sun, J.M. & Liu,T.S. (2006) The age of theTaklimakan Desert.Science, 312, 1621.

Sun, J.M.,Xu,Q.H. &Huang, B.C. (2007) Late Cenozoic mag-netochronology and paleoenvironmental changes in thenorthern foreland basin of theTian Shan Mountains. J. Geo-phys. Res., 112, B04107, doi: 10.1029/2006JB004653.

Sun, J.M., Zhu, R.X. & An, Z.S. (2005a) Tectonic uplift in thenorthernTibetan Plateau since13.7Ma ago inferred from mo-lasses deposits along the Altyn Tagh Fault. Earth Planet. Sci.Lett., 235, 641^653.

Sun, J.M., Zhu, R.X. & Bowler, J. (2004) Timing of theTian-shan Mountains uplift constrained by magnetostratigraphicanalysis of molasses deposits. Earth Planet. Sci. Lett., 219,239^253.

Sun, Z.M.,Yang, Z.Y., Pei, J.L.,Ge, X.H.,Wang, X.S.,Yang,T.S., Li,W.M. & Yuan, S.H. (2005b) Magnetostratigraphy ofPaleogene sediments from northern Qaidam Basin, China:implications for tectonic uplift and block rotation in northernTibetan plateau. Earth Planet. Sci. Lett., 237, 635^646.

Suppe, J., Chou,G.T. &Hook, S.C. (1992) Rates of folding andfaulting determined from growth strata. In: Thrust Tectonics(Ed. by K.R.McClay), pp.105^121. Chapman &Hall, Su¡olk.

Talling,P.J.&Burbank,D.W. (1993)Assessment of uncertain-ties in magnetostratigraphic dating of sedimentary strata. In:


H. Luet al.

Applications of Paleomagnetism to Sedimentary Geology (49) (Ed.by D.M. Aissaoui, D.F. McNeil & N.F. Hurley), SEPM (Soc. Se-diment. Geol.) Spec. Publ., 59^69.

Tapponnier, P. &Molnar, P. (1979) Active faulting and Ceno-zoic tectonics of theTien Shan, Mongolia, and Baykal regions.J. Geophys. Res., 84, 3425^3459.

Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud,N.,Wittlinger,G.& Jingsui,Y. (2001) Oblique stepwise riseand growth of theTibetan Plateau. Science, 394, 1671^1677.

Tucker, G.E. & Slingerland, R. (1996) Predicting sediment£ux from fold and thrust belts.Basin Res., 8, 329^349.

Windley, B.F., Allen,M.B., Zhang, C., Zhao, Z.Y. &Wang,G.R. (1990) Paleozoic accretion and Cenozoic redeformationof the ChineseTien Shan Range, Central Asia. Geology, 18(2),128^131.

Xu, C.M.,He,X.S.,Wu,X.Z. & Yao,X.Y. (1992) Structure ana-lysis and petroleum exploration prospect of Tostai area in Jug-gar Basin. Xinjiang Petrol. Geol., 13(3), 197^205 (in Chinese,English abstract).

Yang, X.P., Deng, Q.D. & Feng, X.Y. (1996) Research on geo-metry, kinematics of Tugulu active reverse fault-fold in thenorthern £ank of theTian Shan Mountains. In: Research of Ac-tive Fault (V) (Ed. byEditing Committee of the Research

of Active Fault), pp. 42^53. Seismology Press, Beijing (inChinese, English abstract).

Yang, Y. & Liu, M. (2002) Cenozoic deformation of the Tarimplate and the implications for mountain building in theTibe-tanPlateau and theTianShan.Tectonics, 21(6),1059, doi: 10.1029/2001TC001300.

Zhang, P.Z., Deng, Q.D., Xu, X.W.,Wu, Z.M., Yang, X.P.,Feng, X.Y., Li, J., Chen,W., Zhao, R.B. & Tang,W. (1994)Tectonic deformation, crustal shortening, and slip rate estima-tion along the Manas reverse fault-fold zone in the northern£ank of theTian Shan Mountains. In: Research of Active Fault(III) (Ed. by Editing of Committee of the Research of ActiveFault), pp. 18^32. Seismology Press, Beijing (in Chinese, Eng-lish abstract).

Zhang, P.Z.,Molnar, P. & Downs,W.R. (2001) Increased sedi-mentation rates and grain sizes 2^4Myr ago due to the in£uenceof climate change on erosion rates.Nature, 410, 891^897.

Zheng,H.B.,Powell,C.M.,An,Z.S.,Zhou, J.&Dong,G.R.(2000) Pliocene uplift of the northernTibetanPlateau.Geology,28(8), 715^718.

Manuscript received 9 July 2008; Manuscript accepted 5 March2009.



BasinResearch ...burbank.faculty.geol.ucsb.edu/Site/Publications_files/Lu...fold-and-faultzoneoftheforeland(Fig.1).Ourpaleomag-netic investigations provide a robust temporal framework

Documents