Sedimentary Record of Mesozoic Deformation and Inception of the Turpan-Hami Basin, Northwest China

Home , Hami, Junggar Basin, Karamay, Tian Shan, Turpan Depression

Geological Society of America Memoir 194 2001

Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin, northwest China

Todd J. Greene* Department of Geological and Environmental Sciences, Stanford University, Stanford, California, 94305-2115, USA Alan R. Carroll* Department of Geology and Geophysics, University of Wisconsin, 1215 West Dayton Street, Madison, Wisconsin 53706, USA Marc S. Hendrix* Department of Geology, University of Montana, Missoula, Montana 59812, USA Stephan A. Graham* Department of Geological and Environmental Sciences, Stanford University, Stanford, California 94305-2115, USA Marwan A. Wartes* Department of Geology and Geophysics, University of Wisconsin, 1215 West Dayton Street, Madison, Wisconsin 53706, USA Oscar A. Abbink* Laboratory of Palaeobotany and Palynology, Utrecht University, Utrecht, The Netherlands

ABSTRACT

The Turpan-Hami basin is a major physiographic and geologic feature of northwest China, yet considerable uncertainty exists as to the timing of its inception, its late Paleozoic and Mesozoic tectonic history, and the relationship of its petroleum systems to those of the nearby Junggar basin. To address these issues, we examined the late Paleozoic and Mesozoic sedimentary record in the Turpan-Hami basin through a series of outcrop and subsurface studies. Mesozoic sedimentary facies, regional unconformities, sediment dispersal patterns, and sediment compositions within the Turpan-Hami and southern Junggar basins suggest that these basins were initially separated between Early Triassic and Early Jurassic time. Prior to separation, Upper Permian profundal lacustrine and fan-delta facies and Triassic coarse-grained braided-fluvialÐalluvial facies were deposited across a contiguous Junggar-Turpan-Hami basin. Permian through Triassic facies were derived mainly from the Tian Shan to the south, as indicated by northward-directed paleocurrent di- rections. This is consistent with the sedimentary provenance of Triassic sandstone (mean & Qm29F29Lt42, Qp23Lvm49Lsm28, and Qm51P25K24) and conglomerate ( 32% granitic clasts) in the northern Turpan-Hami basin. We interpret a relative increase in quartz and feldspar concentration and a relative decrease in volcanic lithic grains in the northern Turpan-Hami basin to reflect unroofing in the Tian Shan and exposure of late Paleozoic granitoid rocks. In addition, two basinwide unconformities of Late PermianÐEarly Triassic and Early TriassicÐmiddle-Late Triassic age attest to deformation within the Turpan-Hami basin and associated continued uplift and erosion of the Tian Shan.

*E-mails: Greene, [email protected]; Carroll, carroll@geology. wisc.edu; Hendrix, [email protected]; Graham, [email protected]. edu; Wartes, [email protected]; Abbink, [email protected].

Greene, T.J., et al., 2001, Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin, northwest China, in Hendrix, M.S., and Davis, G.A., eds., Paleozoic and Mesozoic tectonic evolution of central Asia: From continental assembly to intracontinental deformation: Boulder, Colorado, Geologi- cal Society of America Memoir 194, p. 317Ð340.

317 318 T.J. Greene et al.

By Early Jurassic time, the Turpan-Hami basin and the southern Junggar basin became partitioned by uplift of the Bogda Shan, a major spur of the Tian Shan. In contrast to the thoroughgoing northward-directed Permian-Triassic depositional systems, Lower through Middle Jurassic strata begin to reflect ponded coal-forming, lake-plain environments within the Turpan-Hami basin. These strata contain paleocurrent indicators reflecting flow off the intervening Bogda Shan. Abasinwide Lower JurassicÐ Middle Jurassic unconformity in the Turpan-Hami basin suggests continued uplift and erosion of the Bogda Shan, consistent with a return to more lithic-rich sandstone and volcanic-rich conglomerate compositions. These sedimentary facies, paleocurrent, and provenance data sets provide the best available constraints on the initial uplift of the Bogda Shan and the first documentary evidence of intra-Mesozoic shortening within the basin.

INTRODUCTION by the Bogda Shan, a major mountain range containing peaks as high as 5570 m (Fig. 1A). Few published data bear directly on The present-day Turpan-Hami basin is a major physio- the initiation of the Turpan-Hami basin or its premodern physiographic feature of central Asia. It contains the second lowest el- graphic and depositional history. Post-Carboniferous sedimen- evation on earth (-154 m), yet is bounded on its northern flank tary facies, paleocurrent dispersal patterns, and sandstone

A. E86° E92°

CHINA N JUNGGAR BASIN 0 100 km DUSHANZI N44° N44° + + NORTH + + + + + + TIAN URUMQI S B HAN O BA GD AN RKO A SH L TAG TIAN SHAN TURPAN-HAMI H BASIN TURPAN HAMI CENTRAL TIAN TIAN N42° SOUTH SHAN SHAN N42° + +++ + + + +

KUQA KORLA K TARIM BASIN URUK TAGH

CHINA

PRE-MESOZOIC TRIASSIC JURASSIC CRETACEOUS CENOZOIC E86° E92°

E86° E88° E90° E92° E94° B. N44° K (Li and Jiang, 1987) N44° J T-P T-P C C southern Junggar Basin Urumqi Jimsar

B OG DA SHAN Zaobishan Taoshuyuan T2-3 Shisanjianfang Sandaoling Ha-3 T T2-3 P-T1 * P1-2 P2-T1 C

Kekeya 5 C Aiweiergou C 6 Taibei Sag 4 P1 -

J1-T Taican-2 8 F T Hami Baiyanghe lam * P2 Turpan in TURPAN-HAMI g M T88-625 tn K C Tokesun . Ancan-1 K (Mu, 1994) J2 J1-T3 T88-635 * J2-3 BASIN Sag J2 T-P T2-3 T2-3 J2-3 C T84-200 T-P T1 C P2 CHOL TAGH N42° N42°

E86° E88° ° E90 ° E92 E94°

50 km N Taoshuyuan study locality Legend fault Upper Jurassic (J3) Permian-Triassic (P2 - T1) Late Paleozoic granitoid rocks T88-635 seismic line Lower-Middle Jurassic (J1 + J2) Permian (P1 + P2) Paleozoic gabbroic rocks * well Cenozoic Middle-Upper Triassic (T2 + T3) Carboniferous volcanic rocks Pre-Mesozoic igneous and sedimentary depression Cretaceous (K1 + K2) metamorphic rocks

Figure 1. A: Location map of study area within Xinjiang Uygur Autonomous Region, northwestern China. Boxed area is shown in detail in B. B: Geologic map of Turpan-Hami basin, showing various study locations in basin, location of seismic lines and wells referred to in text, and two main western sedimentary depressions, Tokesun Sag and Tabei Sag (adapted from Chen et al., 1985). Solid black circles show general stratigraphic relationships and unconformities of Carboniferous through Cretaceous strata at various locations; data for each black circle are from this study unless otherwise noted. Fault locations are adapted from Chen et al. (1985) and Allen et al. (1993). Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 319 compositions from the southern Junggar and northern Tarim et al. (1993) also identified two Mesozoic compressional basins (Hendrix et al., 1992; Carroll et al., 1995) suggest that the episodes in the Turpan-Hami basin and suggested that Lower Junggar and Tarim basins have existed as discrete physiographic Permian marine facies in the northern Turpan basin were de- features, separated by the intervening Tian Shan, since late posited as a foreland basin fill related to north-vergent thrust Paleozoic time. However, previous interpretations of the time of faults in the Tian Shan region. inception and early structural style of the Turpan-Hami basin Key to understanding the tectonic and sedimentary history vary widely. Hendrix et al. (1992) suggested, on the basis of of the Turpan-Hami basin is better documentation of the record paleocurrents, sedimentary facies, and isopach data, that the of sedimentary fill within the basin, combined with information Turpan-Hami basin was established as a discrete physiographic regarding the history of uplift and unroofing of the Bogda Shan. feature by Early Jurassic time, in response to compressional up- Unfortunately, very little information has been published about lift of the Bogda Shan. Allen et al. (1995) proposed that the Tur- the structure of the interior of the Bogda Shan; access to key ex- pan basin was one of several major depocenters created by posures is difficult, and there are no seismic lines that transect transtensional rotation within a Late PermianÐTriassic sinistral the range. Tectonic interpretations from the Turpan-Hami basin shear system. In determining the sedimentary provenance of are largely limited either to specific reports on a focused northern Turpan-Hami deposits, Greene et al. (1997) and petroleum-related topic (Huang et al., 1991; Wang et al., 1993, Greene and Graham (1999) suggested that granitic cobbles con- 1998; K. Cheng et al., 1996; L. Liu et al., 1998), or regional tained within Lower Triassic deposits were derived solely from interpretive syntheses that lack detailed data sets and hence are the Central and South Tian Shan blocks, south of the Turpan- impossible to evaluate critically (Fang, 1990, 1994; Li and Hami basin, and inferred that the ancestral Bogda Shan had not Shen, 1990; Shen, 1990; Chen, 1993; Tao, 1994; Allen et al., been uplifted by that time. 1995; Shao et al., 1999b). Shao et al. (1999b), for example, used Better timing constraints for the initiation of the Turpan- basin subsidence curves to infer a period of Late Permian ther- Hami basin are especially important because they bear directly mal subsidence that was followed by a Middle-Triassic to early on the lateral extent of thick, organic-rich Upper Permian la- Tertiary period of flexural subsidence for the Turpan-Hami custrine shale in the southern Junggar basin. Carroll et al. (1992) basin. However, error bars for age control are absent, and paleo- and Clayton et al. (1997) demonstrated that these and equivalent elevation level is set at sea level for the past 250 m.y. in their strata are the source of oils produced at the Karamay oil field analysis. Likewise, numerous Chinese authors have published and other fields along the northwestern Junggar basin margin, reports on Early-Middle Jurassic depositional history and se- which collectively are thought to contain reserves of 2Ð3 billion quence stratigraphy, mainly due to its importance in petroleum barrel oil fields (Taner et al., 1988). If the initial separation of trapping for Turpan-Hami oil production (Wu and Zhao, 1997; these two basins by the Bogda Shan occurred during post-Late Li, 1997; Qiu et al., 1997; and many others). These reports, Permian time, then it is possible that the rich lacustrine source however, are difficult to assess because stratigraphic and sedi- rocks in the southern Junggar basin may extend into the Turpan- mentologic data are not presented. Marine sequence strati- Hami basin (Fig. 1B). graphic nomenclature is often used to describe high-resolution At a more regional level, structural, geochronologic, and lacustrine base-level changes during Early-Middle Jurassic sedimentologic evidence suggests that, during Mesozoic time, time, without presenting accurately located seismic lines, well- the Turpan-Hami and Bogda Shan area existed within a broader log cross sections with appropriate well logs used for correlat- zone of contractile deformation that underwent several episodes ing flooding surfaces or sequence boundaries (e.g., gamma-ray of shortening. Mesozoic contractile structures have been docu- log), or fossil assemblages to justify age control. mented in the subsurface of the Junggar and Tarim basins (Liu In order to better define the timing of initial formation of the et al., 1979; Liu, 1986; Wu, 1986; Li and Jiang, 1987; Peng Turpan-Hami basin and, by extension, the initial uplift of the et al., 1990; Zhao et al., 1991a, 1991b; Li, 1995; Hendrix et al., Bogda Shan, we examined key outcrops of Mesozoic strata, oil 1996; Wu et al., 1996), and Mesozoic fission-track ages from well cores and electrical logs, and seismic lines during the sum- the core of the Tian Shan (Dumitru et al., this volume) suggest mers of 1996, 1997, and 1998. We collected Permian through significant Mesozoic unroofing. Hendrix et al. (1992) docu- Jurassic stratigraphic and sedimentologic data at four different mented several successions of coarse, alluvial conglomerate in localities along the northern basin margin of Turpan-Hami each basin and inferred that they represent episodic renewed (Fig. 1B): Aiweiergou (westernmost corner), Taoshuyuan downcutting of the ancestral Tian Shan. Hendrix (2000) inter- (west), Zaobishan (north-central), north of the town of Shisan- preted variations in the composition of Mesozoic sandstone jianfang (east), and Sandaoling (east). We measured and de- from the southern Junggar and northern Tarim basins to reflect scribed &20 stratigraphic sections from exposed Mesozoic polyphase deformation of the ancestral range. Vincent and Allen strata, and we used those data as the basis for interpreting the en- (this volume) documented several Mesozoic angular unconfor- vironment of deposition for each section. We also examined nine mities and coarse conglomerate successions in the northeastern previously unpublished reflection seismic profiles from the in- Junggar basin and interpreted them to reflect far-field deforma- terior of the basin, along with data from oil well boreholes, and tion originating at the southern continental margin of Asia. Allen we used those to compliment our surface data set. In addition to 320 T.J. Greene et al. providing data on the timing of initiation of the Turpan-Hami et al., 1987; Z. Cheng et al., 1996; Wang et al., 1996), most for- basin as a depositional entity, we sought to examine the structure mations are dated to the epoch level, sufficient for the tectonic of the basin and its record of fill for evidence of Mesozoic short- interpretations in this paper (Fig. 2). ening, as has been suggested from other basins of western China. In this chapter, we summarize our data pertaining to Mesozoic Triassic sedimentary facies relations and sediment dispersal systems within the Turpan-Hami basin. We integrate those data with Taodongou (east Taoshuyuan). At Taodongou (Figs. 1 sandstone composition and conglomerate clast count data. Col- and 3A), 144 m of Lower Triassic strata (T1s; see Fig. 2 for for- lectively, these data provide the best available constraints on mation abbreviations herein) are truncated by an erosional un- Mesozoic shortening within the Turpan-Hami basin, and, indi- conformity and overlain by at least 70 m of Middle to Upper rectly, on the initial uplift of the Bogda Shan. Triassic (T2-3k) conglomerate (Figs. 3B and 4A). The lower 80 m of T1s are mostly red to pink mudstone locally interbed- STRATIGRAPHY AND SEDIMENTOLOGY ded by 4Ð8 m sandy conglomerate interbeds that sharply fine to siltstone. The upper 64 m of the T1s section represent a series of Our sedimentary data set, described in detail in the follow- stacked upward-fining sandy conglomeratic intervals &8Ð16m ing, indicates that several fundamental changes in depositional thick that fine to siltstone; little to no mudstone is exposed. Ero- style occurred in the Turpan-Hami basin in Permian through sional contacts are common at the base of each package. Toward Middle Jurassic time. This depositional history can be summa- the top of the T1s, the upward-fining sand packages pinch out rized as follows: Upper Permian strata are marked by fine- laterally over 20Ð30 m. An erosional contact with &25 m of re- grained, mostly profundal lacustrine, fan-delta, and fluvial lief separates the T1s deposits from the overlying higher energy facies (Carroll et al., 1992; Wu and Zhao, 1997; Carroll, 1998; T2-3k conglomerate (Fig. 3B). The latter consists of clast- Wartes et al., 1998, 1999, 2000), whereas Triassic deposits con- supported conglomerate containing abundant tabular and trough tain coarse-grained braided-fluvialÐalluvial facies. In sharp cross stratification with 71 m relief. Cross-bedded sandy lenses, contrast are Lower through Middle Jurassic strata that reflect the & 1 m thick, are common; little mudstone is present. ponding of water in coal-forming, lake-plain environments. We interpret these deposits to represent two different de- Concomitant with this major change in Jurassic environments positional environments. The lower 144 m of T1s deposits was a shift during Early and Middle Jurassic time in the main record relatively low energy flow associated with a braided sedimentary depocenter, from the western Tokesun Sag to the river flood plain (Nanson and Croke, 1992). The upper 64 m north-central Tabei Sag (Fig. 1B: Huang et al., 1991; Wang record increasing depositional energy and erosive power that et al., 1996; Qiu et al., 1997; Wu and Zhao, 1997). produced minor channels over a braid plain. Following an &25-m-deep incision, a gravel bed braided-fluvial system Age control dominated, incising large channels and depositing a series of gravel bars (Miall and Gibling, 1978). Hendrix et al. (1992) Due to the lack of interbedded datable volcanic units and a described similar deposits from Upper Canfanggou Group paucity of paleomagnetic studies, age assignments in the (lowermost Triassic) strata at Taodongou (Fig. 3A). They de- Turpan-Hami basin are based solely on plant microfossil scribed a detailed 50 m section of coarse conglomerate and (spores, pollen), plant macrofossil, and vertebrate fossil assem- lenticular sandstone and siltstone red beds deposited in a blages. This study relies heavily on new palynological results braided-fluvialÐalluvial environment. (Abbink, 1999) derived from selected mudrocks within bore- Zaobishan. Three measured sections were described on a hole core and outcrop samples (Fig. 2). Age interpretations are meter scale at the Zaobishan locality (Fig. 4, B, C, and D). Sec- based on the first occurrence datum and/or last occurrence da- tions 4B and 4C both record Lower Triassic and Middle-Upper tum of spores and pollen (sporomorphs). Most of the strati- Triassic deposits on the north and south flanks of a large east- graphic ranges of the encountered sporomorphs are not westÐtrending Carboniferous-cored anticline (Figs. 5A and 6), calibrated against marine faunas from northwest China (e.g., and are chronostratigraphically similar to the measured section Huang, 1993; Ouyang, 1996). We therefore conducted a data at Taodongou (Fig. 4A). search to determine the range of the stratigraphic marker taxa Both measured sections 4B and 4C can be divided into a for northwest China based on selected literature references from lower fine-grained section and an upper coarse-grained section, other regions. Furthermore, on the basis of the overall content presumably separated by an erosional unconformity at the 25 m of the sporomorph assemblages, the palynoflora of our samples mark (although we observed an erosional contact in section 4B, are not considered to be endemic; rather, the floral associations this contact is covered at section 4C). The lower finer grained closely resemble those of Europe and of the former USSR. On portion consists of fine- to medium-grained sandstone beds with the basis of our palynological studies, along with independent trough cross-beds and planar bed stratification. The lower half age assignments published by Chinese paleontologists (Liao of section 4B is relatively lower energy than 4C, the former con- Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 321

Schematic Age Formation stratigraphy Well logs on seismic lines Palynological assemblages Ha-3 well (Figure 10C) Qiketai (J2q) 1

Sanjianfang (J2s) 1) Jurassic undifferentiated: Alete Bisaccate pollen 1

Middle (J2) undifferentiated, Xishanyao (J2x) Perinopollenites elatoides, Punctatisporites spp. 3 Jurassic 2 2) Lower Jurassic Sangonghe (J1s) (Hettangian-Aalenian): Chasmatosporites minor, Apiculatisporis globosus,

Lower (J1) Lower Badaowan (J1b) common Striatella spp. Figure 2. Generalized stratigraphic and lithologic chart of Turpan-Hami stratigraphy from Carboniferous through Middle Haojagou (T3h) 2 Jurassic strata with reported fauna and flora assemblages and Ancan-1 well (Figure 10B) stratigraphic positions of major unconformities discussed in text. Palynological assemblages from Triassic Huangshanjia (T3hs) Ha-3 and Ancan-1 wells serve as 2 3) Upper Triassic (Late Norian): Middle/Upper (T2-3) 2 Lordonispora fossulata, age control for seismic line draw- Vallasporites spp., Araucaricites ings. Unless otherwise noted, all australis, Patinasporites denus, palynology-based age interpreta- 3 Stereisporites puncta tions are from Abbink (1999). For 4 Karamay (T2-3k) 4) Middle-Upper Triassic*: Upper Permian age assignment of Alisporites, Cyclogranisporites, sample 6, note that association Duplexisporites with Limatulasporites (=Gor- Lower Shaofangou (T1s) (T1) Jiucaiyuan (T1j) donisporites) and Taeniasporites (=Lunatisporites) is not typical 5 5) Upper Permian*: Late Permian. Although latest Guodikeng (P2g) Limatulasporites, Lundbladispora, Permian samples from Salt Range Taeniaesporites, Bisulcocypris sp. (Pakistan) contain the taxa, these 1 Wutonggou (P2w) 6) Upper Permian (Wordian): taxa are also typical for Early Tri- Upper (P2) 3 Hamiapollenites bullaeformis, assic, in particular in north China Lunatisporites noviaulensis, Quanzijie (P2q) (Ouyang and Norris, 1988). Vittatina spp., Klausipollenites schaubergeri, Weylandites Taierlong (P2t) 6 striata, Cordaitina,

Permian Gardenasporites, Daheyan (P2d) (km) Protohaploxipinus, SP Resistivity Straitoabieites, Gordonisporites ? coal Key vertical scale: Taoxigou (P1t) 100 m mudstone unconformity sandstone

Lower (P1) Lower 6 sampled for palynology 0 conglomerate (km) Carboniferous * from the Tu-Ha Petroleum Bureau, volcanic Ancan-1 well taining paleosol horizons and the latter having more stacked lenticular packages that grade from clast supported at their trough cross-bedded sandstone beds. Paleocurrent indicators scoured bases to matrix supported. point north-northeast for section 4C. At the 25 m mark in both We interpret sections 4B and 4C to be similar in deposi- sections, a sharp increase in grain size occurs, above which tional style to section 4A. The Lower Triassic (T1s) sections are at least 30 m of polymictic conglomerate containing large record lower energy deposition in a braided river flood plain (71 m relief) trough cross-beds mantled by pebbles and small or distal sheetflood environment (Nanson and Croke, 1992). cobbles. The upper half of section 4B consists of sandy con- Middle-Upper Triassic (T2Ð3k) strata recorded a major change glomerate with 1Ð2-m-thick medium- to coarse-grained trough in depositional energy represented by gravelly braided-fluvial cross-bedded sandstone lenses distributed throughout the sec- systems (Miall and Gibling, 1978) with associated sandy over- tion. The conglomerate consists of 2Ð10-m-thick shingled, bank sheet flows. Hendrix et al.’s (1992) description of Triassic 322 T.J. Greene et al.

A. Taodongou locality, east Taoshuyuan deposits at Qijiagou (southern Junggar basin) was similar to that for Triassic strata at Zaobishan and Taodongou. Namely, fine- E88o55' E89o00' grained red beds of lowermost Triassic deposits sharply grade to coarse braided fluvial conglomerate of the Middle-Upper Trias- T1 section * sic Karamay Formation. J1 section N43o15' * section 4A Section 4D was measured in a valley of Lower Triassic deposits (T1s) and is presented as a photomosaic in Figure 5B. Generally, section 4D consists of 2Ð10-m-thick lenses of trough section 7B cross-bedded conglomerate packages interbedded with 1Ð2-m- thick medium- to coarse-grained, trough cross-bedded sand- N43o14' stone. Cobbles commonly mantle the troughs within the

N conglomerate, and scoured basal contacts are ubiquitous. Mea- sured paleocurrent indicators for trough cross-beds point north 2 km and northeast (Fig. 5C). The excellent two-dimensional valley-wall exposure *from Hendrix et al. (1992) shown in Figure 5B contains all the major elements of a classic gravel-bed braided-fluvial system (Miall, 1996), i.e., abundant B. Section 4A, Triassic deposits gravel bars, sandy bedforms, and stacked, channelized sandy Middle-Upper Triassic section 4A conglomerate packages with numerous internal erosion sur- Karamay Formation (T2-3k) faces, lack of fines, and few observed downstream accretion surfaces. Zhao et al. (1991) documented similar T1s deposits in erosional surface the Dalongou area (southern Junggar basin) on the north side of the Bogda Shan, and also interpreted them to reflect a braided- fluvial environment. A modern analog for the coarse-grained Lower Triassic Triassic deposits throughout Turpan-Hami occurs in the western Shaofangou Fm. (T1s) Tian Shan in the Lake Issyk-kul area of Kyrgyzstan, where intermontane basins commonly contain gravel-dominated braided river environments with associated transverse bars (Sgibnev and ~10 m Talipov, 1990; Rasmussen and Romanovsky, 1995). Shisanjianfang. Section 4E depicts Middle-Upper Trias- sic conglomerate just north of the town of Shisanjianfang (Figs. 1 and 4E). Several 4Ð18-m-thick beds of clast- and C. Section 7B, Lower Jurassic deposits matrix-supported conglomerate interfinger with medium- to coarse-grained trough cross-bedded sandstone. The mean paleocurrent indicator direction points north-northeast. We also interpret section 4E to represent deposition in a braided-fluvial system with occasional lower energy sand-rich, matrix-support section 7B deposition similar to those described herein.

Lower and Middle Jurassic

We studied exposures of Lower and Middle Jurassic deposits in the western part of the basin at Aiweiergou, in the north-central part of the basin at Kekeya, at Taodongou (east ~20 m Taoshuyuan), and in the eastern Hami basin at Sandaoling (Figs. 1 and 7). We also studied Middle Jurassic strata in the central part of the basin at Flaming Mountain where thick sandstone Figure 3. A: Corona Satellite image of Taodongou locality in east Taoshuyuan. Black lines show locations of sections 4A and 7B from packages of the Xishanyao and Sanjianfang Formations (J2x this study, as well as Lower Triassic (T1) and Lower Jurassic (J1) mea- and J2s) and lacustrine deposits of the Qiketai Formation (J2q) sured sections from Hendrix et al. (1992). B: Intra-Triassic erosional are exposed (Huang et al., 1991; Schneider et al., 1992; C. Wang surface at Taodongou, where braided-fluvial deposits of Karamay For- et al., 1996; Liu and Di, 1997; H. Wang et al., 1997; Wu and mation (T2-3K) overlie Shaofangou Formation (T1s); stratigraphic Zhao, 1997; L. Liu et al., 1998; Greene et al., 2000). “up” is to upper right of photo. White bar shows location of portion of measured section 4A. C: Entire 66 m measured section 7B (shown as Aiweiergou. Near the Aiweiergou coal mine, we measured white line) representing Lower Jurassic flood-plain deposits; strati- a 240 m section of Lower Jurassic Badaowan Formation (J1b) graphic “up” is to upper left of photo. Triassic outcrops of the Turpan-Hami basin

Taodongou locality: Shisanjianfang T2-3k/T1s Zaobishan locality 4E east Taoshuyuan (south section) (m) locality T2-3k/T1s (m) T2-3k/T1s 48 T2-3k 4A (m) 4B (north section) (m) 40 64 50 4C

48 80 24 40 32

gravel-bed braided fluvial braided gravel-bed 16 N

erosional unconform fluvial braided gravel-bed (~25 m 16 gravel-bed braided fluvial braided gravel-bed

gravel-bed braided fluvial braided gravel-bed 70 8 30

scour) 0 Middle/Upper 0 Triassic (T2-3k) cob vc m mud (n = 51) grain size 144 covered hillslope trough limbs Lower Triassic ity (T1s)

20 128 20 N

112 Zaobishan locality T1s 10 (n = 23) (m) 96 trough limbs 10 4D sheetflood distal braided braided river floodplain river braided

minor channels on braid plain minor channels on braid 70 80

cob vc m mud 0 grain size

~200 m section 64 0 cob vc m mud 60 grain size

48 LEGEND

50 32 conglomerate braided river flood plain river braided sandstone

siltstone 16

mudstone 40

partially covered interval 0 cob vc m mud grain size Ca-rich stringers upward fining upward coarsening 30

organic material concretions gravel-bed braided fluvial braided gravel-bed N

conglomerate lag

20 trough cross-beds mantled with cobbles

large-scale trough (n = 48) cross-bedding (>1 m) trough limbs N low-angle 10 cross-bedding trough cross-bedding plane-bed stratification mud parting 0 (n = 46) cob vc m mud grain size trough limbs

Figure 4. Summary of Triassic measured sections in Turpan-Hami basin. Note erosional unconformity between undifferentiated Middle-Upper Triassic conglomerate scouring into fine sandstone of Lower Triassic deposits in sections 4A, 4B, and 4C. This erosional surface (photo in Fig. 3B) is present at both Zaobishan and Taodon- gou (east Taoshuyuan) as well as in seismic line T88-635. Note that paleocurrent indicators are directed north to northeast in sections 4C, 4D, and 4E. 324 T.J. Greene et al.

A. Zaobishan locality

E90o20' E90o25' E90o30' E90o35' N43o20' C. Paleocurrent indicators

section 4C

section 4D N43o16'

trough limbs N section 4B (n = 143)

5km N43o12'

B. Lower Triassic braided fluvial deposits, section 4D W E

~10 m

section 4D

Figure 5. A: Corona Satellite image of Zaobishan locality showing location of sections 4B, 4C, and 4D. White box shows location of photomosaic pictured in part B. B: Photomosaic of Lower Triassic braided-fluvial deposits at Zaobishan. Black lines highlight interpreted depositional surfaces. White line represents 76-m-long measured section 4D. Stratigraphic “up” is to top of photo. Note numerous large-scale cross-beds indicating flow to right (northeast). C: Photo of ubiquitous trough cross-sets (pencil for scale) measured at locality pictured in B, along with corresponding stereoplot of measured paleocurrent indicators. Mean vector indicates northeast sediment-dispersal direction. deposits (Fig. 7A). The lower 96 m consist mainly of 8Ð16-m- Hendry, 1987). Conglomerate beds have crude horizontal strati- thick, clast-supported, polymictic cobble conglomerate with fication, and they commonly interfinger with large-amplitude several associated 1Ð2-m-thick medium- to coarse-grained, cross-bedded sandy facies. The rippled siltstone beds are most trough cross-bedded and plane-bed laminated sandstone lenses. likely interchannel overbank deposits or abandoned channels Conglomerate intervals commonly contain scoured basal con- and meander scars on the flood plain that are preserved as silt tacts and are laterally continuous over several meters. Con- plugs. Campbell and Hendry (1987) described modern gravel glomerate clasts are 2Ð8 cm and are well rounded. Wood meander lobes on the Saskatchewan River that contain many of fragments are common, as are interbeds of 0.5Ð1-m-thick, the same elements as this portion of the Badaowan Formation: rippled, fine-grained sandstone and siltstone. horizontal gravel sheets, interfingering cross-bedded pebbly We interpret this portion of the Badaowan Formation to sands, and clay-silt plugs. have been deposited in a gravel-sand meandering fluvial envi- The remainder of the section (above the 96 m mark) is rela- ronment (cf. Nijman and Puigdefábregas, 1978; Campbell and tively finer grained and more organic rich than the underlying Summary stratigraphic column for Zaobishan locality Palynology and (m) Paleocurrent Lithofacies/Depositional Environment Indicators

2500 FLOOD PLAIN-ANASTOMOSING FLUVIAL: T2-3 braided fluvial sequence gradually fines upward to interbedded mud and/or siltstone with fine to medium sandstone containing trough cross-bedding. Two coal seams (<1 m thick) are interbedded with

T2-3k siltstone.

TRIASSIC ALLUVIAL-BRAIDED FLUVIAL: Dominated by cobble conglomerate, 2000 with abundant scour features. Individual conglomerate packages upwardly fine to medium-coarse sandstone. The sequence is bounded by a basal erosional unconformity at its base.

sections N FLOOD PLAIN, DISTAL SHEETFLOOD: Upward-fining sand 4B, 4C packages with abundant well-developed trough cross-bedded, regularly interbedded mudstone and siltstone, occasional paleosols, and rare conglomerate.

n = 23, trough limbs

1500 T1sh section 4D N N BRAIDED FLUVIAL: Well-organized conglomerate with cobble-mantled (>1 m) trough sets. Cross-bedded sand lenses are common. Sandy conglomerate beds are rich in alkalic plutonic clasts. T1j n = 46, trough limbs n = 48, trough limbs

P2-T1 LAKE PLAIN-MEANDERING FLUVIAL: Dominantly red mudstone and siltstone interbedded with sandy micrite and occasional plane-laminated sheet sandstone and minor 1000 P2w trough cross-bedded sandstone. P2q

1 1) Middle - early late Permian: MEANDERING FLUVIAL-LACUSTRINE: Upward-fining sequences Vittatina spp., Klausipollenites of trough cross-bedded sandstone; sequences are typically scoured schaubergeri, Hamiapollenites at their base and capped by a rippled surface. The fine-grained

P2t bullaeformis, Cordaitina spp., Gardenasporites, Protohaploxipinus, section is dominated by calcareous siltstone and mudstone

PERMIAN Straitoabieites interbedded with thin, laterally continuous limestone. N BRAIDED FLUVIAL: Dominantly well-organized conglomerate 500 P2d of mostly andesitic clast composition. Abundant trough cross- trough limbs bedding and occasional large silicified trees are present; P2d n = 94 interbedded sandstone beds display well-developed cross-bedding.

ALLUVIAL FAN AND LACUSTRINE: Interbedded limestone and conglomerate with well preserved tufa, overlain by variably laminated, profundal lacustrine mudstone and fine sandstone.

P1y VESICULAR BASALTIC AND ANDESITIC FLOWS

MIXED MARINE CARBONATE AND SILICICLASTICS: 0 Thinly bedded micritic limestone, red mudstone, and chert. CARBON- IFEROUS. ANDESITIC FLOWS

mudstone conglomerate

Figure 6. Measured stratigraphic section of Permian through Triassic deposits at Zaobishan with descriptions of lithostratigraphy and depositional environment (see Figs. 1 and 5 for locations). This generalized section at Zaobishan is typical along north rim of Turpan-Hami basin. Note that paleocurrent indicators for Upper Permian and Lower Triassic deposits are pointed northwest to northeast. Arrow (labeled 1) points to location of P2t mudstone from which we recovered mid-early Late Permian palynomorphs. Aiweiergou locality Lower-Middle Jurassic outcrops of 7A J1b the Turpan-Hami basin (m) 7B LEGEND 240 conglomerate conglomerate lag Palynological assemblages sandstone 1) Lower Jurassic (Toarcian): clast-supported Taodongou locality: abundant Striatella cf. balmei, siltstone conglomerate east Taoshuyuan Quadraeculina anellaeformis sequence mudstone trough cross-beds repeats mantled with cobbles 2) Jurassic undifferentiated: J1b coal Alete bisaccate pollen undiff., trough cross-bedding Deltoidospora spp., (m) covered interval low-angle cross-bedding Perinopollenites elatoides, plane-bed stratification Punctatisporites spp. coal fragments mud parting 200 plant material 3) Middle Jurassic (Bajocian- ripples Bathonian): Neoraistrickia woody material climbing ripples bacclifera, N. truncata, mud rip-up clast Quadraeculina anellaeformis 60 upward-fining 2 concretions sequence root casts 2 sampled for palynology upward-coarsening Sandaoling locality J2x 7C Kekeya locality 7D (m) 50 J1b (m) sandy meandering river 150

photo in Figure 8B 40 40 1 amalgamated

(n = 34)

amalgamated trough limbs

logs and casts (0.5 m) 3 30 100 30

30 0.5 m mud rip-up flood plain adjacent to meander-belt system 20 20 20

photo in Figure 9B sandy braided-fluvial-distal sheetflood sandy braided-fluvial-distal

50 on deltaic flood plain anostomosed river mud rip-ups

deformed mud interbeds 10 10 10 gravel-sand meandering river gravel-sand

0 0 0 0 vc m mud cob vc m mud cob cob m mud mud/coal grain size grain size med fine grain size grain size

Figure 7. Summary of Lower and Middle Jurassic outcrops throughout Turpan-Hami basin (see Fig. 1B for localities). Measured section 7A, at Aiweiergou, represents only portion of thick Lower Jurassic strata in western depression of Turpan-Hami (Tokesun Sag). Section 7B (photo in Fig. 3C) represents contemporaneous deposits in north-central depression (Tabei Sag). Generally, Lower Juras- sic deposits in Tabei Sag are not as thick and coarse as Lower Jurassic deposits in Tokesun Sag. Section 7C is from Kekeya locality. Paleocurrent indicators in measured section 7C are directed south to southeast, signifying reversal from previously north- to northeast-directed paleocurrent indicators in Permian and Triassic deposits. Middle Jurassic deposits from Sandaoling coal mine are described in section 7D. Numbered arrows signify positions of samples studied for palynology; interpreted ages and palynological assemblages for each sample are listed in legend (Abbink, 1999). Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 327

section (Fig. 7A). We interpret this section to represent a sandy A. Kekeya locality meandering depositional environment with associated crevasse E90o05' E90o07' splays, flood-plain, and overbank deposits. This interval consists of three main upward-fining successions (30Ð60-m thick),

o C. Paleocurrent indicators each consisting of basal scouring cross-bedded conglomerate N43 13' and sandstone capped by siltstone, carbonaceous mudstone, and N coal with high mud content. In the lower portion (96Ð134 m), section 7C mudstone is interbedded with tabular fine-grained, trough cross- bedded sandstone beds (61 m thick) and a couple of 3Ð5-m-

N o (n = 34) thick high-mud humic coal seams, abundant wood fragments, N43 11' trough limbs and plant material. The tabular sandstone layers most likely 2 km record periodic flooding events or small chute channels deposited on a meander flood plain. The next overlying section B. Trough cross-bedding in J1b deposits, section 7C (134Ð196 m) upwardly fines from medium-grained, trough cross-bedded sandstone with rare 61 m conglomerate lenses to coal and carbonaceous mudstone containing plant debris and numerous concretions. This section could either represent pro- grading sandy bars or a meandering fluvial channel. The final portion (205Ð236 m) consists of an upward-fining package with a scoured base and a basal conglomerate lag that abruptly fines to a medium- to coarse-grained, trough cross-bedded sandstone with dispersed cobbles. We interpret these repeating upward- fining cycles to reflect small meander channels. Taodongou (east Taoshuyuan). Excellent lateral exposure of J1b deposits can be observed near the Taodongou coal mine Figure 8. A: Corona Satellite photo of Kekeya locality showing loca- in north-central Turpan-Hami (Fig. 3, A and C). Measured section of section 7C. B: Trough cross-bedding associated with sand-rich tion 7B encompasses 65 m (represented by a white line in braided-fluvial facies of Lower Jurassic Badaowan Formation (J1b) described in section 7C (photo is from 40 m mark of section 7C). Strati- Fig. 3C) of mostly green-gray mudstone capped by a 4Ð5-m- graphic “up” is to upper right of photo. C: Stereoplot of measured thick coal seam. An 8-m-thick succession of amalgamated, paleocurrent indicators from section 7C. Mean vector indicates south- fining-upward sandstone beds occurs midway through the sec- southwest sediment-dispersal direction. tion. These beds contain trough cross-beds mantled by pebbles, plane-bed stratification, and scoured basal contacts. Overall, the Taodongou J1b section records deposition on a flood plain, in the Sandaoling coal mine (Figs. 1B and 9). The base of sec- overbank environment with associated crevasse splays that are tion 7D contains a 12-m-thick coal seam, overlain by organic- probably part of larger meander-belt system. Hendrix et al. rich interbedded coal and laminated mudstone with small (1 m (1992) also examined lowermost Jurassic deposits at Taodon- thick) channelized sand bodies containing cross-bedding, coal gou that they interpreted to be from a braided-fluvial flood fragments, and root casts (Fig. 9B). The channels appear to be plain; however, no attempt is made to correlate the two sections. isolated, with minimal evidence for lateral migration, and are Kekeya. Near the Kekeya coal mine (Figs. 1B and 8) we bounded by flood-plain deposits. We interpret the depositional measured a detailed 44-m-thick section of J1b deposits (sec- environment as an anastomosed river deposited on a subaque- tion 7C). The lowermost 4 m consist of coal interbedded with ous, deltaic flood plain, as Smith (1986) described for the Mag- dark gray, carbonaceous shale containing abundant coaly wood dalena River in northwestern Columbia. fragments. These units are overlain by 40 m of amalgamated 2Ð8-m-thick intervals of medium- to coarse-grained, trough REGIONAL UNCONFORMITIES cross-bedded sandstone. Silicified logs and casts to 0.5 m long, mud rip-up clasts, and basal conglomerate lags are all common Pre-Tertiary uplift of the Bogda Shan undoubtedly would throughout the section. Trough cross-bed orientations generally have affected the Mesozoic sedimentary fill of the Turpan-Hami indicate paleoflow to the south-southwest (Fig. 8, B and C). On basin on a regional scale, and therefore might be expressed as the basis of the lack of lateral accretion surfaces and abundant regional unconformities. To test this hypothesis, we examined erosional features with ubiquitous mud rip-up clasts and woody surface exposures, as well as nine different regional seismic re- debris, we interpret this succession to reflect sandy braided- flection lines throughout the Turpan-Hami basin (made avail- fluvial or distal sheetflood sedimentation along a flood plain able by the Chinese National Petroleum Corporation). The (Olsen, 1989; Miall, 1996). seismic lines were shot either parallel (east-west) or perpendicu- Sandaoling. In the eastern Turpan-Hami basin, 50 km lar (north-south) to the basin axis, and were recorded down to northwest of Hami, we measured Middle Jurassic coaly strata &3.0 s (two-way traveltime). Two sets of two crossing seismic 328 T.J. Greene et al.

A. Sandaoling locality An Upper PermianÐLower Triassic (P2-T1) angular unconformity occurs at the Aiweiergou locality (Fig. 11) and on seis- section 7D mic line T88-635 (Fig. 10B). It is important to note that although an Upper PermianÐLower Triassic angular unconformity is present in the southern and western portions of the basin, localities to the north in the southern Junggar basin (e.g., Jimsar; Fig. 1B) show conformable P2-T1 stratigraphy (Yang et al., 1986; Liao et al., 1987; Hendrix et al., 1992; Carroll et al., 1995; Tang et al., 1997). At both Taodongou (east Taoshuyuan) and Zaobishan, an erosional unconformity marks a major sedimentologic break from Lower Triassic sandstone and paleosols (T1) to deeply scoured Middle-Upper Triassic (T2-3) conglomeratic deposits (Fig. 4, A and B). The T1-T2-3 unconformity is also expressed in the eastern side of seismic line T88-635 (Fig. 10B). A slightly younger intra-Upper Triassic unconformity appears in seismic B. Root casts in J2x deposits, section 7D line T88-625 (Fig. 10D). Although we observed no exposures of the Lower-Middle Jurassic (J1-J2) contact, the J1-J2 unconformity is a convincing erosional surface on all of the seismic lines (Fig. 10, A, B, C, and D). In map view, the discordance persists for as much as 50 km on both strike and dip seismic lines (Figs. 1B and 10). It is especially important to understand this feature because Turpan-Hami basin petroleum source-rock intervals and reser- 1 m voirs are within Lower and Middle Jurassic strata (Huang et al., 1991; Wang et al., 1996; Wu and Zhao, 1997). Hence, this unconformity has direct ramifications for the lateral preservation of source rocks and the geometry of petroleum traps. A prominent Jurassic-Cretaceous (J-K) angular unconformity appears in seismic line T89-465 in the eastern Hami area (Fig. 10C). Although we did not observe a J-K unconformity at Flaming Mountain, it is recognized by many Chinese publica- Figure 9. A: Outcrop of anastomosed river deposits described in section 7D (black line) from Middle Jurassic Xishanyao Formation (J2x) tions in the western Turpan-Hami area (Li and Shen, 1990; Zhao at Sandaoling coal mine (Figs. 1B and 7D). Note person for scale at bot- et al., 1991a, 1991b; Mu, 1994; Wang et al., 1994). tom of photo. B: Photo of root casts (indicated by arrows) from 16 meter mark of section 7D. Stratigraphic “up” is to top for both photos. PALEOCURRENT DATA

Paleocurrent indicator data from this study were collected line drawings are presented (see Fig. 1B for locations): T84Ð200 from limbs of trough cross-beds and from imbricated clasts in and T88-635 in the west part of Turpan-Hami (Fig. 10, A and B), fluvial deposits. After correcting for structural dip, the three- and T89-465 and T88-625 in the east (Fig. 10, C and D). Paly- dimensional planes of trough cross-beds were plotted as poles nology from this study, as well as from the Turpan-Hami Petro- on a stereographic plot (lower hemisphere projection). The pole leum Bureau, provide age control for Permian through Jurassic to a statistical best-fit great circle through all of the trough pole reflectors intersecting the Ancan-1 and Ha-3 wells (Figs. 1B, 2, data describes the average paleocurrent direction for the de- and 10, B and C). posits (cf., DeCelles et al., 1983). For clast imbrication data, On the basis of seismic line interpretations and outcrop the pole of each restored imbrication plane was plotted on a rose observations, at least four different regional unconformities can diagram along with the mean vector (bin = 1¡). Although the be identified (Fig. 1B): (1) Upper PermianÐLower Triassic (P2- small number of paleocurrent indicators is not statistically sig- T1); (2) LowerÐMiddle-Upper Triassic (T1-T2-3); (3) Lower- nificant, it is sufficient to define broad temporal and regional Middle Jurassic (J1-J2); and (4) Jurassic-Cretaceous (J-K). trends. Because of the thick Jurassic through Tertiary section in the Figure 12 summarizes paleocurrent data from this study as northern half of the basin (75 km in the Tabei Sag; Wu and well as five other reports for Permian through Jurassic deposits Zhao, 1997), no pre-Jurassic reflectors were imaged close to the of the Turpan-Hami and southern Junggar basins (Carroll, 1991; Bogda Shan. Zhao et al., 1991; Hendrix et al., 1992; Schneider et al., 1992; Legend

base Tertiary unconformity (T) fault base Cretaceous unconformity (K) truncated reflector base Middle Jurassic unconformity (J1-J2) name of seismic crossline T86-658 intra-Triassic unconformity (T1-T2-3 or T3-T3)

A 10 km T84-200

Naiqian-5 well surface control at Flaming Mountain S Naiqian-4 well T88-635 J2 J3 N T84-625 T86-642 T86-646 T87-199 T84-650 T86-654 T86-658 unresolved area 685 m structure unresolved 1.0 s 1.0 s J2 ? J1

2.0 s 2.0 s

B 10 km T88-635 Ancan-1 well W T84-200 E T86-638 T86-140 T86-165 T84-165 T88-175 T194-82 T84-145 T86-150 T85-637 T87-156 (km)

1.0 s 1.0 s

2 J2 T2-3 T1 2.0 s J2 2.0 s T2-3 P2 3 P2 T2 T1

3.0 s 4 3.0 s SP Resistivity C 10 km T89-465 Ha-3 well S T88-625 N T93-613 T89-615E T88-620 T87-630 T88-635 T91-640 T87-665 T89-667 T91-673 T88-677 T89-681 T91-683 T89-687 T89-689 (km) T90-692

(poor data) 1.0s 1 J2 1.0s J1 K 2 J2x

T3 3 T3h 2.0s SP Resistivity ? 2.0s T3hs

T2-3k 3.0s T1 3.0s D 10 km T88-625 W T89-465 E T87-450 T87-620D T88-456 T89-459 T88-462 T93-663 T87-467 T89-470 T88-473 T89-476 T89-478 T89-481 T89-482 T89-484 T96-486 T89-489 T88-493 0.0s 0.0s

1.0s K 1.0s J2x J1b T3h

poor data 2.0s T3hs

T2-3k

Figure 10. Line drawings of four reflection seismic time profiles (seconds are in two-way traveltime) across Turpan-Hami basin (see Fig. 1B for locations). Lines T84-200 (north-south trending) and T88-635 (east-west) are from western portion of basin, near Tabei Sag, and lines T89-465 (north-south) and T88-625 (east-west) are from eastern portion of basin, near Hami. Ancan-1 (total depth 4222 m) and Ha-3 (total depth 3001 m) wells provide stratigraphic control for lines T88-635 and T89-465, respectively. For- mation abbreviations (see Fig. 2 for key) for specific reflectors are indicated where age control is sufficient. Boxed intervals within each well are expanded in Figure 2 with their corresponding palynological age interpretations. Tu-Ha Petroleum Bureau provided seismic ties to wells, based on extensive previous experience with seismic interpretations basinwide. Cross-lines are labeled at top of each seismic line drawing. Note Lower JurassicÐMiddle Jurassic (J1-J2) unconformity that appears on all lines (parts A, B, C, D). Also note intra-Triassic deformation on lines T88-635 (part B) and T88-625 (part D), as well as Permian-Triassic (P2-T1) unconformity near Ancan-1 well (part B). 330 T.J. Greene et al.

~40 m x

Figure 11. A: Angular unconformity at Aiweiergou (westernmost Turpan- Hami) between Upper Permian rocks (P2t) and Lower Triassic red beds (T1j). See Figure 1B for location. B: Line trac- ing interpretation of photo in part A. Squiggle lines denote approximate location of P2-T1 unconformity; truck (highlighted with box) hauling coal provides scale. X indicates sampled for palynology. Middle-Late Permian age is based on presence of Klausipollenites B. schaubergeri and of Alisporites/Falcis- porites cpx. (Abbink, 1999). See Fig- ure 7A (legend) for Lower Jurassic palynological assemblage (section 7A was measured along strike from this exposure). Lower Jurassic

~40 m x Lower Triassic

road Upper Permian

truck

Carroll et al., 1995; Yu et al., 1996). At the Taoshuyuan locality This study presents paleocurrent indicator data from four lo- in northern Turpan-Hami, previous workers noted a reversal calities in the Turpan-Hami basin. Upper Permian cobble clast im- from northward-directed paleocurrent indicators in Permian brication at Aiweiergou suggests east-directed sediment transport strata (Carroll et al., 1995) to southward-directed paleocurrents (Fig. 12A). At Zaobishan, trough cross-bedding in Lower Triassic in Lower Jurassic strata (Hendrix et al., 1992), suggesting an braided-fluvial deposits (Fig. 6) indicates a general northeastward initial uplift of the Bogda Shan between Permian and Early paleocurrent direction, consistent with paleoflow off the ancestral Jurassic time. Yu et al. (1996) reported divergent paleocurrents Tian Shan, south of the Turpan-Hami basin. The Shisanjianfang in Lower Jurassic deposits on either side of the Bogda Shan, im- locality also contains Lower to Middle Triassic braided-fluvial de- plying that the Turpan-Hami and Junggar basins were separated posits with northeastward-directed paleocurrent indicators (Figs. by Early Jurassic time or earlier. However, Shao et al. (1999b) 4E and 12A). These data differ markedly from measurements of presented paleocurrent data from Lin (1993) for both sides of trough cross-bedding from Lower Jurassic fluvial deposits at the Bogda Shan that suggested convergent Lower Jurassic de- Kekeya that indicate transport to the south (Figs. 8C and 12B). positional systems that drained into a Bogda Shan depocenter. Collectively, we interpret these paleocurrent data and data from A. Permian-Triassic

E86 E88 E90 P2w T1 (n = 66) (n = 79) Dalongou5 N44 T2-3 trough axes southern Junggar basin Manas1 (n = 241) (n = 43) 1,2 total Mesozoic 1 Jimsar Qijiagou T1-2 (n = 17)1 2 Permian Urumqi undivided pebble imbrication (n = 12) 2 P2 BO 6 pebble imbrication G Shisanjianfang (n = 10) DA SHAN T2-3 2 6 Taoshuyuan Zaobishan P2 Permian T1 undivided Pd2 trough trough limbs pebble imbrication (n = 51) 6 (n = 48) limbs pebble imbrication Aiweiergou (n = 17) P2d, pebble imbrication (n = 94) trough limbs (n = 48) (n = 143)

Turpan-Hami basin

CHOLTAGH

Legend N42 50 km N paleocurrent references Cenozoic rocks 1 Hendrix et al., 1992 Mesozoic rocks 2 Carroll et al., 1995 and Carroll, 1991 3 Yu et al., 1996 Paleozoic rocks 4 Schneider et al., 1992 E92 Late Paleozoic 5 X. Zhao et al., 1991 granitoid rocks 6 this study

B. Jurassic E86 E88 E90 E92 1 Tianchi 3 N44 J2x, J2t (n = 40) Baiyanghe 1 total Mesozoic Manas 1 southern Junggar basin (n = 241) Xishanyou J3q J1b, pebble imbrication (n = 12)

B 4 J1 OG Toutunhe (n = 46) DA SHAN 1 Taoshuyuan 3 Meiyaogou Kekeya6 3 Baiyang River J1b J1b, trough limbs (n = 11) (n = 34) J1b, foresets J1b, pebble imbrication

Subashigou3 Turpan-Hami basin J1b, pebble imbrication CHOLTAGH

N42

Figure 12. Summary of all paleocurrent indicator data from this study as well as from five other studies (see legend for list of references). Note reversal from north-directed paleocurrents in Permian-Triassic time (A) to south-directed paleocurrents during Jurassic time (B). See text for discussion of paleocurrent indicator data format. 332 T.J. Greene et al. previous reports to reflect northward pre-Jurassic transport in the 1999a; Hendrix, 2000; Fig. 13). We point-counted a small set of Turpan-Hami basin, off the ancestral Tian Shan located to the Triassic sandstone samples (n = 7) from the Turpan-Hami basin, south, followed by a reversal to south-directed paleoflow during using a modified Gazzi-Dickinson method (see Graham et al., Jurassic time as the ancestral Bogda Shan was uplifted. 1993, for detailed techniques; Dickinson and Suczek, 1979; Dickinson, 1985). The raw point-count data (Table 1) were nor- SANDSTONE AND CONGLOMERATE PROVENANCE malized into detrital modes following the methods of Ingersoll et al. (1984) and plotted on standard ternary diagrams (Fig. 13) Provenance studies of Mesozoic sedimentary strata within in order to provide direct comparison with previous provenance the Turpan-Hami basin are an especially powerful tool for pale- studies in the area (e.g., Carroll, 1991; Hendrix, 2000). ogeographic reconstruction, because the Paleozoic rocks ex- Despite the relatively small number of samples counted for posed in the ranges surrounding the Turpan-Hami basin are this study (Table 1), when integrated with data from previous compositionally different from each other (Wen, 1991; Hopson workers, several important trends appear to characterize Per- et al., 1998). The Bogda Shan, located north of the Turpan-Hami mian through Lower Jurassic sandstone samples from the basin, consists almost entirely of intermediate to mafic vol- Turpan-Hami basin. Carboniferous and Permian sandstone con- canics and related plutonic rocks (Chen et al., 1985). Felsic plu- tains abundant volcanic lithic grains and plagioclase feldspar tonic rocks are very minor in modern exposures of the Bogda (Fig. 13; mean Qm4F41Lt55, Qp1Lvm95Lsm4, and Qm10P84K6). Shan (Fig. 1B). In contrast, the central and south Tian Shan We interpret the lithic-rich Carboniferous and Permian samples blocks to the south and west of Turpan-Hami (Fig. 1A) contain to reflect erosional unroofing of volcanic cover strata from the numerous exposures of late Paleozoic granitoids (Wen, 1991; southern Tian Shan, a conclusion supported by north-trending Hopson et al., 1998), emplaced during the final phase of tectonic paleocurrent indicators from these strata (Fig. 12A; Carroll amalgamation of the Tian Shan (Tilton et al., 1986; Coleman, et al., 1990; Shao et al., 1999b). Jurassic sandstone is also rich

1989; Feng et al., 1989; Kwon et al., 1989). In order to use the in lithic-volcanic grains (mean Qm23F21Lt56, Qp12Lvm56Lsm32, composition of Mesozoic sedimentary strata to constrain the ini- and Qm53P27K20), but Hendrix (2000) interpreted these compo- tial timing of uplift of the Bogda Shan, we focused our efforts sitions to reflect erosional unroofing of volcanic strata in the on the northern flanks of the Turpan-Hami basin, located clos- Bogda Shan, consistent with south-directed paleocurrent mea- est to present-day exposures of the Bogda Shan and currently re- surements in the northern Turpan-Hami basin (Fig. 12B). In ceiving erosional detritus from that mountain range (Graham contrast to lithic volcanic-dominated Carboniferous, Permian, et al., 1993). Our rationale was that these localities should have and Jurassic sandstone samples from the Turpan-Hami basin are been the first to receive sediment unequivocally derived from Triassic sandstone compositions containing lower modal per- the Bogda Shan. In our provenance studies, outlined in the fol- centages of total lithic grains (Lt) and higher percentages of lowing, we use sandstone petrography and conglomerate clast potassium feldspar (mean Qm29F29Lt42, Qp23Lvm49Lsm28, and counts to infer that Lower Triassic deposits were derived solely Qm51P25K24). Greene et al. (1997) interpreted these composi- from the central and south Tian Shan blocks, south of the Turpan- tional characteristics to reflect Triassic erosional unroofing of Hami basin, and that the ancestral Bogda Shan had not been late Paleozoic granites in the central and south Tian Shan, con- uplifted by that time. sistent with northeastward-directed Triassic paleocurrent indicators (Fig. 12A). Shao et al. (1999a) described a simple Sandstone framework grains unroofing history based on point counts of Turpan-Hami sandstones that showed mean values in QFLt%Lt decreasing and Sandstone point-count data from this study are compared QFLt%Q increasing from Permian through Tertiary time; no- with similar data from upper Paleozoic and Mesozoic sandstone tably, they showed no increase in quartz or feldspar content for from the Turpan-Hami basin (Carroll et al., 1995; Shao et al., Triassic samples. Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 333

Q Qm m Figure 13. Summary of sandstone provenance point-count data of Turpan-Hami basin. Means (closed squares) and standard deviation

Continental Recycled Jurassic fields (polygons; 1s) are shown for Block Orogen (n = 6) each time slice. Global provenance suites proposed by Dickinson and Circum-Pacific Volcanoplutonic Triassic Suczek (1979) are provided for (n = 7) comparison. Qm, monocrystalline Jurassic quartz; F, total feldspars; Lt, to- Triassic (n = 6) (n = 7) tal lithic grains; K, potassium Magmatic feldspar; P, plagioclase; Qp, poly- Arc crystalline quartz + chert; Lvm, F Carboniferous-Permian L P Carboniferous-Permian K volcanic and metavolcanic lithics; (n = 4) t (n = 4) Lsm, sedimentary + metamorphic lithics. All Mesozoic data, except Q Triassic data from Turpan-Hami p basin (this study), are from Hendrix (2000); all Paleozoic data are from Carroll (1991). Although data are too few to be statistically significant, they are sufficient to define subduction complex broad temporal and regional Collision suture, changes in sandstone composition. fold-thrustbelt Note relative enrichment in Triassic QmFLt%Qm and QmPK%K for (n = 7) Triassic sandstone as well as overall enrichment in Qm for Mesozoic Carboniferous-Permian sandstone relative to Paleozoic (n = 4) Jurassic sandstone. L (n = 6) vm Arc Orogen Lsm

Conglomerate data High percentages of felsic plutonic clasts in Lower Triassic deposits at Zaobishan contrast sharply with Permian and Juras- We conducted conglomerate clast lithology counts at four sic conglomerate containing few or no felsic plutonic clasts localities along the northern basin rim: Taoxigou (west (Fig. 14). We interpret this to reflect deeper erosion and unroof- Taoshuyuan), Aiweiergou, Zaobishan, just north of the town of ing of exposed Paleozoic granites in the central and south Tian Shisanjianfang, and Kekeya (Figs. 1B and 14). Permian con- Shan, consistent with sandstone compositions and northward- glomerate at Taoxigou consists mainly of limestone clasts most directed paleocurrents. Granitic clasts are also very rare in likely derived from the underlying marine Upper Carboniferous Upper Permian strata at Aiweiergou and Zaobishan, where con- deposits (Carroll et al., 1995). At Aiweiergou, Upper Permian glomerate is dominated by intermediate volcanic clasts most conglomerate (P2d) consists mainly of intermediate and mafic likely derived from erosion of Carboniferous andesitic cover volcanic clasts. At Zaobishan, conglomerate clasts in Lower Tri- strata (Carroll et al., 1995). assic fluvial deposits contain anomalously high percentages of pink granitic and aplitic dike rock cobbles (32%), whereas DISCUSSION Middle-Upper Triassic conglomerate is dominated by vein quartz and mafic and intermediate volcanic compositions Tectonic setting of the North Tian ShanÐBogda Shan block (Fig. 14). At Shisanjianfang, Middle-Upper Triassic clast lithology is dominantly intermediate and mafic and felsic volcanics Details regarding the Paleozoic tectonic setting of the North as well as granitic clasts. Lower Jurassic conglomerate at Tian ShanÐBogda Shan (NTS/BS) block are uncertain, yet have Kekeya has a well-mixed population of clast lithologies, al- obvious implications for the subsequent fill of the Turpan-Hami though granitic clasts are notably absent. Yu et al. (1996) also basin. For example, if the Bogda Shan existed as an active island reported conglomerate clast percentages for Lower Jurassic arc until the Early Permian, as has been proposed by various au- strata at Baiyang River and Taoshuyuan: high proportions of thors (Hsü, 1988; Coleman, 1989; Wang et al., 1990;Allen et al., volcanic clasts (80% and 95%, respectively) and low percent- 1991, 1992; Fang, 1994; Mu, 1994), then it is unlikely that the ages of granitic clasts (0% and 15%, respectively). Turpan and Junggar basins were depositionally linked during 334 T.J. Greene et al.

LEGEND: CLAST TYPES

vein quartz felsic volcanic sedimentary intermediate-mafic plutonic

chert intermediate/mafic volcanic felsic plutonic unknown

Formation Conglomerate clast counts Location (Formation) Age number of counts Xishanyao (J2x)

Songonhe (J1s) Kekeya (J1b; section 7C) Jurassic Badaowan (J1b)

Lower/Middle n = 119

Haojagou (T3h)

Huangshanja (T3hs) Zaobishan (T2-3k; section 4C) Middle/ Upper Karamay (T2-3k) n = 125 Shisanjianfang (T2-3k; section 4E)

Triassic n = 104 Zaobishan Shaofangou (T1s) (T1s; section 4D) Lower Jiucaiyuan (T1j) n = 202

Guodikeng (P2g) Wutonggou (P2w) Quanzijie (P2q) Taierlong (P2t) Aiweiergou (P2d) Daheyan (P2d) n = 113 Permian Taoxigou (P1t) Taoxigou:

Lower Upper west Taoshuyuan (P1t) n = 115 Carbon- basement iferous 0% 50% 100%

Figure 14. Composite stratigraphic chart with summary of conglomerate clast count data collected in Turpan-Hami basin. Although mostly mafic to intermediate clast compositions are dominant, felsic plutonic clasts in Lower Triassic conglomerate (T1s) at Zaobishan (measured section 4D) are higher than in underlying or overlying units.

Late Permian time. However, this scenario is in direct conflict unified Junggar-Turpan-Hami lake system (Liu et al., 1979; with concordant, north-directed paleocurrent indicators in the Taner et al., 1988; Nishaidai and Berry, 1991; Carroll et al., Turpan-Hami and Junggar basins (Fig. 12A) and isotopic prove- 1992; Ren et al., 1994; Brand et al., 1993; Wang et al., 1996; nance data that support post-Permian uplift of the Bogda Shan Greene, 1997; Wartes et al., 1998, 1999, 2000). (Greene et al., 1997; Greene and Graham, 1999). Moreover, on Many Chinese authors refer to the Bogda Shan as either a the basis of facies relations and organic geochemical attributes, late Paleozoic intracontinental rift belt disrupting a DevonianÐ many authors have reported contemporaneous Upper Permian Middle Carboniferous passive margin (Huang et al., 1991; organic-rich lacustrine deposits on both sides of the Bogda Shan Fang, 1990; Chen, 1993; Lin, 1993; Wang et al., 1996; Wu et al., in the southern Junggar and Turpan-Hami basins, implying a 1996; Wu and Xue, 1997; Wu and Zhao, 1997; Y. Liu et al., Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 335

1998), or as a focal point of extension due to mantle diapirism Upper Carboniferous rocks, which are in turn overlapped by beneath the Bogda Shan (Zhu and Zhao, 1992; Tao, 1994). Upper Permian lacustrine rocks. In addition, on the basis of These works, however, are largely conceptual in nature and pre- stratigraphic and sedimentologic relationships, Carroll et al. sent no outcrop, subsurface, or geochemical data that support a (1999) reported a large north-southÐtrending Lower Permian late Paleozoic Bogda rift. normal fault with as much as 3 km of displacement. Both fea- Others agree that oceanic crust of unknown width and an tures are interpreted as indicators of Early Permian extension in associated island arc complex, termed the North Tian a direction normal to earlier Devonian-Carboniferous east- ShanÐBogda Shan block, was being subducted under the north- westÐtrending compressional features. ern margin of the Central Tian Shan block (also referred to as the Yili microcontinent) during Carboniferous time; arc-related Paleogeographic models magmatism ceased during the Late Carboniferous (Z. Wang et al., 1986; C. Wang et al., 1990; Zhou, 1987; Coleman, 1989; On the basis of the results of our study, along with data and Hopson et al., 1989; Carroll et al., 1990, 1995; Windley et al., interpretations by other workers, we offer the following sce- 1990; Allen et al., 1991, 1992; Wen, 1991; S¸engör et al., 1993; nario for the paleogeographic and tectonic evolution of the Z. Cheng et al., 1996; Wu et al., 1996; Wu and Zhao, 1997). Turpan-Hami basin (Fig. 15). The precise temporal and spatial distribution and the polarity Late Permian. Shortening and folding of Upper Permian of subduction represented by the North Tian ShanÐBogda strata occurred in the western and southern parts of Turpan- Shan volcanic rocks, however, are uncertain (see Carroll et al., Hami (Fig. 15A). Subsequent latest PermianÐearliest Triassic 1990, 1995; Windley et al., 1990; Allen et al., 1991, for discus- uplift of the Central Tian Shan resulted in beveling of Upper sion of possible paleogeographic scenarios). Carroll et al. Permian strata, and deposition of Lower Triassic strata, as (1995) interpreted 1 km of shallow-marine deposits and demonstrated in outcrop at Aiweiergou (Fig. 11) and by seis- andesitic and dacitic volcanic flow rocks in the south and mic line T88-635 (Fig. 10B). Zhou (1997) and Dumitru et al. southwest Bogda Shan as representing an emergent Late Car- (this volume) inferred a period of rapid Permian-Triassic cool- boniferous Bogda Shan island arc. If arc-related magmatism ing in the Central Tian Shan, farther to the west, based on shut off soon after, this allows a time span of 20Ð30 m.y. for apatite fission-track data. Notably, the northern part of Turpan- any topographic highs to be eroded before Late Permian depo- Hami, as well as a southern Junggar locality (Jimsar), contain sition. We infer that such an episode of erosion is manifested in continuous Permian through Triassic sections (Fig. 1B; Yang et the basinwide CarboniferousÐLower Permian angular uncon- al., 1986; Liao et al., 1987; X. Zhao et al., 1991; Tang et al., formity observed throughout the Turpan-Hami basin (Fig. 1B; 1997), suggesting that these more distal localities were not af- Liao et al., 1987; Carroll et al., 1999). This interpretation per- fected by the deformation. mits the Bogda Shan to be a part of a Late Carboniferous island Several lines of evidence point to continuous Permian de- arc chain without being a physiographic barrier between the position across the Turpan-Hami and Junggar basins. Wartes Turpan and southern Junggar basin during Late Permian and et al. (1998, 1999, 2000) reported the existence of a large Late Early Triassic time. Permian Junggar-Turpan-Hami lake system represented by a Allen et al. (1995) proposed that PermianÐTriassic mag- complex association of lake marginal and basinal facies on both matism in the Turpan-Hami basin was the result of transten- sides of the Bogda Shan. There are other reports of widespread sional rotation within a Late PermianÐTriassic sinistral shear Upper Permian lacustrine deposition within at least four main system. In the Turpan-Hami basin, they based this on mapped depocenters in the Turpan-Hami basin; however, no sedimento- (but undated) Lower Permian mafic volcanic rocks and logic or stratigraphic data have been presented (Allen et al., tholeitiic dike rocks intruding late Paleozoic turbidites. Gab- 1993, 1995; Mu, 1994; Wang et al., 1996; Wu and Zhao, 1997). broic bodies in the western Bogda Shan were interpreted as evi- Paleocurrent indicators suggest sediment dispersal from south dence of Permian magmatism by Windley et al. (1990) and to north, through the present-day Bogda Shan (Fig. 12A; Allen et al. (1991, 1995). However, an alternative interpretation Carroll, 1991; X. Zhao et al., 1991; Hendrix et al., 1992; Carroll is that the mapped gabbroic bodies of Chen et al. (1985) are hy- et al., 1995). Coeval sandstone and conglomerate is extremely pabyssal intrusives from diorite-trondhjemite magmas that are volcanic rich, indicating erosion of the extinct Carboniferous typical of arc magmatism (Clifford Hopson, 1999, personal arc (Figs. 13 and 14). commun.). These rocks were analyzed by Hopson et al. (1989) Although we infer a proximal foreland setting for the and yielded a U-Pb radiometric age of 328 ; 10 Ma (Carbonif- Upper Permian deposits in Figure 15A, isopach trends for Per- erous); confirming the pre-Permian arc-related history of the mian strata are ambiguous, given the currently available data. western Bogda Shan (Carroll et al., 1995). Wu and Guo (1991) published north-south seismic lines be- Carroll et al. (1999) reported several north-southÐtrending tween Urumqi and Jimsar indicating northward thinning of mafic dike rocks in the Lower Permian Taoxigou Group at the Upper Permian deposits; in contrast, Carroll et al. (1992, 1995) Taoshuyuan locality (Fig. 1B). Their inferred Early Permian age and Wartes et al. (2000) reported Upper Permian thicknesses in derives from crosscutting relationships with shallow-marine southern Junggar that greatly exceed those of Turpan-Hami. Late Permian A.

Z' JUNGGAR-TURPAN-HAMI LAKE Z Z' x Junggar-Turpan-Hami lake

x x extinct extinct NTS x x NTS island arc island arc x x x x +++ NTS Fault CTS "suture zone" NTS-BS extinct island arc

CENTRAL TIAN SHAN extinct NTS island arc Z Triassic B. Z' Z' Z x x x x xx x

granitic clasts NTS Fault + ++ Tian Shan x x wedge-top foredeep x x x x x fold-thrust belt x x x Z CENTRAL TIAN SHAN x x

Early-Middle Jurassic C.

Z' Z' Z x x x x

Bogda Shan ++

NTS Fault Tian Shan fold-thrust belt proto Bogda Shan x x x x x x x x x x x x coarse clastics Central Tian Shan (CTS) arc unconformity CENTRAL TIAN SHAN x Z North Tian Shan (NTS) and paleocurrent sandy clastics Bogda Shan (BS) island arc direction x + + arc/within plate granites 100 km N fine clastics x + fault Tian Shan mafic dikes reverse/normal no vertical scale fold-thrust belt fault Figure 15. Schematic, nonpalinspastic paleogeographic model of study area from Late Permian to Early-Middle Jurassic time. A: Subsequent to Carboniferous collision between the Central Tian Shan arc and the North Tian ShanÐBogda Shan island arc, north-vergent Tian Shan fold-thrust belt began shedding debris into contiguous Turpan-Hami-Junggar basin. Large Late Permian lacustrine system spanned most of basin, with sediment dispersal generally directed north. During this time, Bogda Shan had not been uplifted (its present-day position is outlined). Greene and Graham (1999) also reported existence of Permian within-plate granites (cf. Pearce et al., 1984) in present-day Bogda Shan. See Wartes et al. (2000) and Carroll et al. (1995) for more detailed descriptions of Permian paleogeography. B: Triassic time was dominated by coarse-clastic deposition as compression in north-vergent Tian Shan fold-thrust belt continued. Several intra-Triassic unconformities in Turpan-Hami basin suggest repeated uplift and erosion of Tian Shan, although there is no break in sedimentation between Permian and Triassic time in southern Junggar basin. Dissection of Tian Shan fold-thrust belt unroofed Central Tian Shan granitoid rocks (e.g., T1s deposits at Zaobishan; Fig. 14) as sediment dispersal was still directed to north (Greene et al., 1997). C: By Early Jurassic time, shortening of Tian Shan fold-thrust belt continued, and Bogda Shan became significant physiographic feature partitioning southern Junggar basin from Turpan-Hami. As thrusting initiated in Bogda Shan area, amalgam of faulted Carboniferous island-arc basement, recycled Permian and Triassic deposits, mafic dike rock, and within-plate granites became structurally displaced to create proto-Bogda Shan. Paleocurrents diverged away from Bogda Shan as sediment was ponded within intermontane Turpan-Hami basin. Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 337

However, if we assume that the Turpan-Hami basin was in a based on total thickness of Lower and Middle Jurassic strata wedge-top position relative to a northward-vergent Tian Shan (provided by the Tu-Ha Petroleum Bureau), we infer a shift in fold-thrust belt, then basinward stratal thickening of Upper Per- Jurassic depocenters from the western Tokesun Sag to the north- mian deposits toward the southern Junggar basin would be en- central Taibei Sag (Fig. 1B). Notably, many other studies report tirely permissible (cf. DeCelles and Giles, 1996). This would a widespread Triassic-Jurassic unconformity throughout the explain the abundant regional and local unconformities near the Turpan-Hami basin, although no age control data are presented orogenic wedge, the coarseness and extreme immaturity of Up- (Wang et al., 1994; D. Li, 1995; W. Li, 1997; Qiu et al., 1997; per Permian deposits, localized lacustrine deposits that possibly Wu and Xue, 1997; Wu and Zhao, 1997). formed in isolated piggy-back basins, and the northward-thick- Southward-directed paleocurrent indicators in the northern ening of Upper Permian strata toward the Junggar foredeep Turpan-Hami basin contrast with northward-directed indica- (Fig. 15A). tors in the southern Junggar basin and thus suggest divergent Triassic. An abrupt change to coarser, more fluvial- sediment-dispersal patterns on both sides of the Bogda Shan for alluvialÐdominated environments characterizes most of the Tri- Lower Jurassic strata (Fig. 12B; Hendrix et al., 1992; Schneider assic in the Turpan-Hami basin (Figs. 4 and 15B). As the central et al., 1992; Yu et al., 1996). Jurassic sandstone compositions and south Tian Shan became more dissected, felsic plutons were contain higher percentages of lithic-volcanic detritus, reflecting unroofed and contributed erosional detritus to Lower Triassic Jurassic erosion of the Bogda Shan (Fig. 13). Hendrix (2000) fluvial deposits across the Turpan-Hami basin. The T1-T2-3 noted increased percentages of radiolarian chert from Jurassic unconformity observed at Zaobishan (Fig. 4B), Taodongou samples on the northern and southern flanks of the Bogda Shan (Figs. 3B and 4A), and in seismic line T88-635 (Fig. 10B) could and suggested that the chert may have been derived from the reflect deformation due to renewed uplift of the Tian Shan to Jurassic unroofing of small, intraarc basins within the range. In the south, although this idea remains to be further tested. At a addition, Chinese facies maps show Jurassic internal drainage more regional scale, Hendrix et al. (1992) inferred an episode patterns confined by the Bogda Shan to the north and the Tian of Late Triassic uplift of the Tian Shan that resulted in large Shan to the south (Huang et al., 1991; Qiu et al., 1997; Wu and coarse-clastic deposition in the southern Junggar and northern Zhao, 1997), suggesting that by this time, the Turpan-Hami Tarim basin. basin had become a single depositional basin, separated from Conglomerate and sandstone provenance data also sup- the southern part of the Junggar basin. port Triassic unroofing of the Tian Shan via an increase of K-feldsparÐrich felsic clasts and decrease in total lithic grains in Lower and Middle Triassic braided-fluvial deposits (Figs. 14 CONCLUSIONS and 15). Based on isotopically enriched eNd initial values of these felsic clasts, Greene and Graham (1999) inferred a Cen- 1. Uplift of the Bogda Shan occurred later than Early Triassic tral Tian Shan provenance. In addition, paleocurrent indicators time. The composition of sandstone and conglomerate in at Zaobishan and Shisanjianfang are directed north to northeast Lower Triassic strata on the north flanks of the Turpan-Hami away from the ancestral Tian Shan (Fig. 12A). The paucity of basin suggests derivation from late Paleozoic granite in the granitic pebbles in modern drainages coming off the Bogda Tian Shan to the south. This remains to be tested for south- Shan at Zaobishan and Taoxigou (west Taoshuyuan) is further ern Junggar Lower Triassic deposits. Because volumetrically circumstantial evidence supporting a non-Bogda Shan Triassic significant granites are absent in the Bogda Shan to the north, provenance for the northern Turpan basin. All of these data to- we infer that the Bogda Shan was not being erosionally un- gether imply that the Bogda Shan was not a significant positive roofed during Early Triassic time. physiographic feature during Triassic time. 2. Pre-Early Jurassic initial uplift of the Bogda Shan is sup- Early Jurassic through Middle Jurassic. Our analysis of ported by paleocurrent indicators from Lower Jurassic strata Jurassic strata reveals a distinct change in depositional style in that are north directed north of the Bogda Shan but south di- Lower Jurassic deposits (Figs. 7 and 15C) represented by an in- rected south of the Bogda Shan. Pre-Jurassic uplift of the crease in swampy, lacustrine, and meander-belt systems. A ma- Bogda Shan is also supported by lithic-volcanicÐrich Juras- jor reorganization of basin physiography and possibly the onset sic sandstone exposed along the northern and southern flanks of internal drainage in the basin is implied by the existence of a of the Bogda Shan, inferred to have been derived from un- J1-J2 unconformity across much of the basin (Fig. 10). Figure roofing of Carboniferous arc-related volcanics in the ances- 10 shows four examples of seismic lines where truncated Lower tral mountain range. Jurassic reflectors are overlapped by Middle Jurassic reflectors. 3. Regional seismic reflection lines and outcrop studies indi- Our palynological studies provide age control both above and cate several intra-Mesozoic shortening events affected the below the unconformity in the Ha-3 well (Figs. 2 and 10C); Turpan-Hami basin. We interpreted angular unconformities Jurassic age control for seismic lines T84-200 and T88-635 of Upper PermianÐLower Triassic (P2-T1), Lower-Middle- (Fig. 10, A and B) was provided by the Tu-Ha Petroleum Bureau Upper Triassic (T1-T2-3), Lower-Middle Jurassic (J1-J2), for the Naiqian-4, Naiqian-5, and Ancan-1 wells. In addition, and Jurassic-Cretaceous (J-K) age, indicating that the 338 T.J. Greene et al.

intra-Mesozoic shortening recorded in other basins of western Allen, M.B., Windley, B.F., Zhang, C., Zhao, Z.Y., and Wang, G.R., 1991, Basin China also affected physiography in the Turpan-Hami basin. evolution within and adjacent to the Tien Shan Range, northwest China: 4. Continued uplift of the Bogda Shan during Early Jurassic Geological Society of London Journal, v. 148, p. 369Ð379. Allen, M.B., Windley, B.F., and Zhang, C., 1992, Paleozoic collisional tecton- time caused a major reorganization of depositional systems ics and magmatism of the Chinese Tian Shan, central Asia: Tectono- in the Turpan-Hami basin and erosional beveling of Lower physics, v. 220, p. 89Ð115. Jurassic and older strata. Jurassic depositional systems in the Allen, M.B., Windley, B.F., Zhang, C., and Guo, J., 1993, Evolution of the Tur- Turpan-Hami basin were inundated with the supply of sedi- pan basin, Chinese central Asia: Tectonics, v. 12, p. 889Ð896. ment derived from erosion of the Bogda Shan, resulting in a Allen, M.B., S¸engör, A.M.C., and Natal’in, B.A., 1995, Junggar, Turfan, and Alakol basins as Late Permian to ?Early Triassic extensional structures basinwide J1-J2 angular unconformity. This feature could in a sinistral shear zone in the Altaid orogenic collage, central Asia: have been associated with a basinwide shift of Jurassic de- Geological Society of London Journal, v. 152, p. 327Ð338. pocenters, from Early Jurassic deposition principally in the Brand, U., Yochelson, E.L., and Eagar, R.M., 1993, Geochemistry of Late Per- western depression (Tokesun Sag) to Middle Jurassic depo- mian non-marine bivalves: Implications for the continental paleo- sition principally in the north-central depression (Tabei Sag). hydrology and paleoclimatology of northwestern China: Carbonates and Evaporites, v. 8, p. 199Ð212. 5. The post-Early Triassic constraints on initial uplift of the Campbell, J.E., and Hendry, H.E., 1987, Anatomy of a gravelly meander lobe in Bogda Shan described in this chapter suggest that organic- the Saskatchewan River, near Nipawin, Canada, in Ethridge, F.G., et al., rich Upper Permian lacustrine deposits in the Turpan-Hami eds., Recent developments in fluvial sedimentology: Society of Eco- basin were likely contiguous with voluminous organic-rich nomic Paleontologists and Mineralogists Publication 39, p. 179Ð189. lacustrine deposits in the southern and central Junggar basin. Carroll, A.R., 1991, Late Paleozoic tectonics, sedimentation, and petroleum potential of the Junggar and Tarim basins, northwest China [Ph.D. thesis]: Because these deposits are the primary petroleum source Stanford, California, Stanford University, 405 p. rock in the Junggar basin, the suggestion that their updip Carroll, A.R., 1998, Upper Permian lacustrine organic facies evolution, southern equivalents are present in the Turpan-Hami basin increases Junggar Basin, northwest China: Organic Geochemistry, v. 28, p. 649Ð667. the petroleum source-rock potential of that basin. Carroll, A.R., Graham, S.A., Hendrix, M.S., Chu, J., McKnight, C.L., Xiao, X., and Liang, Y., 1990, Junggar basin, northwest China: Trapped late Paleo- zoic Ocean: Tectonophysics, v. 181, p. 1Ð14. ACKNOWLEDGMENTS Carroll, A.R., Brassell, S.C., and Graham, S.A., 1992, Upper Permian lacustrine oil shales, southern Junggar basin, northwest China: American Associa- tion of Petroleum Geologists Bulletin, v. 76, p. 1874Ð1902. The Chinese National Petroleum Corporation and the Tur- Carroll, A.R., Graham, S.A., and Hendrix, M.S., 1995, Late Paleozoic amalga- pan-Hami Petroleum Bureau arranged invaluable logistical and mation of northwest China: Sedimentary record of the northern Tarim, technical support as well as access to outcrop and subsurface northwestern Turpan, and southern Junggar basins: Geological Society data. K. Cheng, X. Zeng, W. Wang, T. Hu, A. Su, and X. Zhang of America Bulletin, v. 107, p. 571Ð594. assisted greatly in interpreting the stratigraphy of the Turpan- Carroll, A.R., Wartes, M.A., and Greene, T.J., 1999, Sedimentary evidence for Early Permian normal faulting, southern Bogda Shan, northwest China: Hami basin as well as providing access to key regional seismic Geological Society of America Abstracts with Programs, v. 31, no. 7, lines and wells. A. Hessler provided outstanding field assistance p. A369ÐA370. during the 1997 season. Discussions with G. Ernst, A. Hanson, Chen, M., 1993, Structural styles of the Turpan-Hami basin: Petroleum Explo- C. Hopson, J. Hourigan, L. Hsiao, C. Johnson, M. McWilliams, ration and Development, v. 20, no. 5, p. 1Ð7. B. Ritts, X. Ying, Y. Yue, and D. Zhou improved this manuscript. Chen, Z., Wu, N., Zhang, D., Hu, J., Huang, H., Shen, G., Wu, G., Tang, H., and Hu, Y., 1985, Geologic map of Xinjiang Uygur Autonomous Region: We are grateful to C. Cooper, M. Allen, S. Vincent, and T. Law- Beijing, Geologic Publishing House, scale 1:2 000 000. ton for their constructive reviews. Acknowledgment is made to Cheng, K., Su, A., Zhao, C., and He, Z., 1996, Study of coal-generated oil in the the donors of The Petroleum Research Fund, administered by Tuha basin: 30th International Geological Congress, Progress in geol- the American Chemical Society, for partial support of this re- ogy of China: Beijing, Geological Publishing House, p. 796Ð799. search (ACS-PRF32605-AC2). Additional funding came from Cheng, Z., Wu, S., and Fang, X., 1996, The Permian-Triassic sequences in the southern margin of the Junggar basin, and the Turpan basin, Xinjiang, the American Association of Petroleum Geologists and the Stan- China, in International Geologic Congress, 30th, Field Trip Guidebook ford University McGee Fund. Financial support during the term T394: Beijing, Geological Publishing House, 25 p. of this research project was also provided by the Graduate Clayton, J.J., Yang, J., King, J.D., Lillis, P.G., and Warden, A., 1997, Geo- School of the University of Wisconsin and the Stanford-China chemistry of oils from the Junggar basin, northwest China: American Industrial Affiliates, an industrial consortium that has included Association of Petroleum Geologists Bulletin, v. 81, p. 1926Ð1944. Coleman, R.G., 1989, Continental growth of northwest China: Tectonics, v. 8, Agip, Arco, Chevron, Exxon, Japan National Oil Corporation, p. 621Ð636. Mobil, Phillips, Shell, Statoil, Texaco, Triton, and Unocal. DeCelles, P.G., and Giles, K.A., 1996, Foreland basin systems: Basin Research, v. 8, p. 105Ð123. DeCelles, P.G., Langford, R.P., and Schwartz, R.K., 1983, Two new methods of REFERENCES CITED paleocurrent determination from trough cross-stratification: Journal of Sedimentary Petrology, v. 53, p. 629Ð642. Abbink, O.A., 1999, Palynology of outcrop and core samples from the Turpan- Dickinson, W.R., 1985, Interpreting provenance relations from detrital modes Hami basin, northwest China: Utrecht, The Netherlands, Laboratory of of sandstone, in Zuffa, G.G., eds., Provenance of arenites: Hingham, Palaeobotany and Palynology Report 9916, 18 p. Massachusetts, D. Reidel Publishing Company, p. 333Ð361. Sedimentary record of Mesozoic deformation and inception of the Turpan-Hami basin 339

Dickinson, W.R., and Suczek, C.A., 1979, Plate tectonics and sandstone com- Kwon, S.T., Tilton, G.R., Coleman, R.G., and Feng, Y., 1989, Isotopic studies positions: American Association of Petroleum Geologists Bulletin, v. 63, bearing on the tectonics of the west Junggar region, Xinjiang, China: p. 2164Ð2182. Tectonics, v. 8, p. 719Ð727. Fang, G., 1990, Initial studies about Bogda Mountains late Palaeozoic aulaco- Li, D., 1995, Hydrocarbon occurrences in the petroliferous basins of western gen: Xinjiang Geology, v. 8, p. 133Ð141. China: Marine and Petroleum Geology, v. 12, p. 26Ð34. Fang, G., 1994, Paleozoic plate tectonics of eastern Tianshan mountains Xin- Li, G., and Shen, S., 1990, On the formation and evolution of Turpan-Hami jiang, China: Acta Geologica Gansu, v. 3, p. 34Ð40. basin and its characteristic of bearing oil and gas, Xinjiang: Bulletin of Feng, Y., Coleman, R.G., Tilton, G., and Xiao, X., 1989, Tectonic evolution of Xi’an Institute Geological and Mineral Resources no. 28: Beijing, Chi- the west Junggar region, Xinjiang, China: Tectonics, v. 8, p. 729Ð752. nese Academy of Geological Sciences, p. 25Ð36. Graham, S.A., Hendrix, M.S., Wang, L.B., and Carroll, A.R., 1993, Collisional Li, J., and Jiang, J., 1987, Survey of petroleum geology and the controlling fac- successor basins of western China: Impact of tectonic inheritance on sand tors for hydrocarbon distribution in the east part of the Junggar basin: composition: Geological Society ofAmerica Bulletin, v. 105, p. 323Ð344. Oil and Gas Geology, v. 8, p. 99Ð107. Greene, T.J., 1997, Petroleum geochemistry of Upper Permian and Middle Juras- Li, W., 1997, Sequence stratigraphy of Jurassic in Taibei Sag, Turpan-Hami sic source rocks and oils of the Turpan-Hami basin, northwest China: basin: Oil and Gas Geology, v. 18, p. 210Ð215. American Association of Petroleum Geologists Bulletin, v. 81, p. 1774. Liao, Z., Lu, L., Jiang, N., Xia, F., Sung, F., Zhou, Y., Li, S., and Zhang, Z., 1987, Greene, T.J., and Graham, S.A., 1999, Isotopic provenance of Devonian-aged Carboniferous and Permian in the western part of the east Tianshan granitic cobbles and geochemistry of Bogda Shan granitoids marking Mountains: Beijing, Eleventh Congress of Carboniferous Stratigraphy the inception of the Turpan-Hami basin, northwest China: Geological and Geology, Guide Book Excursion 4, 50 p. Society of America Abstracts with Programs, v. 31, no. 7, p. A374. Lin, J.Y., 1993, Sedimentary sequence of the Bogda Rift: Discussion about the Greene, T.J., Carroll, A.R., Hendrix, M.S., and Li, J., 1997, Permian-Triassic formation and evolution of the entire intercontinental sedimentary basin evolution and petroleum system of the Turpan-Hami Basin, Xin- basin in the northern Xinjiang [Ph.D. thesis]: Xi’an, Northwest Uni- jiang Province, northwest China: American Association of Petroleum versity, 104 p. Geologists and Society of Economic Paleontologists and Mineralogists Liu, H., 1986, Geodynamic scenario and structural styles of Mesozoic and Annual Meeting Abstracts, v. 6, p. 42. Cenozoic basins in China: American Association of Petroleum Geolo- Greene, T.J., Carroll, A.R., Hendrix, M.S., Cheng, K., and Zeng, X.M., 2000, gists Bulletin, v. 70, p. 377Ð395. Sedimentology and paleogeography of the Middle Jurassic Qiketai For- Liu, H., Lian, H., Cai, L., Xia, Y., and Liu, L., 1979, Evolution and structural mation, Turpan-Hami basin, northwest China, in Gierlowski-Kordesch, style of Tian Shan and adjacent basins, northwestern China: Earth Sci- E., and Kelts, K., eds., Lake basins through space and time: American ence, v. 19, p. 727Ð741. Association of Petroleum Geologists Studies in Geology 46, p. 141Ð152. Liu, L., and Di, S., 1997, Characteristics of Middle Jurassic sedimentation and Hendrix, M.S., 2000, Evolution of Mesozoic sandstone compositions, south- reservoir pore evolution in Turpan Depression: Oil and Gas Geology, ern Junggar, northern Tarim, and western Turpan basins, northwest v. 18, p. 17Ð18. China: A detrital record of the ancestral Tian Shan: Journal of Sedi- Liu, L., Liu, Y., and Di, S., 1998, Sedimentation and diagenesis of Qiketai Forma- mentary Research, v. 70, p. 520Ð532. tion, in north Turpan Depression: Oil and Gas Geology, v. 19, p. 238Ð243. Hendrix, M.S., Graham, S.A., Carroll, A.R., Sobel, E.R., McKnight, C.L., Liu, Y., Wu, T., Cui, H., and Feng, Q., 1998, Paleogeothermal gradient and geo- Schulein, B.J., and Wang, Z., 1992, Sedimentary record and climatic im- logic thermal history of the Turpan-Hami basin, Xinjiang: Science in plications of recurrent deformation in the Tian Shan: Evidence from China, ser. D, v. 41, p. 62Ð68. Mesozoic strata of north Tarim, south Junggar, and Turpan basins, north- Miall, A.D., 1996, The geology of fluvial deposits: Sedimentary facies, basin west China: Geological Society of America Bulletin, v. 105, p. 53Ð79. analysis, and petroleum geology: Berlin, Springer-Verlag, p. 208Ð211. Hendrix, M.S., Graham, S.A., Amory, J.A., and Badarch, G., 1996, Noyon Uul Miall, A.D., and Gibling, M.R., 1978, The Siluro-Devonian clastic wedge of syncline, southern Mongolia: Lower Mesozoic sedimentary record Somerset Island, Arctic Canada, and some regional paleogeographic im- of the tectonic amalgamation of central Asia: Geological Society of plications: Sedimentary Geology, v. 21, p. 85Ð127. America Bulletin, v. 108, p. 1256Ð1274. Mu, Z., 1994, Permian and Triassic formation distribution and palaeogeo- Hopson, C., Wen, J., Tilton, G., Tang, Y., Zhu, B., and Zhao, M., 1989, Paleo- graphical pattern of Turpan-Hami basin: Xinjiang Petroleum Geology, zoic plutonism in east Junggar, Bogdashan, and eastern Tianshan, v. 14, p. 14Ð20. northwest China: Eos (Transactions, American Geophysical Union), Nanson, G.C., and Croke, J.C., 1992, A genetic classification of flood plains: v. 70, p. 1403Ð1404. Geomorphology, v. 4, p. 459Ð486. Hopson, C.A., Wen, J., and Tilton, G.R., 1998, Isotopic variation of Nd, Sr, and Nijman, W., and Puigdefábregas, C., 1978, Coarse-grained point bar structure Pb in Paleozoic granitoid plutons along an east Junggar-Bogdashan- in a molasse-type fluvial system, Eocene Castisent sandstone formation, Tianshan transect, northwest China: Geological Society of America Ab- south Pyrenean Basin, in Miall, A.D., ed., Fluvial sedimentology: Cana- stracts with Programs, v. 30, no. 5, p. 20. dian Society of Petroleum Geologists Memoir 5, p. 487Ð510. Hsü, K.J., 1988, Relict back-arc basins: Principles of recognition and pos- Nishidai, T., and Berry, J.L., 1991, Geological interpretation and hydrocarbon sible new examples from China, in Kleinspehn, K.L., and Paola, C., potential of the Turpan basin (NW China) from satellite imagery: Jour- eds., New perspectives in basin analysis: New York, Springer-Verlag, nal of Petroleum Technology, v. 13, p. 35Ð58. p. 245Ð263. Olsen, H., 1989, Sandstone-body structures and ephemeral stream processes in Huang, D., Zhang, D., Li, J., and Huang, X., 1991, Hydrocarbon genesis of the Dinosaur Canyon Member, Moenave Formation (Lower Jurassic), Jurassic coal measures in the Turpan Basin, China: Organic Geochem- Utah, U.S.A.: Sedimentary Geology, v. 61, p. 207Ð221. istry, v. 17, p. 827Ð837. Ouyang, S., 1996, Spore-pollen assemblage from Bquinshan Group, Qinghai Huang, P., 1993, An Early Jurassic sporopollen assemblage from the north- and its geological age: Acta Palaeontologica Sinica, v. 35, p.1Ð25. western margin of the Junggar Basin, Xinjiang: Acta Micropalaeonto- Ouyang, S., and Norris, G., 1988, Spores and pollen from the Lower Triassic logica Sinica, v. 10, p. 77Ð88. Heshanggou Formation, Shaanxi Province, North China: Review of Ingersoll, R.V., Fullard, T.F., Ford, R.L., Grimm, J.P., Pickle, J.D., and Sares, Palaeobotany and Palynology, v. 54, p. 187Ð231. S.W., 1984, The effect of grain size on detrital modes; a test of the Gazzi- Pearce, J.A., Harris, N.B.W., and Tindle, A.G., 1984, Trace element discrimi- Dickinson point-counting method: Journal of Sedimentary Petrology, nation diagrams for the tectonic interpretation of granitic rocks: Journal v. 54, p. 103Ð116. of Petrology, v. 25, p. 956Ð983. 340 T.J. Greene et al.

Peng, X., Hu, B., and Liu, L., 1990, A restudy for pre-Bogda mountain folded Wang, Z., Wu, J., Lu, X., Zhang, J., and Liu, C., 1986, An outline on the tectonic zone: Xinjiang Petroleum Geology, v. 11, p. 276Ð295. evolution of the Tian Shan of China: Chinese Academy of Geological Qiu, Y., Xue, S., and Ying, F., 1997, Continental hydrocarbon reservoirs of Sciences Institute of Geology Bulletin, v. 15, p. 81Ð92. China: Beijing, Petroleum Industry Press, p. 87Ð104. Wartes, M.A., Greene, T.J., and Carroll, A.R., 1998, Permian lacustrine paleo- Rasmussen, K.A., and Romanovsky, V.V., 1995, Late Holocene climate change geography of the Junggar and Turpan-Hami basins, northwest China: and lake-level oscillation; Issyk-kul, Kyrgyzstan, Central Asia: Society American Association of Petroleum Geologists Annual Convention, Ex- of Economic Paleontologists and Mineralogists Program and Abstracts, tended Abstracts, v. 2, p. A682. v. 1, p. 103. Wartes, M.A., Carroll, A.R., and Greene, T.J., 1999, Permian sedimentary evo- Ren, Z., Jiang, H., Liu, Y., and Li, W., 1994, Organic geochemical characteriza- lution of the Junggar and Turpan-Hami basins, northwest China: Geo- tion of the Permian-Jurassic source rocks in Aiweiergou and Taoyuanzi logical Society ofAmericaAbstracts with Programs, v. 31, no. 7, p.A291. sections on the bordering areas of the Tu-Ha basins: Experimental Pe- Wartes, M.A., Carroll, A.R., Greene, T.J., Cheng, K., and Ting, H., 2000, Per- troleum Geology, v. 16, p. 1Ð9. mian lacustrine deposits of northwest China, in Gierlowski-Kordesch, Schneider, W., Zhao, X., Long, N., Zhao, Y., and Falke, M., 1992, Sedimentary E., and Kelts, K., eds., Lake basins through space and time: American environment and tectonic implication of Jurassic in Toutunhe area, Jung- Association of Petroleum Geologists Studies in Geology 46, p. 123Ð132. gar basin: Xinjiang Geology, v. 10, p. 192Ð203. Wen, J., 1991, Geochronological and isotopic tracer studies of some granitoid S¸engör, A.M.C., Natal’in, B.A., and Burtman, V.S., 1993, Evolution of the rocks from Xinjiang, China: Constraints on Paleozoic crustal evolution Altaid tectonic collage and Paleozoic crustal growth in Eurasia: Nature, and granitoid petrogenesis [Ph.D. thesis]: Santa Barbara, University of v. 364, p. 299Ð307. California, 163 p. Sgibnev, V.V., and Talipov, M.A., 1990, Evolution of the Issyk-kul sedimenta- Windley, B.F., Allen, M.B., Zhang, C., Zhao, Z.Y., and Wang, G.R., 1990, tion basin (Tien-Shan) during Quaternary [abs.]: 13th International Sedi- Paleozoic accretion and Cenozoic redeformation of the Chinese Tien mentological Congress, v. 13, p. 202. Shan range, central Asia: Geology, v. 18, p. 128Ð131. Shao, L., Li, W., and Yuan, M., 1999a, Characteristic of sandstone and its tec- Wu, C., and Xue, S., 1997, Sedimentology of petroliferous basins in China: Bei- tonic implications of the Turpan basin: Acta Sedimentologica Sinica, jing, Petroleum Industry Press, p. 384Ð400. v. 17, p. 95Ð99. Wu, T., and Zhao, W., 1997, Formation and distribution of coal measure oil-gas Shao, L., Stattegger, K., Li, W., and Haupt, B.J., 1999b, Depositional style and fields in Turpan-Hami Basin: Beijing, Petroleum Industry Press, 262 p. subsidence history of the Turpan Basin (NW China): Sedimentary Ge- Wu, T., Zhang, S., and Wang, W., 1996, The structural characteristics and ology, v. 128, p. 155Ð169. hydrocarbon accumulation in Turpan-Hami coal-bearing basin: Acta Shen, J., 1990, The characteristics of petroleum geology in Chaiwopu basin: Petrolei Sinica, v. 17, p. 12Ð18. Xinjiang Petroleum Geology, v. 11, p. 297Ð310. Wu, Z., 1986, Characteristics of evolution and division of tectonic structure in Smith, D.G., 1986, Anastomosing river deposits, sedimentation rates and basin Junggar basin and the appraisal of gas and oil: Xinjiang Geology, v. 4, subsidence, Magdalena River, northwestern Colombia, South America: p. 20Ð34. Sedimentary Geology, v. 46, p. 177Ð196. Wu, Z., and Guo, F., 1991, Second discussion of Bogda nappe tectonic and its Taner, I., Kamen-Kaye, M., and Meyerhoff, A.A., 1988, Petroleum in the Jung- oil-gas accumulation: Xinjiang Geology, v. 9, p. 40Ð49. gar basin, northwestern China: Journal of Southeast Asian Earth Sci- Yang, J., Qu, J., Zhou, H., Cheng, Z., Zhou, T., Hou, J., Li, P., Sun, S., Li, Y., ences, v. 2, p. 163Ð174. Zhang, Y., Wu, X., Zhang, Z., and Wang, Z., 1986, Permian and Triassic Tang, Z., Parnell, J., and Longstaffe, F.J., 1997, Diagenesis and reservoir po- strata and fossil assemblages in the Dalongkou area of Jimsar, Xinjiang: tential of Permian-Triassic fluvial/lacustrine sandstones in the southern Ministry of Geology and Mineral Resources Geologic Memoir, ser. 2, Junggar basin, northwestern China: American Association of Petroleum no. 3: Beijing, Geological Publishing House, 262 p. Geologists Bulletin, v. 81, p. 1843Ð1865. Yu, C., Jiang, Y., and Liu, S., 1996, Jurassic sedimentary boundary between the Tao, M.X., 1994, Tectonic environmental analysis of Turpan-Hami basin: On Junggar and Turpan-Hami basins in Xinjiang: Sedimentary Facies and the genetic relationship between basin and orogenic belt of continental Palaeogeography, v. 16, p. 48Ð54. inner plate: Acta Sedimentologica Sinica, v. 12, p. 40Ð50. Zhao, W., Li, W., and Yan, L., 1991a, Types, characteristics of oil-gas pools and Tilton, G.R., Kwon, S.T., Coleman, R.G., and Xiao, X., 1986, Isotopic studies hydrocarbon distribution regularities in Turpan-Hami basin: Oil and Gas from the West Junggar Mountains, northwest China: Geological Society Geology, v. 12, p. 351Ð363. of America Abstracts with Programs, v. 18, p. 773. Zhao, W., Yuan, F., and Zeng, X., 1991b, The structural characteristics of Wang, C., Ma, R., and Ye, S., 1990, Allochthonous terranes in eastern Tianshan, Turpan-Harmy basin: Acta Petrolei Sinica, v. 13, p. 9Ð18. northwest China, in Wiley, T.J., et al., eds., Terrane analysis of China and Zhao, X., Lou, Z., and Chen, Z., 1991, Response of alluvial-lacustrine deposits Pacific rim: Circum-Pacific Council for Energy and Mineral Resources, of Permian-Triassic Canfanggou Group to climate and tectonic regime Earth Science Series, v. 13, p. 257Ð260. in Dalonggou area, Junggar basin, Xinjiang: China Earth Sciences, v. 1, Wang, C., Luo, B., and Zheng, G., 1993, Organic geochemical characteristics p. 343Ð354. and genesis of crude oils from the Terpan basin, China: Acta Sedimen- Zhou, D., 1997, Studies in the tectonics of China: Extensional tectonics of the tologica Sinica, v. 11, p. 72Ð81. northern margin of the south China sea; amalgamation and uplift of the Wang, C., Cheng, K., Xu, Y., and Zhao, C., 1996, Geochemistry of Jurassic coal- Tian Shan; and wedge extrusion model for the Altyn Tagh fault [Ph.D. derived hydrocarbon of Turpan-Hami basin: Beijing, Petroleum Indus- thesis]: Stanford, California, Stanford University, 354 p. try Press, 247 p. Zhou, R., 1987, The advance on isotope geochronology of Xinjiang: Xinjiang Wang, C., Fu, J., Sheng, G., Zhang, Z., Xia, Y., and Cheng, X., 1998, Labora- Geology, v. 5, p. 15Ð105. tory thermal simulation of liquid hydrocarbon generation and evolution Zhu, Y., and Zhao, J., 1992, Quaternary paleoglacier and neotectonic movement on of Jurassic coals from the Tu-Ha basin: Acta Geologica Sinica, v. 72, the northern slope of Bogda Mountain: Xinjiang Geology, v. 10, p. 40Ð50. p. 276Ð284. Wang, H., Liu, W., Chen, Y., Mou, Z., Li, B., and Zhu, H., 1997, Sedimen- tary microfacies and petroleum productivity of Sanjianfang For- Manuscript Accepted by the Society June 5, 2000 mation in Wenxi-I and Wen-V blocks: Oil and Gas Geology, v. 18, p. 252Ð256. Wang, S., Zhang, W., Zhang, H., and Tan, S., 1994, Petroleum geology of China: Beijing, Petroleum Industry Press, p. 411Ð414. Printed in the U.S.A.