Sedimentation and Tectonics of the Sylhet Trough, Bangladesh

Sedimentation and tectonics of the Sylhet trough, Bangladesh

SAMUEL Y. JOHNSON U.S. Geological Survey, M.S. 939, Box 25046, Denver Federal Center, Denver, Colorado 80225 ABU MD. NUR ALAM Geological Survey of Bangladesh, Segun Bagicha, Dhaka-1000, Bangladesh

ABSTRACT mentary lithic fragments compared to older cene), to (2) a foreland basin linked to the Indo- rocks, reflecting uplift and erosion of the sed- Burman ranges (Oligocene and Miocene), to The Sylhet trough, a sub-basin of the Ben- imentary cover of the Shillong Plateau. If the (3) a foreland basin linked to south-directed gal Basin in northeastern Bangladesh, con- Dauki fault has a dip similar to that of other overthrusting of the Shillong Plateau (Pliocene tains a thick fill (12 to 16 km) of late Mesozoic Himalayan overthrusts, then a few tens of to Holocene). and Cenozoic strata that record its tectonic kilometers of horizontal tectonic transport evolution. Stratigraphic, sedimentologic, and would be required to carry the Shillong GEOLOGIC SETTING petrographic data collected from outcrops, Plateau to its present elevation. Uplift of the cores, well logs, and seismic lines are here Shillong Plateau probably generated a major The Cretaceous to Holocene Bengal Basin, used to reconstruct the history of this trough. (-300 km) westward shift in the course of the underlying the deltaic parts of Bangladesh and The Sylhet trough occupied a slope/basinal Brahmaputra River. adjacent India as well as the Bay of Bengal setting on a passive continental margin from (Fig. 1), forms a "remnant ocean basin" late Mesozoic through Eocene time. Sub- INTRODUCTION (Mitchell and Reading, 1986) at the juncture of sidence may have increased slightly in Oligo- the Indian plate and the Burma platelet (Curray cene time when the trough was located in the The Sylhet trough of northeastern Bangladesh and others, 1983). Basin development began in distal part of a foreland basin paired to the is a tectonically complex province of the Bengal the Early Cretaceous epoch (ca. 127 Ma) when Indo-Burman ranges. Oligocene fluvial-deltaic Basin (Figs. 1-3). This trough is bounded on the the Indian plate rifted away from Antarctica. strata (Barail Formation) were derived from north by the Shillong Plateau, on the east and After a plate reorganization ca. 90 Ma, the incipient uplifts in the eastern Himalayas. southeast by the Chittagong-Tripura fold belt of Indian plate migrated rapidly northward and Subsidence increased markedly in the Mio- the Indo-Burman ranges, and on the west by the collided with Asia between ca. 55 and 40 Ma cene epoch in response to western encroach- Indian Shield platform; the trough is open to the (Curray and others, 1983; Molnar, 1984). ment of the Indo-Burman ranges. Miocene to south and southwest to the main part of the The onshore part of the Bengal Basin has been earliest Pliocene sediments of the Surma Bengal Basin. Published estimates of the thick- divided into platform or shelf, slope or "hinge," Group were deposited in a large, mud-rich ness of late Mesozoic and Cenozoic strata in the and basinal facies (Evans, 1964; Salt and others, delta system that may have drained a signifi- Sylhet trough range from about 13 to 17 km 1986). Crystalline basement of the Indian plate cant proportion of the eastern Himalayas. (Evans, 1964; Hiller and Elahi, 1984), and underlies the shelf and bounds the Bengal Basin Subsidence rates in the Sylhet trough in- much of this strata is Neogene in age. The Sylhet to the west and north; the deeper basin to the creased dramatically (3-8 times) from Mio- trough has minimal topographic relief and is ac- east and southeast may be floored by oceanic or cene to Pliocene-Pleistocene time when the tively subsiding. Although many workers have transitional crust (Desikachar, 1974; Curray and fluvial Tipam Sandstone and Dupi Tila For- reported on the geology and hydrocarbon poten- others, 1983). Paleocene and Eocene strata in- mation were deposited. This dramatic subsid- tial of the Sylhet trough (for example, Lietz and clude carbonate platform and clastic shelf depos- ence change is attributed to south-directed Kabir, 1982; Hiller and Elahi, 1984, Khan and its of the Jaintia Group and their thick basinal overthrusting of the ShOlong Plateau on others, 1988), detailed sedimentologic and pet- correlatives, the Disang Group (Fig. 4). The the Dauki fault for the following reasons. rologic studies of basinal sediments are lacking basin has been characterized by deltaic sedimen- (1) Pliocene and Pleistocene strata thin for two reasons. (1) Outcrops are of poor quality tation since the Oligocene epoch. Today, the markedly away from the Shillong Plateau, and occur only on the uplifted basin margins. onshore part of the Bengal Basin is the site of the consistent with a crustal load emplaced on (2) The Sylhet trough is a frontier petroleum world's largest delta (about 60,000 km2) formed the northern basin margin. (2) The Shillong province that has not been extensively drilled, by rivers (Ganges, Brahmaputra/Jamuna, Pad- Plateau is draped by Mesozoic to Miocene and therefore only limited core is available for ma, Meghna) that drain a large proportion of the rocks, but Pliocene and younger strata are study. Our outcrop and core studies of Neogene Himalayas. This subaerial delta feeds the world's not represented, suggesting that the massif and Oligocene strata, together with observations largest submarine fan (Bengal Fan), which ex- was an uplifted block at this time. (3) South- based on examination of well logs and seismic tends more than 3,000 km south into the Bay of directed overthrusting of the Shillong Plateau data, allow reconstruction of the sedimentary Bengal (Curray and Moore, 1974). and tectonic evolution of the Sylhet trough. We is consistent with gravity data and with re- The Bengal Basin gradually is being en- propose that the Sylhet trough evolved from cent seismotectonic observations. Sandstone croached on by the Indo-Burman ranges, an (1) a passive continental margin (pre-Oligo- in the Tipam has a marked increase in sedi- ~230-km-wide, active orogenic belt associated

Geological Society of America Bulletin, v. 103, p. 1513-1527, 12 figs., 1 table, November 1991.

1513

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 Figure 1. Schematic map showing the location of the Sylhet trough and surrounding tectonic elements. Abbreviations: C, Chittagong; Ca, Calcutta; CTFB, Chittagong-Tripura fold belt; D, Dhaka; HDT, Haflong-Disang thrust system; M, Mandalaya; MBT, Main Boundary thrust fault. Shillong Plateau is shaded. A-A', B-B', and C-C' are lines of section shown in Figure 6. Circled numbers show the approximate locations of the composite stratigraphie columns shown in Figure 4. Based on Ray (1982), Acharya and others (1986), and Salt and others (1986).

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 Age Upper Assam, Southwest Shillong Surma Basin, Naga Hills east of India Plateau, India Bangladesh Surma Basin, India 0 Ma Dihing Formation Dalu and Dupi Tila Dupi Tila Fm 1 (300-1500 m) p Rangapani Fms, Formation and Dihing Group Namsang Fm. , (95-130 m) _ (up to ?—-—r (1500-2500 m) 6375 m) Tipam Tipam Sandstone Sandstone (55-900 m) 5- Bilkona Formation Surma Group (450-880 m) Surma Group (3100 m) (2800-3250 m) Tipam Sandstone Surma Group - Figure 3. Remote sens- Angario li 10- (100-1450 m) ing image of the Sylhet Formation Boka Bil and Boka Bil and Bhuban Bhuban trough area, Bangladesh, (400 - 1170 m) and Formations Formations showing the Chittagong- Boldamgiri Tripura fold belt (TCFB), Surma Group? Formation the Sylhet trough (ST), 20- (260 m) (260-860 m) Barail Formation and the Shillong Plateau Barail Formation Barail Group Barail Group (SP). International border (1150 m) (800-1000 m) (2665 m) (1000 m) _ between Bangladesh and 40- Kopili Formation Kopili Formation Kopili and ' India is highlighted in (400-600 m) r (520 m) h Sylhet upper Disang white, aiding comparison Sylhet Limestone Sylhet Limestone (600- Group with Figure 2 which (400 m) (200 m) 800 m)' (1250-3000 m) shows approximately the Tura Formation .TuraFm. (400 m) 60- undifferentiated same area. Note the (250? m- ) Langpar and Cherra Formations Cretaceous and plunge of north-trending (70 m) early Tertiary folds into the subsurface Dergaon Mahadeo rocks (5000 m?) of the Sylhet trough, and Formation Formation deposited in shelt lower Disang the abundance of lakes 80- (546 m) (165 m) (northeast) to Group (darkest hues) in the basinal (2310+ m) trough. Scale bar (lower (southwest) left) is 20 km long. From granitic and Gumaghat setting 100- SPARRSO (1984). metamorphic Formation complex (150 m)

Sylhet Traps Sylhet Traps 144- H (500 m?) and rift volcanics ? granitic and metamorphic complex

Figure 4. Chart showing the stratigraphy of the Sylhet trough and surrounding areas. The approximate locations of the numbered stratigraphic columns are shown in Figure 1. The latest Cretaceous Langpar Formation, Cherra Formation, Tura Formation, Sylhet Limestone, and Kopili Formation comprise the Jaintia Group. Ages of nonmarine units are based on palynol- ogy. Data from Baksi (1965), Chakraborty (1972), Holtrop and Keizer (1970), Gupta (1976), Murthy and others (1976), Baneiji (1981,1984), and Murty (1983). Note that the time scale is not linear. Wavy lines show unconformities. Time scale from Palmer (1983).

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 SEDIMENTATION AND TECTONICS OF SYLHET TROUGH, BANGLADESH 1515

with eastward subduction of the Indian plate 92° below Myanmar (Burma) (Brunnschweiler, 1966; LeDain and others, 1984; Sengupta and MmtMMmmMmmM: others, 1990). Folds and thrust faults in the Indo-Burman ranges trend north-south (Figs. 1, 3), consistent with this eastward subduction. Earthquake data (Ni and others, 1989; Chen and Molnar, 1990), however, suggest that basement of the Indian plate below the Indo-Burman ranges is moving north. Thus the shortening in the overlying rocks is at least partly decoupled from the basement. Because of this apparent de- coupling, the Indo-Burman ranges have been 25°- termed the "Burma platelet" (Curray and others, 1983). Most of the northward motion of the Indian plate relative to Asia is taken up east of the Indo-Burman ranges on the right-lateral Sagaing fault (Fig. 1). At their northernmost extension, the Indo-Burman ranges merge with the west-trending Himalayas in a complex structural zone, the Assam Syntaxis. The Main Boundary thrust fault, initiated in latest Miocene or Plio- cene time (LeFort, 1975), is regarded as the present thrust front of the Himalayas and forms the northern margin of the Himalayan foredeep. This foredeep is relatively narrow in this eastern Himalayan area and is bounded to the south and separated from the Bengal Basin by the Shillong Plateau.

THE SYLHET TROUGH

The Sylhet trough is a sub-basin of the Bengal Basin in northeastern Bangladesh (Figs. 1-3). This trough is characterized by a large, closed, negative gravity anomaly (as low as -84 milligals), has minimal topography (elevations of about 5 to 20 m) and numerous lakes and swamps, and is actively subsiding. Estimates of the thickness of basin fill range from about 13 to 17 km (Evans, 1964; Hiller and Elahi, 1984). The eastern part of the Sylhet trough lies in the frontal deformation zone of the Indo-Burman ranges. North-trending folds that are uplifted in the Chittagong-Tripura fold belt plunge Figure 2. Schematic geologic map of the Sylhet trough, northeastern Bangladesh (location northward into the Sylhet trough subsurface shown in Fig. 1). Abbreviations as follows: A, Atgram borehole; B, Beani Bazar gas field; (Figs. 1-3). The anticlines (Fig. 5) are com- C, Chhatak gas field; CA, Chhatak anticline; F, Fenchuganj borehole; GT, Goyain trough; monly faulted, and many produce gas. These H, Habiganj gas field; HA, Habiganj anticline; J, Jaintiapur; K, Kailas Tila gas field; KT, folds involve Pliocene and Pleistocene deposits Kushiara trough; P, Patharia borehole; PA, Patharia anticline; R, Rasidpur gas field; S, Sylhet (dipping as steeply as 60° or more) and are gas field; SA, Sylhet anticline; Sh, Shillong; T, Tama Bil. See Figure 4, column 3, for Sylhet either presently active or have stopped growing trough stratigraphic units. Approximate isopach of Pliocene to Holocene strata is based on recently. Structural relief between paired anti- limited well and seismic data, and on interpretations by Lietz and Kabir (1982) and Hiller and clinal crests and adjacent synclinal troughs may Elahi (1984). A-A' is line of section for Figure 6. Locations of Figure 5 seismic lines (5A, 5B) be as much as 7,000 m (Hiller and Elahi, 1984), are also shown. and the synclines have acted as major late Neo- gene and Quaternary depocenters. The folds de- The Sylhet trough is bounded to the north by zoic and Cenozoic sedimentary cover drapes crease in amplitude westward, and are not the Shillong Plateau (Figs. 1-3), which is under- portions of the southern Shillong Plateau and present west of about 91° (Lietz and Kabir, lain by a basement complex of Archean gneiss generally dips south in a monocline (Evans, 1982), where the Sylhet trough merges with the and minor greenstone and upper Proterozoic 1964). The Shillong Plateau has a maximum main part of the Bengal Basin. granite (Acharya and others, 1986). Late Meso- elevation of about 2,000 m. Thus, it has been

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1515 SEDIMENTATION AND TECTONICS OF SYLHET TROUGH, BANGLADESH

postulated that there may be as much as 15-19 km of structural relief between the Shillong Pla- teau and the basement of the Sylhet trough (Hiller and Elahi, 1984). This structural relief is characterized by a steep gravity gradient (Fig. 6), as much as 94 milligals in 40 km (Rahman and others, 1990). The poorly exposed Dauki fault (Figs. 1, 2, 6) forms the contact between the Shillong Pla- teau and the Sylhet trough. The 5-km-wide fault zone has a straight trace across essentially flat topography and is characterized by extensive fracturing (Evans, 1964) and steep (up to 85°) dips of Pliocene and Pleistocene strata (Khan, 1978). Within about 40 km south of the Dauki fault, the north-trending folds in the subsurface of the Sylhet trough are deflected to a northeast trend. Evans (1964) proposed -250 km of right- lateral displacement on the Dauki fault, envi- sioning the Shillong Plateau as a displaced spur of the Indian plate. He based this model on local juxtaposition of disparate stratigraphic sequences and correlative Eocene shelf (Sylhet Limestone) and basin (Disang Group) deposits (Fig. 4), and on greater inferred amounts of shortening in the Indo-Burman ranges north of the latitude of the Dauki fault than to the south. Subsequently, several workers (Desikachar, 1974; Dasgupta, 1977; Brunnschweiler and Khan, 1978; Rao, 1983) disputed the rationale for large lateral dis- placements and instead emphasized vertical dis- placements. Recently Seeber and Armbruster (1981) and Molnar (1987) suggested that the Sylhet trough and its basement are underthrusting the Shillong Plateau and that rupture on this thrust zone produced the great (magnitude 8) 1897 Assam earthquake. Support for underthrusting comes from interpretation of gravity anomalies (Fig. 6; Nandy, 1986; Verna and Krishna Kumar, 1987; Chen and Molnar, 1990). The elevation of the Shillong Plateau should be compensated by a thick crustal root and yield a negative gravity anomaly of about -100 milligals, not the positive or near positive anomalies shown in Figure 6. Instead, the Shil- long Plateau seems to be supported by a strong lithosphere, as would be the case if the Indian plate was thrust below it (Chen and Molnar, 1990). The thrust-fault hypothesis for the southern margin of the Shillong Plateau and the Figure 5. A. Portion of BOGMC seismic line PK-SU-13 (unmigrated) across the Habiganj Dauki fault is most consistent with Pliocene and anticline near the Habiganj gas field (Fig. 2), showing the structural style and reflector charac- younger depositional patterns in the Sylhet teristics of the Surma Group, the Tipam Sandstone, and the Dupi Tila Formation. The Tipam trough, as discussed below. To the east, the Sandstone thins from about 1,500 m on the east flank of the fold (1.65 to 2.5 sec TWT) to about Dauki fault may bend into the Haflong-Disang 1,125 m (0.65 to 1.4 sec TWT) over the crest of the fold. The increase in seismic velocity with thrust fault (with a change in dip) as originally depth makes this thinning appear less pronounced than described. Note reverse fault on west proposed by Evans (1964). To the west, offset limb. B. Line drawing of portion of BOGMC seismic line PK-SU-3 (unmigrated) across the on the Dauki fault may diminish, and(or) the Sylhet anticline (Fig. 2) showing reflector geometry and pronounced thinning of the Pliocene- Dauki fault may be truncated by a north- Pleistocene Tipam Sandstone and Dupi Tila Formation over the anticlinal crest. Dotted lines trending structure, also discussed below in this show inferred continuation of formation contacts and are constrained by well data near the paper. anticlinal crest.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1518 JOHNSON AND NUR ALAM

South Figure 6. Schematic cross sections showing A' Bouguer gravity anomalies and the thickness Bouguer gravity of the Pliocene-Pleistocene section in the Sylhet trough. Lines A-A' and B-B' show marked north-to-south thinning away from the Shillong Plateau. Line C shows Pliocene- Pleistocene folds. Locations of lines shown 2 5 rsrs Shillong Plateau << Sylhet trough on Figure 1; A-A' is also shown in Figure 2. Ê 2- OLL Gravity data are from Rahman and others Ì 0- mainly Indian plate (1990). Dip on the Dauki fault is inferred (see c 4-2- Tipam Sandstone and Dupi Tila Formation I "" basement rocks text for discussion). Abbreviations on C-C' TO g Surma Group are as follows: CA, Chattak anticline; GT, "05 -8 - Goyain trough; KT, Kushiara trough; PA, Pa- v.e. = 2 tharia anticline; SA, Sylhet anticline. Line South North C-C' is partly modified from Hiller and Elahi B' 30-? Bouguer gravity (1984). Vertical exaggeration *2.

« -2o°J -40 £ -60- -80-1 SYLHET TROUGH STRATIGRAPHY, Shillong Plateau Sylhet trough < < SEDIMENTOLOGY, AND 2 O U- 0- PETROLOGY mainly Indian Tipam Ss. and Pupi Tila Fm. -2- plate basement -4 _ rocks Surma Group -6 The focus of this paper is the post-Eocene v.e = 2 evolution of the Sylhet trough. The pre- Oligocene history of the Sylhet trough is briefly west Bouguer gravity discussed below to provide a framework for € subsequent discussions. -60-r -80^ Pre-Oligocene Rocks 2-, CA GT SA KT PA 0 -2- _Tipam Ss. and Dupi Tila Fm Stratigraphy. Pre-Oligocene rocks crop out -4- on the northern (north of and within the Dauki -6- Surma Group fault zone) and eastern margins of the Sylhet -8 20 km v.e.=2 trough; they have not been penetrated by wells in the trough subsurface. The pre-Oligocene section (about 1,500 m thick) on the southern Shil- long Plateau reflects trasgressive sedimentation TABLE 1. COMPOSITION OF SYLHET TROUGH SANDSTONE AND SAND on a passive margin (Baneiji, 1981; Salt and others, 1986). East of the Sylhet trough in the Slraligraphic unit Barail Surma Tipam Dupi Tila Modem, Modern, Naga Hills of India (Figs. 1 and 4, col. 4), pre- Formation Group Sandstone Formation Hari River Dauki River Oligocene rocks comprise the Disang Group X SD X SD X SD X SD X SD X SD (3,560 to 5,310 m thick), divided by Rao (1983)

Monocrystalline quartz 70.0 11.9 58.5 6.6 47.3 6.7 62.9 19.0 66.9 3.8 50.0 9.6 into lower and upper units comprising basinal Polycrystalline quartz 3.7 2.8 3.3 2.0 6.1 3.6 5.3 3.3 5.1 4.4 0.8 0.7 and shallow-marine facies, respectively. These Chert 1.1 0.8 0.7 0.7 1.5 1.0 0.4 0.6 0.3 0.3 0.4 0.5 Plagioclase feldspar 3.8 2.4 7.8 2.5 4.6 1.6 2.9 2.7 6.2 2.2 11.0 0.4 thickness relationships indicate a pre-Oligocene Potassium feldspar 2.3 1.3 8.9 2.2 10.5 3.7 5.3 3.2 9.3 2.3 20.2 1.0 Sedimentary lithic 7.2 3.7 7.1 2.6 16.9 0.4 14.3 18.9 3.6 0.8 4.0 1.2 shelf-to-basin transition in the vicinity of the Volcanic lithic 0.9 2.6 0.6 0.6 1.6 1.2 0.2 0.4 0.1 0.2 0.4 0.6 Sylhet trough. Facies relations described by Rao Metamorphic lithic 3.8 2.5 2.4 1.6 2.6 2.2 0.7 1.4 0.0 0.0 1.0 1.0 Mica 6.5 9.5 9.3 7.5 6.5 8.8 7.4 10.7 3.0 2.3 5.9 4.6 (1983) suggest that by late Eocene time, pro- Accessories 0.8 0.5 1.7 0.9 2.3 2.8 0.6 0.9 5.5 4.0 6.5 5.5 grading shelf and slope deposits had filled in the Number of samples 8 14 8 7 3 3 proximal portions of the adjacent basin. Plagiodase/total feldspar 0.61 0.19 0.47 0.10 0.31 0.08 0.34 0.12 0.40 0.06 0.35 0.01 West of the Chhatak anticline in the north- QFL 81,7,12 70,19,11 61,16,23 74,10,16 79,17,4 58,36,6 western part of the Sylhet trough (Fig. 2), on QmFLt 75,7,18 66,19,15 52,16,32 67,10,23 73,17,10 57,36,7 QmPK 92,5,3 78,10,12 75,9,17 88 4,8 81,8,11 61,14,25 Bangladesh Oil, Gas, and Minerals Corporation (BOGMC) seismic line PK-SU-6, the Sylhet Note-, percentage of grain types is presented as a mean (X) and standard deviation (SD). Statistical parameters calculated alter Dickinson (1985). Percentages Limestone reflector is underlain by —2,500 m based on point counts of more than 300 framework grains (quartz, feldspar, lithic fragments) per thin section or grain mount using the "Gazzi-Dickinson" method (Ingersoll and others, 1984). For the purpose of comparison, compositional data above are normalized without showing proportion of diagenetic phases (cements (between 4.2 and 5.1 sec, two-way traveltime = and matrix) and porosity. Raw petrographic data used in constructing this table are presented in Johnson and Alam (1990). TWT) of stratified rock. (Inferred thicknesses of

stratigraphic units from seismic data are based Q largely on the time-depth plot of Lietz and Kabir, 1982.) If one assumes that the overlying Kopili Formation is the same thickness in the Sylhet trough as on the Shillong Plateau (about 400-600 m), then the pre-Oligocene section in the Sylhet trough is a minimum of about 2,900 m thick. Previously Khan and others (1988) suggested on the basis of seismic data that pre- Kopili rocks in the Sylhet trough are 5,000 m thick.

Oligocene Rocks

Stratigraphy. Oligocene rocks comprising the Barail Formation crop out on the northern and eastern margins of the Sylhet trough and have been penetrated in the subsurface only in the Rasidpur 2 and the Atgram IX wells (Fig. 2). In the southwestern Shillong Plateau (Figs. 1, 4), the correlative Barail Group consists of about 1,000 m of sandstone and carbonaceous shale of inferred delta-plain and shallow-marine origin (Baksi, 1965; Chakraborty, 1972; Baneiji, 1981). East of the Sylhet trough in the Naga Hills (Fig. 1), the Barail Group is much thicker (2,665 m) and consists of deltaic and prodeltaic deposits (Rao, 1983). In the northwestern part of the Sylhet trough on the west end of BOGMC seismic line PK- SU-6 (west of the Chhatak anticline; Fig. 2), the top of the Surma Group and the top of the Sylhet Limestone form prominent markers at about 3,200 m (2.2 sec, TWT) and 7,700 m (4.25 sec, TWT), respectively. If one assigns the Miocene Surma Group (Fig. 4) in this area a thickness similar to that in the eastern Sylhet trough (about 3,100 m, see next section), then the cumulative thickness for the Kopili Forma- tion and the Barail Formation is about 1,400 m. We infer that the Barail Formation in the western Sylhet trough is about 800-1,000 m thick, similar to the thickness on the Shillong Plateau. Other authors (for example, Holtrop and Keizer, + Surma Group, n = 14 O modern sand, Hari River, n = 3 1970) have suggested much greater thicknesses a Tipam Sandstone, n = 8 • modern sand, Dauki River, n = 3 (2,450-3,350 m) for the Barail Formation in the Sylhet trough, but we could not reconcile these Figure 7. QFL (total quartz, feldspar, lithic fragments) and QmFLt (monocrystalline quartz, inferences with our interpretation of the avail- feldspar, total lithic fragments) ternary diagrams showing the mean (symbols) and standard able seismic data. In the present geographic derivation (solid hexagonal outlines) of compositions of Sylhet trough sandstones and modern framework, the Barail Formation clearly thick- sands, and the relevant sandstone provenance fields (enclosed by dashed lines) of Dickinson ens basinward to the east and southeast, main- (1985). Pétrographie data are summarized in Table 1, based on the raw data and plots taining the pre-Oligocene pattern of a prograd- presented in Johnson and Alam (1990). ing shelf that filled the adjacent basin. Sedimentology. The Barail Formation was examined in four poorly exposed areas in the stone; mottled reddish-orange and gray siltstone (1) fluvial channel deposits, consisting mainly Dauki fault zone near Jaintiapur and along the and mudstone; and yellowish-brown to reddish- of fining-upward sequences of trough-cross- Dauki River (Fig. 2). In these outcrops, the Ba- brown intraformational conglomerate. The bedded, parallel-bedded, and ripple-laminated, rail Formation consists of multicolored (yellow- lower part of the Barail Formation consists of fine- to medium-grained sandstone; and (2) ish-brown, light-red, light-brown, gray) sand- alternating intervals (several meters thick) of flood-plain deposits consisting mainly of mottled

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1520 JOHNSON AND NUR ALAM

silty mudstone (inferred paleosols) and thin (5-15 cm), lenticular bed;; of fine-grained sandstone that form bedsets as; thick as 150 cm (inferred crevasse-splay deposits). Collectively, the alternating channel and flood-plain deposits of the lower Barail form cycles of mixed-load, probably meandering-riveir origin. The upper part of the Barail Formation consists of alternating sequences (tens of centimeters to meters thick) of massive to cross-bedded sandstone and intraformational conglomerate, bounded by erosion surfaces. Intraformational sandstone and mudstone clasts (as much as 6 cm in maximum dimension) are commonly ce- mented or rimmed by hematite. The coarser texture of these strata and the abundant erosional bounding surfaces reflect deposition in rivers of greater competence than those that deposited the lower Barail. Sandstone Petrology. Eight samples of moderately well-sorted, fine- to medium-grained sandstone of the Barail Fornjation were examined petrographically (Table 1; Fig. 7). The sandstones have a high quartz content and are relatively low in feldspar. Four samples have substantial hematite cements and(or) matrix, which formed by the alteration of mafic minerals and(or) lithic fragments. This process re- sulted in an increased proportion of modal quartz relative to the primary detrital sediment. The samples contain a diverse suite of lithic grains, including sedimentary rock fragments (as much as 10.0%), volcanic rock fragments (as much as 6.1%), quartz-mica tectonite fragments (as much as 5.4%), and low-grade metamorphic Figure 8. Strata from cores of the Surma rock fragments (as much as 3.2%). Many mono- Group. A. Facies A parallel-laminated mud- crystalline grains have remnant quartz over- stone and siltstone to very fine grained sand- growths, further indicating a major source in stone. Note rare ripple lamination (arrow) sedimentary rocks. The samples fall in the "re- and small load structures. B. Horizon of soft- cycled orogen" and the "quartzose recycled oro- sediment deformation of inferred slumping gen" provenance field on QFL and QmFLt origin in facies A deposits. Note also the ball- diagrams, respectively (Fig. 7; Dickinson, 1985). and-pillow structures at bottom part of photograph. Scale (bar = 1 cm) in upper part. Miocene to Early Pliocene Rocks C. Facies B ripple-laminated sandstone and plane-laminated mudstone. Note concave un- Stratigraphy. The Surma Group has been dulating lower contacts to ripple sets, planar divided into the Bhuban and Boka Bil Forma- to rounded ripple crests, and opposing dips in tions in the Sylhet trough (Khan and others, ripple foresets. Cores are from the Fenchuganj 1988). We detected no signicant lithologic or 2 borehole (Fig. 2). petrologic differences between the Bhuban and Boka Bil, and we discuss them as a single unit. A complete thickness of the Surma Group was penetrated in the Atgram IX (2,700 m) and Baneiji, 1984). The contact with the underlying Acharya and others, 1986). Holtrop and Keizer Rasidpur 2 (3,109 m) wells (Fig. 2). Approxi- Barail Formation appears to be a transgressive (1970) inferred that the Surma Group in the mately 3,870 m of Surma Group was drilled in onlap, placed approximately at the Oligocene- Sylhet trough is all Miocene; but Lietz and the Fenchuganj 2 well without reaching its base. Miocene boundary (Banerji, 1984; Salt and oth- Kabir (1982), Baneiji (1984), and unpublished The Surma Group (Fig. 4) has a reported thick- ers, 1986). The contact with the overlying BOGMC palynological reports suggest that its ness of 2,800-3,250 m in the Naga Hills to the Tipam Sandstone is diachronous from middle upper part is early Pliocene. east (Rao, 1983), and 660 to 2,030 m in the Miocene in Assam to early Pliocene in Tripura, Sedimentology. We examined cores and southwest Shillong Plateau (Chakraborty, 1972; India (Shrivastava and others, 1974; Rao, 1983; borehole geophysical logs of the Surma Group

from the Beani Bazar, Fenchuganj 2, Atgram Facies D prodelta and delta-front deposits of a large mud- sandstone rich delta system similar to the modern Bengal IX, Habiganj 1, and Rasidpur 1 wells, and logs í (5-20 m) from six additional Sylhet trough wells (Fig. 2). delta (Fig. 1). Facies D sandstones are inferred We also examined several outcrops on the west abrupt or distributary-channel and proximal distributary gradational mouth-bar deposits. Thick, massive, amalga- flank of the Patharia anticline, and on the north- contact ern basin margin near Jaintiapur and along the mated sandstone beds with mudstone rip-up n clasts suggest deposition in channels. Anoma- banks of the Hari River (Fig. 2). Four lithofacies o a> were recognized. oo <3 O lously thick sets of ripple and climbing-ripple m Facies A (Figs. 8A, 8B) is by far the most "+ lamina suggest rapid deposition, probably where O) abundant, consisting of parallel-laminated gray the channels broadened and rapidly deposited to black mudstone and gray to yellowish-gray a> their sediment load on bars. Low-angle and(or) o o siltstone to very fine grained sandstone. The o a) hummocky bedding suggests reworking of ro Li- sandstone:mudstone ratio ranges from 1:5 to <¿ mouth-bar sediments by storm waves. 1:20. The coarser siltstone and sandstone lami- Facies B interbedded sandstones and mud- nae are generally 1-5 mm thick (Fig. 8A), have stones were probably deposited on deeper parts planar to slightly undulatory upper and lower of channel-mouth bars, but also above wave boundaries, and are laterally continuous (for as base. Sands and muds were probably trans- much as 20 m in outcrops) to lenticular. Thicker ported to the mouth bar by flood and tidal cur- laminae (3=2 mm) are uncommonly ripple lami- rents, and then the sands were reworked by nated. Several intervals as thick as 30-40 cm are Figure 9. Schematic diagram showing oscillating currents into wave ripples. Mud characterized by soft-sediment folding and con- lithofacies and thickness of Surma Group drapes were deposited from suspension during volute lamination (Fig. 8B). The thickness of coarsening-upward sequences determined quiescent periods. Mud-dominated facies A these deformed intervals suggests that they from well-log analysis. Mean and standard strata were probably deposited in deeper parts of formed by slumping. Small (2 mm > diameter) deviation for thickness based on measure- the mouth-bar or prodelta shelf; the common load casts and ball-and-pillow structures are also ment of 23 sequences. Facies A, B, and C are soft-sediment deformation of inferred slumping common. indistinguishable on logs; however, core origin also suggests prodelta slope deposition. Strata of facies A are gradational with less studies suggest that facies A is most abundant, Facies A deposits resemble the "streaked mud" common facies B strata, which consist of thinly facies B is less common, and facies C is rare. facies of De Raaf and others (1977). They inter- interbedded and interlaminated, very fine preted the coarser sand or silt layers as suspen- grained to fine-grained sandstone and silty mud- sion deposits of storms, slightly reworked by stone (Fig. 8C). The sandstone:mudstone ratio waves. Mudstone laminae are inferred intra- ranges from about 5:1 to 1:5. Sandstone beds are mon, the lack of stratification and the mixing of storm suspension deposits. Facies A deposits as thick as 5 cm and are ripple to parallel lami- sediments of different grain size indicate exten- may have also been deposited by density under- nated. Ripple sets are as thick as 2 to 3 cm and sive bioturbation. flows associated with ebb-tidal currents (John- son and Alam, 1990). The intensive bioturba- are characterized by concave, undulating lower Facies D consists of massive, low-angle to tion of facies C reflects decreased sediment contacts. Foreset laminae within single ripple hummocky-bedded, and ripple-laminated, fine- accumulation rates on the shelf and slope, possi- forms and within thin (less than 10 cm) strati- to medium-grained sandstone. Ripple sets are bly associated with delta-lobe abandonment. graphic horizons commonly dip in opposing di- asymmetric, trough-shaped, and anomalously rections. Ripple cross-laminae are commonly thick (3-10 cm), and climbing ripples are com- On the modern Bengal delta, the mouth bar strongly tangential to the lower set boundary mon. Only thick (several meters) massive beds or shelf extends about 70 km offshore to a depth and in some cases ascend the adjacent ripple were observed in cores. Uncommon abrupt tex- of about 20 m, and sand is deposited as the form. Ripple crests are symmetric, asymmetric, tural inversions in these beds (for example, mean grain size in some areas out to the shelf- or planed off, and are commonly draped by medium-grained sandstone overlying fine- to slope break (Kuehl and others, 1989). The base mudstone laminae. The ripple marks are com- medium-grained sandstone) indicate that the of the steep part of the slope is about 90 m in monly form discordant, that is, the morphology thicker beds are amalgamated and represent depth and about 90 km from shore. Using this of the ripple crest does not reflect the internal several depositional events. delta as an analogue, paleobathymetry for structure and geometry of the ripple. These rip- Based on well-log interpretation, the Surma Surma Group deposits probably ranged from a ples closely resemble those described in the "len- Group is characterized by sequences that range few meters (channel and mouth bar) to less than ticular bed" facies of De Raaf and others (1977) in thickness from ~40 to 300 m (Fig. 9). Sedi- 100 m (prodelta slope). Given this limited pa- and are of inferred wave-generated origin. Soft- mentologie analysis suggests that the lower parts leobathymetric range and the long time period sediment deformation structures, mainly load of the sequences consist mainly of facies A, less inferred for Surma Group deposition (Fig. 4), it casts and ball-and-pillow structures (10 cm > common facies B, and rare facies C, and they is surprising that neither subaerial nor deeper- diameter), characterize intervals as thick as 10 show no systematic grain-size trends. The upper water deposits were recognized. Sediment ac- cm in facies B. several meters of each sequence commonly cumulation and subsidence appear to have Strata of facies A and B are transitional to, coarsens upward and is capped by facies D maintained a remarkably long-lived equilib- and interbedded with, rare facies C strata, which sandstone. The facies D sandstone beds are rium. The prodelta shelf and slope of this Mio- consist of massive, silty to sandy mudstone. probably responsible for the continuous to len- cene delta was broad (>70 km wide), compa- Dispersed sand grains and isolated small (<5- ticular banded reflectors that characterize the rable to that of the modern Bengal delta mm) clots and curved lenses of sand are com- Surma Group on seismic data (Fig. 5). (Coleman, 1969). mon. Although well-defined burrows are uncom- Surma Group strata are interpreted as the The contact of the Surma Group and the

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1522 JOHNSON AND NUR ALAM

overlying Tipam Sandstone is exposed on the Surma Group were examined (Table 1; Fig. 7). Pliocene to Pleistocene Rocks west flank of Patharia anticline and along the Strata are generally arkosic arenites and have Hari River (Figs. 2, 10A). At the Patharia anti- very similar compositions (mean QFL = Stratigraphy. The Tipam Sandstone and the cline locality, there is a significant, coarsening- 70,19,11). Quartz is generally monocrystalline Dupi Tila Formation comprise the Pliocene and upward facies change in the upper ~ 10-15 m of and includes some grains with remnant over- Pleistocene deposits of the Sylhet trough. These the Surma Group. Facies remain mainly paral- growths. Lithic fragments are mainly sedimen- units crop out only on the uplifted northern and lel- and ripple-laminated, but the sandstone: tary (as much as 10.4% in one sample), and eastern basin margins, and they can be identified mudstone ratios rises significantly, sandstone quartz-mica tectonite fragments are also present easily on seismic records (Fig. 5). The Tipam laminations and beds become thicker and more in significant amounts (as much as 3.9%). Mica, Sandstone is a nearly massive unit having minor continuous, and there are two lenticular in- mainly biotite, is very common and in some internal lithologic variability. Hence it is seismi- terbeds (10-20 cm thick) of massive, poorly samples abundant (as much as 20.9%). The QFL cally homogeneous and has few internal reflec- sorted conglomerate (probably intraformation- modes plot in the "recycled orogen" provenance tors. This massive seismic character is easily al) containing angular to rounded clasts. All of field, and the QmFLt modes plot in the contrasted with the banded reflectors of the un- these facies changes suggest shoaling, consistent "mixed," "transitional continental," and "quartz- derlying Surma Group and the overlying Dupi with the upward transition to the nonmarine ose recycled orogen" fields (Dickinson, 1985). Tila Formation. The Dupi Tila, and to a lesser Tipam Sandstone. The contact between the The abundance of mica, the presence of sedi- degree the Tipam, each thin over anticlinal Tipam and the Surma Group is erosional and mentary and quartz-mica tectonite fragments, crests (Fig. 5; Hiller and Elahi, 1984), indicating may be a local unconformity. An erosional sur- and the evidence for a recycled source of some that they were deposited synchronously with face also characterizes the Tipam-Surma contact quartz grains favors a partial sedimentary or folding. The Tipam-Dupi Tila contact appears along the Hari River (Fig. 10A). The transi- metasedimentary source. There is minimal over- conformable in synclines but is commonly an tional, shoaling-upward facies seen in the Lonely lap between the Surma Group and Barail angular unconformity near anticlines (for exam- Chara section, however, are not present in the Formation sandstones on ternary diagrams ple, see Fig. 5). The upper contact between the Hari River outcrops. (none on the standard deviation plots of Fig. 7); Dupi Tila Formation and modern sediments is also commonly unconformable in anticlinal Sandstone Petrology. Fourteen thin sections they are distinguished on the basis of more areas. of fine- to medium-grained sandstone from the quartz and less feldspar in the Barail.

Figure 10. Outcrops of the Tipam Sandstone and Dupi Tila Formation along the Hari River (Fig. 2). A. Erosional contact (ar- rows) between interbedded mudstone and sandstone of the Surma Group (lower right) and massive sandstone of the Tipam Sandstone (upper left). Section in photograph is about § m thick. B. Typical exposure of cross-bedded, coarse-grained sandstone of the Tipam Sandstone. Section in photograph is ~7 m thick. C. Alternating resistant, coarse-grained, fluvial-channel deposits, and less resistant, more fine-grained flood-plain deposits of the Dupi Tila Formation. Man (sitting) for scale.

The maximum penetration of the Tipam- flanked by flood plains on which fine-grained alteration of mafic minerals and lithic fragments. Dupi Tila section in a well is about 2,440 m at sediments were deposited and subsequently Partly dissolved feldspar grains are also present Beani Bazar (Fig. 2); however, Sylhet trough eroded. in most of the samples. Thus, most of the Dupi wells have been drilled on anticlinal crests where Sedimentology of the Dupi Tila Formation. Tila samples are enriched in quartz relative to the minimum section is encountered. In a map- Outcrops of the Pliocene and Pleistocene Dupi the original detrital sediment. Because of the var- ping study near Jaintiapur (Fig. 2) on the north- Tila Formation were observed along the Hari iable amount of alteration, sandstones plot in ern margin of the Sylhet trough, Khan (1978) River east of Jaintiapur and on the west flank of broad fields on ternary diagrams. The least al- reported thicknesses of 2,271 m for the Tipam the Patharia anticline (Fig. 2). In both areas, the tered samples, which most accurately reflect the Sandstone and 2,393 m for the Dupi Tila For- Dupi Tila consists of alternating intervals of primary detrital sediment, plot in the "recycled mation. Seismic lines reveal even greater thick- fluvial-channel and flood-plain deposits (Fig. orogen" provenance fields of Dickinson (1985). nesses in synclinal areas. In the Goyain and 10C). Channel deposits are generally 2-4 m These least altered samples are similar in com- Kushiara synclinal troughs on the flanks of the thick, have erosional bases, consist of trough position to the Tipam Sandstone. Sylhet anticline (Figs. 2, 5, 6), the Tipam-Dupi cross-bedded and flat-bedded to low-angle- Tila section is about 6,375 m thick (3.65 sec bedded, fine- to coarse-grained sandstone, and Modern Sediments TWT) and 6,000 m thick (3.3 sec TWT), re- generally fine upward. Trough cross-bed set spectively, interpretations consistent with Lietz thicknesses and widths generally range from 10 Petrology of River Sands. For comparison, and Kabir (1982) and Hiller and Elahi (1984). to 50 cm and 50 to 100 cm, respectively. Flood- six samples of medium sand collected from the The Tipam-Dupi Tila section thins to the plain deposits consist of mottled, yellowish- beds of the Dauki and Hari Rivers a few kilome- south (Figs. 2,6), where the maximum thickness brown to gray, silty clay paleosols. Root casts ters south of the Dauki fault (Fig. 2) were also in a synclinal trough (between Rasidpur and and ferruginous nodules are common; primary analyzed petrographically (Table 1; Fig. 7). The Habiganj anticlines) is about 3,200 m (2.2 sec, sedimentary structures have been obliterated. compositions of these samples vary from river to TWT, BOGMC seismic line T-3). Farther to the The alternating channel and flood-plain deposits river, reflecting local Shillong Plateau sources. west in the undeformed part of the Sylhet trough of the Dupi Tila Formation in the two areas are Hari River sand is richer in quartz, plotting in (about 60 km west of the Chhatak anticline), the organized into fining-upward cycles of probable the "transitional continental" to "craton inte- Tipam-Dupi Tila interval thins from about meandering river origin. The best examples of rior" provenance fields on QFL diagrams and in 3,200 m (2.2 sec TWT, line PK-SU-6) about 10 these cycles (thicknesses of 6-12 m) are exposed the "transitional continental" field on the km south of the Dauki fault, to 1,500 m thick along the Hari River. QmFLt diagram (Dickinson, 1985). Dauki (1.15 sec TWT, line R4) about 100 km to the Sandstone Petrology of the Tipam Sand- River sands contain more feldspar, and plot in south. It is possible that some of this thinning (a stone. Eight samples of fine- to medium-grained the "transitional continental" provenance field few hundred meters?) may reflect interfingering sandstone were examined petrographically on both ternary diagrams. The Dauki River between the Tipam Sandstone and the underly- (Table 1; Fig. 7). These sandstones are lithic sand composition does not overlap with any of ing Surma Group (Fig. 6). Regional Pliocene- arenites. Monocrystalline quartz grains have the ancient samples, and the Hari River sand Pleistocene correlatives of the Tipam and Dupi common remnant overgrowths, indicating deri- overlaps only partly with the sandstone of the Tila are thick and extensive in the Assam part of vation from a sedimentary source. Sedimentary Surma Group in the QFL diagram. These sand the Himalayan foreland basin; correlatives are rock fragments are the most common lithic samples are fresh and have not been affected by thin and occur very locally in the Shillong Pla- fragment (16.9% + 10.4%), but quartz-mica tec- surface weathering or diagenesis. There is no teau and Naga Hills (Fig. 4). tonite fragments are also common (2.3% ± 2.0%). secondary process that could lead to an increase in the proportion of lithic fragments, which Sedimentology of the Tipam Sandstone. Mica, especially biotite, is a common accessory would be needed for more overlap between The Pliocene Tipam Sandstone was examined mineral. Sandstone compositions plot in the "re- these samples and Sylhet trough sandstone along the Hari River east of Jaintiapur and on cycled orogen" provenance field on a QFL dia- samples. the west flank of the Patharia anticline (Fig. 2). gram, and near the boundary of the "quartzose At these localities, the Tipam (Fig. 10B) consists recycled orogen," "transitional recycled oro- of thick, vertically stacked beds of yellowish- gen," and "mixed" provenance fields on a BASIN EVOLUTION brown to yellowish-orange, medium-grained to QmFLt diagram (Dickinson, 1985). Based on pebbly sandstone. Strata are characterized by their larger proportion of lithic fragments, The Sylhet trough has undergone a complex large-scale scours (laterally continuous for tens Tipam Sandstone samples are clearly distinct evolution recording the transition from a pas- of meters with tens of centimeters of erosional from samples of the Barail Formation and sive, rifted continental margin to a foreland relief at their base), trough and planar cross- Surma Group (Fig. 7). basin on the margins of two mobile belts, the bedding (sets as thick as 150 cm), and flat to Sandstone Petrology of the Dupi Tila Indo-Burman ranges and the Himalayas. The low-angle bedding. Clasts include both crystal- Formation. Seven fine- to medium-grained geohistory and paleogeographic diagrams of line and sedimentary lithologies (including sandstone samples of the Dupi Tila Formation Figures 11 and 12 illustrate this evolution. Five mudstone rip-up clasts). Well logs and seismic were examined petrographically (Table 1; Fig. points were used in constructing the geohistory data similarly suggest little lithologic variability 7). These samples showed a far greater range of plots, conforming to the above discussions. We and a dominantly sandstone composition. The composition than the older suites of samples an- assumed that sedimentation was almost contin- coarse texture, large-scale sedimentary struc- alyzed, in large part due to the effects of weath- uous and that unconformities are either local or tures, and vertical sedimentary sequence of the ering and early diagenesis. The least altered of short duration (Fig. 4), consistent with seis- Tipam suggest that it formed in a bedload- sample contains 41.5% sedimentary lithic frag- mic interpretations. Deposition is assumed to dominated (probably braided) fluvial system. ments and 4.3% biotite. The most quartz-rich have begun ca. 120 Ma, shortly following rifting The presence of fine-grained intraformational samples contain the most hematite cement and opening of the Indian Ocean (Curray and clasts indicates that these rivers were at times and(or) matrix, which clearly formed through others, 1983). A thickness of 5,500 m was as-

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1524 JOHNSON AND NUR ALAM

signed to the section between the basement and the top of the Kopili Formation (Khan and others, 1988). Although we were not able to con- firm this thickness with available seismic data, it is consistent with reported thicknesses from the Naga Hills to the east in India (Fig. 4; Rao, 1983). On the basis of our seismic analysis, a thickness of 900 m was assigned to the Oligo- cene Barail Formation. A thickness of 3,100 m was assigned to the Surma Group, an interme- diate value between maximum and minimum thicknesses penetrated in wells. The thicknesses for both the Barail Formation and the Surma Group may vary within the Sylhet trough; however, these variations cannot be documented with the available well and seismic data. For the Tipam Sandstone and Dupi Tila Formation, we show thicknesses of 6,37:5 m, representing the deepest synclinal depocenters (Fig. 11 A), and 3,200 m, representing the section in the western, flat-lying part of the Sylhet trough (Fig. 1 IB). Age (million years) A9e (million years) Strata in the Sylhet trough therefore have a total thickness of about 12,700 to 15,875 m. Figure 11. Geohistory diagrams (Van Hinte, 1978) showing the subsidence history of two The pre-Oligocene history (Fig. 12A) repre- areas in the Sylhet trough: A. the Goyain trough (Fig. 2); B. the western, flat-lying part of the sents the passive-margin phase of basin evolu- basin. Dashed lines represent subsidence of the basement corrected for the load induced by the tion. Shelf facies to the northwest graded weight of sediment through time and, thus, the inferred amoung of "tectonic subsidence" southeast into basinal facies, with the interven- produced either by extension (pre-Oligocene?) or thrust loading (Oligocene and younger?). ing slope probably occurring near the north- Due to limitations of lithologic, bathymetric, and thickness data described in text, the plots are western part of the Sylhet trough. The shelf approximate. facies appear to have prograded over this slope in the Sylhet trough by late Eocene time (Rao, 1983). Due to lack of subsurface data, a con- stant subsidence rate was assumed on the geohis- in this more proximal setting. It therefore seems ward over at least parts of the Shillong Plateau tory diagrams between 120 Ma and the unlikely that Barail Formation sandstones on the (Figs. 1, 12C; Banerji, 1984; Salt and others, Eocene-Oligocene boundary (Fig. 11). It seems northern margin of the Sylhet trough were de- 1986). This greater rate of subsidence probably more likely, however, that subsidence would re- rived from the Indo-Burman ranges (they would reflects increased proximity to, and loading semble that of other passive margins, which dis- need to have been transported across the in- from, the encroaching Indo-Burman ranges. play concave burial history curves (Watts and ferred marine embayment to get there). Instead, Crustal loading from the Himalayas may have Steckler, 1979). Barail sediments were probably derived from had a minor effect; however, the thrust front Subsidence during Oligocene time is shown northern and northwestern sources in incipient (Main Central thrust fault) of the uplifting Him- as continuing at about the same rates as in the Himalayan uplifts, with rivers flowing toward alayas was located much farther from the Sylhet pre-Oligocene but, given the general trends for the marine trough in the most rapidly subsiding trough than the Indo-Burman Ranges at this passive margins described above, the slope of the part of the basin. Barail sandstone compositions time, and the Miocene section in this region does curve may represent a small acceleration. (Table 1; Fig. 7) indicate recycled orogenic not thicken to the north (Fig. 4). Surma Group Brunnschweiler (1966) and Rao (1983) sug- sources, consistent with this inference. Oligo- sandstone compositions record derivation from gested that orogenic uplift began in the Indo- cene clastic detritus buried the Eocene carbon- recycled orogenic, mixed, and transitional con- Burman ranges at this time (Fig. 12B), and it is ate/clastic shelf and marks the transition from tinental sources, and they contain more feldspar likely that crustal loading from thrusting in this passive-margin to convergent-margin sedimenta- than do Barail Formation sandstones (Fig. 7). nearby orogeny affected the Sylhet trough and tion in the eastern Himalayas. In Figure 12B, This increased feldspar content indicates deriva- areas to the east. This inference provides an ex- Oligocene deltas are shown slightly north of the tion from crystalline rocks, uncommon in the planation for the greater thickness of the Barail Sylhet trough, a reconstruction that subtracts the Indo-Burman ranges but a significant compo- Group east of the Sylhet trough in the Naga inferred southward tectonic transport of the Shil- nent of the eastern Himalayas (Ray, 1982; Mol- Hills (Figs. 1, 4), which would have occupied a long Plateau discussed in the next section. nar, 1984). The Shillong Plateau, also underlain position in the more rapidly subsiding, proximal Subsidence rates in the Sylhet trough clearly by crystalline rocks, was submerged at this time, part of a foreland basin. Rao (1983) interpreted accelerated during the Miocene epoch (Fig. 11) and so sediment derivation from the Himalayas the Barail of the Naga Hills as mainly deltaic when the Surma Group was deposited during a seems highly likely. The transition in composi- and showed a marine embayment (his Fig. 27C) prolonged transgression that extended north- tion from the Barail Formation to the Surma

away and could not have created this large ef- incipient Vvu? Himalayan w fect. We propose that this dramatic subsidence was forced by south-directed overthrusting of the Shillong Plateau along the Dauki fault. Two additional lines of evidence support this inference. First, the noted southward thinning of the Tipam-Dupi Tila interval (Figs. 2, 6) supports crustal loading on the northern basin margin. Second, both gravity and seismotectonic data are most consistent with a north-dipping thrust below the Shillong Plateau (Seeber and Arm- bruster, 1981; Molnar, 1987). Major uplift of the Shillong Plateau clearly B. Oligocene -gi-r-i-i-; began in the Pliocene Epoch. There is no record of significant Pliocene-Pleistocene deposition on the Shillong Plateau (Fig. 4), consistent with its interpretation as an emerging block. The facies change between the mainly fine-grained, prodelta Surma Group to the coarse-grained, braided-fluvial Tipam Sandstone indicates an abrupt change in basin geometry and tectonics. . s / -Shillon g / fi// < / •If Plateau /• w / The Surma Group-Tipam Sandstone transition ' * ^ I Ë ' is also marked by a change in sandstone petrol- Dauki ogy (Fig. 7), which indicates a new sediment fault source. The significant increase in sedimentary 17 lithic fragments in the Tipam reflects erosion of the sedimentary cover of the uplifting Shillong D. Pliocene Plateau and, to a lesser extent, the Indo-Burman ranges. The petrology of the Dupi Tila Forma- Positive area Delta-front and tion similarly reflects a recycled orogen source. prodelta deposits Only very recently, as recorded in modern flu- Limestone Basinal deposits vial sands that plot in transitional continental and craton interior provenance fields (Fig. 7), Fluvial and delta- -A— Thrust fault plain deposits has the crystalline basement of the Shillong Pla- Anticline, showing plunge teau been exposed. Uplift of the Shillong Plateau also led to a Figure 12. Schematic paleogeographic reconstructions showing four stages in the Cenozoic major reorganization of rivers draining the Him- evolution of the Sylhet trough. Reconstructions in A, B, and C restore inferred Shillong Plateau alayas. The Brahmaputra River, which probably southward displacement. MBT, Main Boundary thrust fault. See text for discussion. Other flowed into the Surma Group delta during the schematic paleogeographic reconstructions for this area are in Rao (1983) and Baneiji (1984). Miocene epoch (Fig. 12C), was apparently deflected about 300 km to the west to its present course skirting the Shillong Plateau (Fig. 1). Group records progressive Himalayan unroof- stone thin over anticlinal crests (Hiller and ing. Sedimentary fades data indicate that the Elahi, 1984), indicating that the frontal zone of DISCUSSION Surma Group delta system was large and later- Indo-Burman ranges deformation had advanced ally extensive, suggesting that the rapidly subsid- into the eastern Sylhet trough in Tipam Sand- Interpretation of the Dauki fault as an over- ing trough on the western flank of the Indo- stone time (early Pliocene). The Sylhet trough thrust is consistent with Molnar's (1987) inter- Burman ranges (of which the Sylhet trough was continued to subside rapidly after the folding pretation of the seismotectonics of the region. a part) drained a large proportion of the eastern associated with the advancing Indo-Burman Geologic data presented here, however, allow its Himalayas. ranges commenced. Along strike to the south, history as a thrust fault to be extended to early The Pliocene and Pleistocene history of the these folds have considerable topographic relief Pliocene time and provide the basis for crude Sylhet trough (Fig. 12D) is characterized by in the Tripura area of India (Fig. 3). It is there- estimates of the amount of overthrusting. The much higher (3 to 8 times) subsidence rates (as fore unlikely that much of this increased rate of Shillong Plateau subsided more slowly than the much as 1,200 m/m.y.) than those recorded for subsidence can be attributed to increased crustal Sylhet trough during Cretaceous to Miocene Miocene deposition of the Surma Group (Fig. loading from this orogenic belt. The frontal time, having a composite section for this interval 11). As described above, the Dupi Tila Forma- thrust of the Himalayas (by this time, the Main of -5,000 m (Fig. 4) and a maximum preserved tion and, to a lesser degree, the Tipam Sand- Boundary thrust fault) was more than 150 km section in any one place of only -3000 m

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021 1526 JOHNSON AND NUR ALAM

(Chakraborty, 1972). Thus, to uplift the base- styles of convergence change. The course of the trough on the west flank of the Indo-Burman ment of the Shillong Plateau to its present max- Jamuna River (the main distributary of the ranges. Barail Formation sediments were proba- imum elevation of about ~2 km, about 5-7 km Brahmaputra) now follows a major north- bly derived from incipient Himalayan uplifts. of vertical displacement are required. Molnar trending lineament in this region that is charac- Subsidence rates accelerated in the Miocene (1984, his Fig. 1) and Ni and Barazangi (1984, terized by recent faulting (Morgan and Mclntire, epoch, mainly in response to continued en- their Fig. 14) show dips of about 5° to 10° for 1959; Coleman, 1969; D. A. Coates, 1989, oral croachment of the Indo-Burman ranges. Mio- the Main Boundary thrust fault in the Hima- commun.). Alternatively, the amount of offset cene strata of the Surma Group were predomi- layas. If the Dauki fault has a similar dip or on the Dauki fault may diminish to the west nantly deposited in a large, mud-rich delta shallows rapidly to a similar dip, then the (implying clockwise rotation of the Shillong Pla- system that drained a large proportion of the amount of overthrusting required to produce the teau around a pivot point near its southwest eastern Himalayas. inferred minimum 5-7 km of vertical displace- border), as suggested by (1) the decreased thick- Subsidence rates increased dramatically from ment ranges from 28 kin (10° dip, 5 km of ness of the Tipam Sandstone-Dupi Tila Forma- Miocene to Pliocene time in the Sylhet trough. vertical uplift) to 80 km (5° dip, 7 km of vertical tion interval in the western, flat-lying part of The transition from the Surma Group to the uplift). Shallower or steeper dips on the inferred Sylhet trough (Fig. 6), suggesting less crustal Tipam Sandstone is marked by major facies Dauki thrust fault would require more or less loading to the west, and (2) the increased thick- (prodelta to braided river) and petrologic horizontal displacement, respectively. Increased ness of sedimentary cover on the Shillong Pla- changes. These changes are attributed primarily thickness estimates for the amount of sedimen- teau at its west end (Chakraborty, 1972), to the effects of south-directed overthrusting of tary cover eroded from the Shillong Plateau indicating decreased vertical uplift in that area. the Shillong Plateau on the Dauki fault for the would increase the estimates of both vertical up- Burbank and Raynolds (1988) described a following reasons. (1) Pliocene and Pleistocene lift and horizontal transport. This approximate synchronous partitioning of convergence be- strata are markedly thinner away from the Shil- estimate of a few tens of kilometers of Pliocene tween the Main Boundary thrust fault and the long Plateau, consistent with a crustal load to Holocene horizontal tectonic transport pro- Salt Range thrust, and significant rotation in emplaced on the northern basin margin. vides a mechanism for some of the facies thrust sheets, in the vicinity of the Northwest (2) South-directed overthrusting of the Shillong and(or) thickness changes in Mesozoic and Ce- Syntaxis at the west end of the Himalayas in Plateau is consistent with gravity data and with nozoic strata between the Shillong Plateau and Pakistan. This style of deformation has not been recent seismotectonic observations (Seeber and the adjacent Sylhet trough and Naga Hills reported along the linear (~l,800-km-long) Armbruster, 1981; Molnar, 1987). (3) Pliocene (Evans, 1964). front of the main part of the Himalayas, and rocks are not represented on the Shillong Pla- Accepting the extremely tenuous estimates of therefore may be a general characteristic of short- teau, suggesting that this massif was an emergent offset above, possible rates of horizontal dis- ening in syntaxial areas. block at this time. If the Dauki fault dips northward at an angle similar to that of other major placement on the Dauki fault range from 5.6 to Finally, it is emphasized that, if our model for Himalayan overthrusts (about 5° to 10°), then a 16 mm/yr (28-80 km over 5 m.y.). The present Pliocene to Holocene overthrusting of the Shil- few tens of kilometers of horizontal tectonic rate of convergence between the Indian and long Plateau is correct, then the overthrusting transport would be required to carry the Shil- Eurasian plates is about 50 mm/yr, but Molnar occurred simultaneously with folding of the long Plateau to its present elevation. (1984) estimated that only 10-25 mm/yr of this Indo-Burman ranges in the Sylhet trough. These amount is accommodated across the Himalayas. two contractile deformations, with nearly per- Thus, a significant proportion of the conver- pendicular trends, are occurring at the same ACKNOWLEDGMENTS gence in the eastern Himalayas over the past 5 upper-crustal level. The Indo-Burman ranges m.y. may have been taken up on the Dauki fold belt is being depressed by the Shillong Pla- This research was conducted as a part of the fault. Overthrusting on the Dauki fault may teau into a foreland basin. Geological Survey of Bangladesh project, "Ac- have an oblique (dextral) component, as sug- celerated Exploration for Mineral Resources gested by the northeastward deflection of CONCLUSIONS and Modernization of the Geological Survey of Pliocene-Pleistocene folds in the Sylhet trough Bangladesh," funded by the Asian Development (Fig. 2). The pre-Oligocene history of the Bengal Basin Bank. We gratefully acknowledge the research The Main Boundary thrust fault, the present was characterized by sedimentation on a passive and logistical support provided by S.K.M. Ab- frontal thrust of the Himalayas, was also in- continental margin associated with Mesozoic dullah, Director General of the Geological Sur- itiated in the Pliocene Epoch. It therefore seems rifting and opening of the Indian Ocean. The vey of Bangladesh; M. R. Khan, M. D. likely that, since Pliocene time, convergence in Sylhet trough likely occupied a transitional set- Kleinkopf, M. J. Terman, and J. W. Whitney. this area has been partitioned between thrust ting on this margin between shelf deposits to the We thank the Bangladesh Oil, Gas, and Miner- faulting below both the Himalayas and the Shil- northwest, and basinal deposits to the southeast. als Corporation (BOGMC) for permission to long Plateau. It follows that the rate of under- Subsidence rates may have increased slightly describe and sample core, to examine and re- thrusting of the Brahmaputra valley just north of during the Oligocene Epoch as a response to produce seismic data and well logs, and for the Shillong Plateau beneath the Himalayas is crustal loading from the developing Indo- accommodations at Barlehka and Sylhet. We less than that of the Indian Shield beneath the Burman ranges. Oligocene strata (Barail Forma- were helped in the field by A. Akhter, M. Ali, Himalayas west of the plateau (Molnar, 1987). tion) on the northern (probably allochthonous) E. M. Brouwers, S. M. Hossain, M. R. Khan, A significant discontinuity having right-lateral margin of the Sylhet trough are fluvial, but they N. Uddin, and the Bangladesh Rifles (national displacement may occur, therefore, on the west- probably passed abruptly (to the southeast) into border guard). We benefitted from the technical ern margin of the Shillong Plateau, where the delta and prodelta deposits in a rapidly subsiding help of V. F. Nuccio and from stimulating dis-

cussions with D. W. Burbank, M. Ismail, De Raaf, J.F.M., Boersma, J. R., and Van Gelder, A., 1977, Wave-generated Murty, K. N., 1983, Geology and hydrocarbon prospects of Assam shelf- structures and sequences from a shallow marine succession, lower Car- Recent advances and present status: Petroleum Asia Journal, v. 6, P. Molnar, W. J. Perry, Jr., C. J. Potter, A.H.M. boniferous, County Cork, Ireland: Sedimentology, v. 24, p. 451-483. p. 1-14. Desikachar, S. V., 1974, A review of the tectonic and geologic history of eastern Nandy, D. R., 1986, Tectonics, seismicity, and gravity of northeastern India Shamsuddin, and J. C. Yount. D. W. Burbank, India in terms of plate tectonics theory: Geological Society of India and adjoining region, in Geology of Nagaland ophiolite: Geological A. C. Christiansen, T. D. Fouch, K. J. Franczyk, Journal, v. 15, p. 137-149. Survey of India Memoirs, v. 119, p. 13-17. Dickinson, W. R., 1985, Interpreting provenance relations from detrital modes Ni, J. F., and Barazangi, M., 1984, Seismotectonics of the Himalayan collision E. A. Johnson, Neil Wells, and J. C. Yount of sandstones, in Zuffa, G. G., ed., Reading provenance from arenites: zone—Geometry of the underthrusting Indian plate beneath the Hima- Dordrecht, The Netherlands, Riedel, p. 333-361. laya: Journal of Geophysical Research, v. 89, p. 1147-1163. provided helpful reviews. Published with Evans, P., 1964, The tectonic framework of Assam: Geological Society of India Ni, J. F., Guzman-Speciale, M„ Holt, W. E., Wallace, T. C., and Seager, W. R., the approval of the Geological Survey of Journal, v. 5, p. 80-96. 1989, Aocretionary tectonics of Burma and the three dimensional Gupta, V. J., 1976, Indian Cenozoic stratigraphy: Delhi, India, Hindustan geometry of the Burma subduction zone: Geology, v. 17, p. 68-71. Bangladesh. Publishing Corporation, 344 p. Palmer, A. R., 1983, The Decade of North American Geology 1983 geologic Hiller, K., and Elahi, M., 1984, Structural development and hydrocarbon en- time scale: Geology, v. 11, p. 503-504. trapment in the Surma Basin, Bangladesh (northwest Indo-Burman fold Rahman, M. A., Mannan, M. A., Blank, H. R., Jr., Kleinkopf, M. D„ and REFERENCES CITED belt): Singapore, Fifth Offshore Southeast Conference, p. 6-5-6-63. Kucks, R. P., 1990, Bouguer gravity anomaly map of Bangladesh: Geo- Hohrop, J. F., and Keizer, J., 1970, Some aspects of the stratigraphy and logical Survey of Bangladesh, scale 1:1,000,000. Acharya, S. K., Mitra, N. D., and Nandy, D. R., 1986, Regional geology and correlation of the Surma Basin wells, East Pakistan: ESCAFE Mineral Rao, A. R., 1983, Geology and hydrocarbon potential of a part of Assam- tectonic setting of northeast India and adjoining region, in Geology of Resources Development Series no. 36, p. 143-154. Arakam basin and its adjacent region: Petroleum Asia Journal, v. 6, Nagaland ophiolite: Geological Survey of India Memoirs, v. 119, Ingersoll, R. V., Bullard, T. F., Ford, R. L., Grimm, J. P., Pickle, J. D., and p. 127-158. p. 6-12. Sares, S. W., 1984, The effect of grain size on detrital modes—A test of Ray, D. K., 1982, Tectonic map of south and east Asia: Hyderabad (Geological Baksi, S. K., 1965, Stratigraphy of Barail Series in southern part of Shillong the Gazzi-Dickinson point-counting method: Journal of Sedimentary Survey of India), Commission for the geological map of the world— Plateau, Assam, India: American Association of Petroleum Geologists Petrology, v. 54, p. 103-116. Subcommission for south and east Asia, 7 sheets, scale 1:5,000,000. Bulletin, v. 49, p. 2282-2294. Johnson, S. Y., and Alam, A.M.N., 1990, Sedimentation and tectonics of the Salt, C. A., Alam, M. M„ and Hossain, M. M., 1986, Bengal Basin—Current Baneiji, R. K., 1981, Cretaceous-Eocene sedimentation, tectonism, and bio- Sylhet trough, northeastern Bangladesh: U.S. Geological Survey Open- exploration of the hinge zone of southwestern Bangladesh: Singapore, facies in the Bengal Basin, India: Palaeogeography, Palaeoclimatology, File Report 90-313,44 p. 6th Offshore Southeast Asia Conference, p. 55-67. Palaeoecology, v. 34, p. 57-85. Khan, M.A.M., 1978, Geology of the eastern and the northeastern part of Sadar Seeber, L., and Armbruster, J., 1981, Great detachment earthquakes along the 1984, Post-Eocene biofacies, paleoenvironments, and paleogeography subdivision, Sylhet District, Bangladesh: Geological Survey of Bangla- Himalayan arc and long-term forecasts, in Simpson, D. W., and of the Bengal Basin, India: Palaeogeography, Palaeoclimatology, Pa- desh Records, v. 2,20 p. Richards, P. G., eds., Earthquake prediction—An international review: laeoecology, v. 45, p. 49-73. Khan. M.A.M., Ismail, M„ and Ahmad, M., 1988, Geology and hydrocarbon Washington, D.C., American Geophysical Union, Maurice Ewing Brunnschweiler, R. O,, 1966, On the geology of the Indo-Burman ranges prospects of the Surma Basin, Bangladesh: Singapore, Seventh Offshore Series, Volume 4, p. 259 277. (Arakan coast and Yoma, Chin Hills, Naga Hills): Geological Society of Southeast Asia Conference, p. 364-387. Sengupta, S., Ray, K. K., Acharya, S. K., and deSmeth, J. B., 1990, Nature of Australia Journal, v. 13, p. 137-194. Kuehl, S. A., Hariu, T. M., and Moore, W. S., 1989, Shelf sedimentation off the ophiolite occurrences along the eastern margin of the Indian plate and Brunnschweiler, R. O., and Khan, M.A.M., 1978, With Sherlock Holmes in the Ganges-Brahmaputra river system—Evidence for sediment bypassing to their tectonic significance: Geology, v. 18, p. 439-442. Bengal Basin: Singapore, Second Offshore Southeast Asia Conference, the Bengal fan: Geology, v. 17, p. 1132-1135. Shrivastava, P. K., Ganeshan, S., and Rady, D., 1974, Tipam Group in the 5 p. LeDain, A. Y., Tapponier, P., and Molnar, P., 1984, Active faulting and subsurface of upper Assam valley south of Brahmaputra: Geological Burbank, D. W„ and Raynolds, R.G.H., 1988, Stratigraphic keys to the timing tectonics of Burma and surrounding regions: Journal of Geophysical Society of India Journal, v. 15, p. 165-181. of thrusting in terrestrial foreland basins—Applications to the north- Research, v. 89, p. 453-472. SPARRSO (Space Research and Remote Sensing Organization, Government western Himalaya, in Kleinspehn, K. L., and Paola, C„ eds., New LeFort. P., 1975, Himalayas—The collided range—Present knowledge of the of the Peoples Republic of Bangladesh), 1984, Bangladesh LANDSAT perspectives in basin analysis: New York, Springer-Verlag, p. 331-352. continental arc: American Journal of Science, v. 275-A, p. 1 -44. satellite digital mosaic: SPARRSO, 1 sheet, scale 1:1,000,000. Chakraborty, A., 1972, On the rock stratigraphy and tectonics of the sedimen- Lietz. J. K..and Kabir,J., 1982, Prospects and constraints of oil exploration in Van Hinte, J- E., 1978, Geohistory analysis—Application of micropaleontology tary belt in the southwest of the Shillong Plateau, Meghalaya: Oil and Bangladesh: Singapore, Fourth Offshore Southeast Asia Conference, in exploration geology: American Association of Petroleum Geologists Natural Gas Commission (India) Bulletin, v. 9, p. 133-141. p. 1-6. Bulletin, v. 62, p. 202-222. Chen, W. P., and Molnar, P., 1990, Source parameters of earthquakes beneath Mitchell, A.H.G., and Reading, H. G., 1986, Sedimentation and tectonics, in Verna, R. K., and Krishna Kumar, G.V.R., 1987, Seismicity and the nature of the Shillong Plateau and the northern Indo-Burman ranges: Journal of Reading, H. G., ed.. Sedimentary environments and fades: Oxford, plate movement along the Himalayan arc, northeast India and Arakan- Geophysical Research, v. 95, p. 12,527-12,552. England, Blackwell Scientific Publications, p. 471-519. Yoma—A review: Tectonophysics, v. 134, p. 153-175. Coleman, J. M., 1969, Brahmaputra River—Channel processes and sedimenta- Molnar, P., 1984, Structure and tectonics of the Himalaya—Constraints and Watts, A. B., and Steckler, M. S., 1979, Subsidence and eustacy at the continen- tion: Sedimentary Geology, v. 3, p. 129-239. implications of geophysical data: Earth and Planetary Sciences Annual tal margin of eastern North America: Washington, D.C., American Curray, J. R., and Moore, D. G., 1974, Sedimentary and tectonic processes in Reviews, v. 12, p. 489-518. Geophysical Union, Maurice Ewing Series, Volume 3, p. 218 234 . the Bengal deep sea fan and geosyncline, in Burke, C. A., and Drake, 1987, The distribution of intensity associated with the great 1897 As- C. L., eds., The geology of continental margins: New York, Springer- sasm earthquake and bounds on the extent of the rupture zone: Geologi- Verlag, p. 617-628. cal Society of India Journal, v. 30, p. 13-27. Curray, J. R„ Emmel, F. J., Moore, D. G,, and Rant, R. W., 1983, Structure, Morgan, J. P., and Mclntire, W. G., 1959, Quaternary geology of the Bengal tectonics, and geological history of the northeastern Indian Ocean, in Basin. East Pakistan and India: Geological Society of America Bulletin, Nairn, A.E.M., and Stelhi, F. G., eds., The ocean basins and margins, v. 70, p. 319-342. Volume 6, The Indian Ocean: New York, Plenum, p. 399-450. Murthy, M.V.N,, Chakrabarti, C., and Talukdar, S. C., 1976, Stratigraphic MANUSCRIPT RECEIVED BY THE SOCIETY JULY 12,1990 Dasgupta, A. B., 1977, Geology of Assam Ar&kam region: Mining and Metal- revision of the Cretaceous-Tertiary sediments of the Shillong Plateau: REVISED MANUSCRIPT RECEIVED FEBRUARY 25,1991 lurgy Society of India Quarterly Journal, v. 49, p. 1-54. Geological Survey of India Records, v. 107, p. 80-90. MANUSCRIPT ACCEPTED APRIL 2,1991

Printed in U.S.A.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/103/11/1513/3381148/i0016-7606-103-11-1513.pdf by guest on 01 October 2021