LACUSTRINE SEDIMENTOLOGY OF THE TSAGANTSAVFORMATION: TAVAN HAR, SOUTHEAST MONGOLIA

JUSTIN GOSSESFranklin and Marshall College

HASHGEREL BAT-ERDENEMongolian University of Science and Technology

Sponsor: Carol DeWet

INTRODUCTIONThe late Mesozoic geologic history ofsoutheastern Mongolia is poorly understood.The English-language literature in particular issparse. The purpose of this study is to interpretthe depositional environments of thelowermost Cretaceous Tsagantsav Formationin the Tavan Har area, and identify the mainvariables that controlled deposition.Understanding the depositional environmentsof the Tsagantsav Formation will helpdelineate the regional geology. For example,early Cretaceous extension in the North China– Mongolia block is not well understood.Proposed drivers include rotational strike slipwith extension, gravitational collapse, escapetectonics, and ocean slab break off, resultingin lithosphere stretching and magmatic under-plating (Meng, 2003; Graham et al., 2001).Interpretation of the lower CretaceousTsagantsav Formation lithologies aids ininterpreting coeval tectonic extension.Lacustrine strata of the Tsagantsav Formationhave been identified as a possible source rockfor hydrocarbons in southern Mongolia(Johnson et al., 2003). Lower Cretaceousdeposits source the Zuunbayan oil field, theonly producing oil field in southern Mongolia.Several large petroleum systems of similar ageand origin have also been discovered to thesouth in China. Understanding the forcescontrolling deposition in the TsagantsavFormation could help in understanding thepetroleum resources of this area.

On a larger scale, this work might also assistin understanding processes of rift evolutionand deposition in areas where volcanicprocesses are a significant sediment source.

GEOGRAPHYThe Tsagantsav Formation occurs in multiplenortheast-southwest oriented grabens or half-grabens, filled with fluvial and lacustrinedeposits mixed with occasional bimodalvolcanic deposits and conglomerate (Grahamet al., 2001). The Tavan Har locality is locatedsouthwest of Zuunbayan, with the Unegt sub-basin to the north and the Zuunbayan sub-basin to the south. The North Zuunbayan faultborders the north side of Tavan Har. Fivestratigraphic sections were measured fromwhat seismic images indicate is a graben alongthe south side of the North Zuunbayan fault(Johnson, C.L., in press).

DEPOSITIONALENVIRONMENTSSection 2: Basal Tsagantsav strata includegreen and brown conglomerate mixed withvarying amounts of reworked ash. Above 12meters of conglomerate, very coarse sandstonemixed with ash is inter-bedded with theconglomerate for 36 meters. Gastropod,mollusk, and plant fossils occur in the sandbeds. Coarse sands are replaced by mudstoneover only a few centimeters. The mudstonecontinues uninterrupted for over 130 meters.In the fine-grained deposits, invertebratefossils become sparse and less diverse.

Figure 1: Measured sections 3, 4, 2, 6-7, and 5. The lower right diagram shows lateral and stratigraphicrelationships for all five sections.

The conglomerate beds represent an alluvialenvironment that is gradually becoming morefluvial as the proportion of sand-sized grainsand fossils increase up-section. This changecould result from decreased faulting orshifting drainage patterns. The sudden andcomplete nature of the change to fine-grainedmudstone seems to suggest expansion of alow-energy lake environment. The change infossil content may indicate the lacustrineenvironment was more alkaline relative to thefluvial environment.

Figure 2: Depositional environment for section 6-7

Section 6-7: This composite section ischaracterized by 151 meters of basalticconglomerate, inter-bedded with very fine-grain mudstones (Fig. 1). Thin-section revealsthere are two types of mudstone in section 6-7.The white mudstone consists of volcanic ashdeposited in an environment sufficiently low-energy to preserve crystal grains settling outbefore glass shards. Heulandite is a commonzeolite in this lithology. Gray-tinted mudstoneconsists of very fine chert-like materialclouded by clays. The difference between thegray and white mudstone is believed to bevarying degrees of alteration, with the graymudstone being substantially more altered.This may be due to differences in waterchemistry or burial rate. The conglomerateclasts are most likely reworked from basalticflows known to occur in the lower TsagantsavFormation (Graham et al., 2001). Some ofthese basaltic lava flows were observed to thenorth and southwest of the section. A possibledepositional environment that would interbedvery coarse conglomerate with fine -grainmudstone and exclude any fluvial materials isshown in Figure 2. A fault that may haveserved as the border fault is located just northof the measured section.

The lacustrine strata in sections 2 and 6-7 arevery similar in appearance and are interpretedas being stratgraphically comparable (Fig. 1).Putting these strata into the lake-type schemedevised by Carroll and Bohacs (1999) isdifficult. This might be due to a lack ofsedimentary structures throughout the section6-7 or that their system assumesprecipitation/evaporation correlates to thesediment fill rate. In sections 2 and 6-7, thesediment is largely water-laid ash or basalticconglomerate. These strata are largelyvolcanic or tectonic in nature and have littlecorrelation with precipitation/evaporation, sothe system devised by Carroll and Bohacs(1999) must be used with caution in settingswhere fluvial input is minor relative to othersources of sediment.Section 5: The first half of section 5 isdominated by sandstone with smaller amountsof mudstone. The mudstone is either white anderodible or porcelanitic. The second half of theoutcrop is dominated by porcelanitic mudstone(Fig. 1).Medium-grained sand, upward finingsediment, climbing ripples, cross bedding, andsoft sediment deformation suggest a fluvial-marginal lacustrine facies with substantialvolcanic ash input indicated by occasionallytuffaceous sand and porcelanitic mudstone.The non-laterally continuous 58 meters ofporcelanitic mudstones towards the top of thesection suggests large ash falls that may haveoverwhelmed transport pathways. Thisscenario would have ash deposits filling in theoriginal deltaic channels, water eroding newchannels, then the next ash fall filling them inagain. The thickness of these beds makes thema good marker horizon for further work. Underthe scheme devised by Carroll and Bohacs(1999), this section would be an overfilledlake-type.Section 4: Section 4 is defined by a seriesof transgressions. The bottom layer in thisseries is carbonate sandstone that finesupwards into a uniform wavy mudstone,followed by white mudstone with interbeddedgreen and red hues that most likely representvarying oxidization conditions, followed byfinely laminated sediment representing a

deeper anoxic environment. This pattern isrepeated with a distinct erosion surfacesometimes being observed at the base of thesandstone units. Section 4 is an example of thebalanced-fill lake-type (Carroll and Bohacs,1999).Section 3: Section 3 consists of cyclicinterbeds of carbonate and non-carbonatesand, silt, and mudstone. Sediment input isfluvial, volcanic ash fall, and reworked ash(Fig. 1). Carbonate is secondary. Pore fluidbecomes less oxidizing over time. Orangemudstone with root casts, illite clays instead ofthe more common smectite clays, and areduction in chemical alteration of grains,indicate subaqueous conditions. Fish fossilsand pyrite in finely laminated mudstoneindicate deeper anoxic environments. Thecyclic appearance of section 3 is a compositeof three factors. Water depth varies accordingto accommodation space and/or climate. Grainsize is affected by fluxes in volcanic ash,fluvial sand, and distance from source.Carbonate is secondary and depends onporosity, which depends on grain size. Thepresence of montmorillonite and volkonskoitealso seemed to limit porosity. Section 3 bestfits the balanced-filled lake-type.

VOLCANIC SEDIMENTIONEvery measured section in the field area hasvolcanic ash in it. If the basaltic conglomerateis included as volcaniclastic material, section6-7 has the highest percentage ofvolcaniclastic material with an estimated>70%. Section 4 had the least with <10%.Depending on the amount of reworking, othersediment, degree of weathering, and speed ofburial, ash deposits in Tavan Har took onmany different appearances. Volcanic ashincludes both glass and crystals. Crystals arelargely felsic. Sericitization of glass, feldspars,and plagioclase is very common. XRD datasuggest secondary alteration products includeheulandite, smectite clays, amorphousmaterial, albite, and volkonskoite. Thepresence of heulandite, instead of analcime asfound in a core near Zuunbayam, could resultfrom different burial depths or geochemicalconditions in Tavan Har. The current workinghypothesis for the presence of volkonskoite, a

clay mineral high in chromium, is that chromewas altered from volcanic glass shards, thenconcentrated in volkonskoite with the help ofbacteria. Work is currently in progress tobetter define the petrology.

CONCLUSIONSBoth regional and local forces controldeposition of the Tsagantsav Formation atTavan Har. The basement derived, ubiquitousbasal conglomerate indicates regional tectonicactivity. Basaltic conglomerate further up-section can be connected to basaltic flowsfound throughout eastern Mongolia. Outcrops3, 4, and 5 indicate overfilled to balance-filledlake conditions. These conditions, typical oflacustrine strata in the Tsagantsav Formation,suggest a relative balance betweenprecipitation/evaporation and the potentialaccommodation rate. The sudden andcomplete change from fluvial to lacustrinedeposits in section 2 and the lithologies ofsection 6-7 suggest that localized drainagepatterns occasionally limited fluvial input(Fig. 2).

REFERENCES CITEDCarroll, A.R. and Bohacs, K.M., 1999, Stratigraphic

classification of ancient lakes: Balancing tectonicand climatic controls: Geology, v. 27, p. 99-102.

Graham, S.A., Hendrix, M.S., Johnson, C.L.,Badamgarav, D., Badarch, G., Amory, J., Porter,M., Barsbold, R., Webb, L.E., and Hacker, B.R.,2001, Sedimentary record and tectonic implicationsof Mesozoic rifting in southeast Mongolia: GSABulletin, v. 113, p. 1560-1579.

Johnson, C.L., Greene, T.J., Zinniker, D.A., Moldowan,M.J., Hendrix, M.S., and Carroll, A.R., 2003,Geochemical characteristics and correlation of oiland nonmarine source rocks from Mongolia: AAPGBulletin, v. 87, p. 817-846.

Johnson, C.L., in press, Polyphase evolution of the EastGobi basin: Sedimentary and structural records ofMesozoic-Cenozoic intraplate deformation inMongolia: Basin Research, v. 26

Meng, Q., 2003, What drove late Mesozoic extension ofthe northern China-Mongolia tract?: Tectonphysics,v. 369, p. 155-174.