sign in sign up
<<
Home , Bouma sequence, Clastic rock, Dish structure

GeoscienceCanada, Volume 3, Number 1, Februi

currents were known in lakes and thousands have been measured and reservoirs,and they appeared to be used to reconstruct paleoflow competent to transport for patterns in hundreds of turbidiie fairly long distances. Many of these basins. different lines of evidence were pulled 3) Within the graded beds. together by Kuenen and Migliorini in many different sedimentary 1950 when they publishedtheir structures were recorded. By thelate experimental results in a now classic 1950s, some authors were proposing paper on "Turbidity currents as a cause models, or ideal turbidiies. of graded bedding". A full review of why based upon a generalization of these and how the concept was established in and the Facies Models has recently been published sequence in which they occurred. (Walker. 1973). This generalization is akin tothe 2. and After its introduction in 1950, the distillation process discussed in the interpretation was previous paper, and the final Associated applied to rocks of many different ages. distillation and publication of the Coarse Clastic in many different places. Emphasis was presently accepted model was done laid upon describing a vast and new by Arnold Bouma in 1962. A version Deposits assemblage of sedimentary structures. of the Bouma model is shown in and using those structures to interpret Figure 1 paleocurrent directions. In the absence Roger G. Walker of a turbidite lacies model (see pevious Th. BOUMT~rbldll. hCi08 MOdd Department of Geology article in this issue of Geoscience The Bouma sequence, or model(Figs. 1, McMaster University Canada),there was no norm with which 2) can be considered as a very simple Hamilfon, Ontar10L8S 4M7 lo compare individual examples, no facies model that effectively carries out framework for organizing observations. all ol the four functions of facies models no logical basis for prediction in new discussed in the previous article. Iwill lntroduetlon situations, and no basis for a consistent examine these in turn, bdh toshed light To the sedimentologist. Me turbidity hydrodynamic interpretation.Yet upon turbidites in general, and to use current concept is both simple and gradually duringthe years 1950-1960, a turbidites as an illustration of a facies elegant. Each turb~dite(defined as the relatively small but consistent set of model in operation. I have described the deposit of aturbidity current) is the result sedimentary features began to be model as very simple because it of a single, short lived event, and once associated with turbidites. These are contains relatively few descriptive deposited, it is extremely unlikely to be considered in the following list, and can elements, and because it is narrowly rewciked by other currents. The now be taken as a set of descriptors for focussed upon sandy and siltyturbidites concept is elegant because it allows the classical turbidites: only. I shall later refer to these as interpretation of thousands of graded 1 ) Sandstone beds had abrupt, sharp "classical" turbidites. sandstone beds, alternating with . bases, and tended to grade upward as the result of a series of similar events. into finer . and mud. Some of I. The Bouma modelas a NORM. The and it can safely be stated that nosimilar the mud was introduced into the model (Fig. 1 ) as defined by Bouma volume of clastic can be interpreted basin by the turbidity current (it consists of five divisions. A-E, which so simply. conta~nedshallow benthonic occur in a fixed sequence. Bouma did In this review. I will begin by studying forams), but the uppermost very fine not give normalized thicknesses for the the "classical" turbidite, and will then mud contained bathyal or abyssal divisions, and this type ol information is gradually broaden the scale to benthonic forams and represented still unavailable. In Figure 3,l have encompass turbidites and related the constant slow rain of mud onto sketched three individual turbidites coarse clastic rocks in their typical the ocean floor. which clearly contain some of the depositional environments - deep sea 2) On the undersurface (sole) of the elements of the Bouma model, yet which fans and abyssal plains. there were abundant obviously differ from the norm. They can The concept of turbidites was markings, now classified into three be characterized as AE, BCE and CE introduced to the geological profession types: tool marks, carved into the beds. Without the model. we could ask in 1950. At that time, nobody had underlying mud by rigid tools (sticks. no more questions about these three observed a modern turbidity current in stones) in the turbidity current; scour turbidites, but with the norm, we can ask the ocean, yet the evidence for density marks, cut into the underlying mud by why certain divisions of the sequence currents had become overwhelming. fluid scour: and organic markings - are missing. I will try and answer this The concept accounted for graded trails and burrows - filled in by the rhetorical question later. sandstone beds that lacked evidence of turbidity current and thus peserved shallow water reworking, and it on the sole. The tool and scour 2. The Bouma model as a framework accounted for transported shallow water markings give an accurate indication andguide lor description. The model has forams in the sandstones, yet bathyal or of local flow directions of theturbidity served as the basis for description in a abyssal benthonic forams in currents, and by now, many large number of studies, particularly in interbedded shales. Low density Canada, U.S.A. and Italy. With the BOUMA DIVISIONS INTERPRETATION ...... :...... :...... I.. . FINES IN TURBIDITY CURRENT, FOLLOWED BY PELAGIC --', 3 p-' LOWER ...... TRACTION IN FLOW REGIME ...... UPPER T: 2

RAPID DEPOSITION, ? QUICK BED

I

Figure 1 lo ernphasfze that alln wealhered or fecton~zed Five drnsions of the Bourna model for OutcroDs ~tcannot be separared from E- ru,o orcs A-qrzoeo or mass .e wnosrone pe r c r , son [~

such quantities and at such a rate that water is forcibly expelled upward, and momentarily. the grainlwater mixture becomes fluidized (or "quick"). The flu~dizationwould destroy any possible sedimentary structures. The second phase of deposttion involves traction of grains on the bed. Flow velocities are lower, and the rate of depositton from suspension is much lower. By direct comparison wlth many experimental studies, division B represents the upper flow regime plane bed, and dtvlsion C, the lower flow regtme rippled bed. The thtrd phase of deposition lnvolves slow deposition of fines from the tail of the current. The origin of the delicate laminations in division (0)is not understood, and I preferto placedivlsion Figure 2 (D)In brackets, tmplyingthat in all butthe Com~lele'Bourna"twbldrle (see FIO 11 C D,nsrons fDI and E were broken off thrs -. , , cleanest outcrops. (D) cannot be shoi,ngpelitjc d~wsonE ofloierbed(bottorn speornen, whfch is from the Care Frechette leftl: oraded drvision A, oarallel larnrnaled roadcul. Levs Forrnatlon lCarnbrian1. separated from E. In the uppermost part drvrsron B and npple cross larnrnated drvrsron Quebec of divlsion E, there may be some true pelagic with a deep water (bathyal or abyssal) benthonic fauna frameuorn provtdea oy the mode one 3 Tnr.moor1 ;IS 3 bass 101 (forams in Tert~aryand younger rocks). rt!crod,nnrn r ir:rrr[~rttctl!u,lTne can aJlcK v loo a seawnce of t~rb~ o~ tes as AEIBCEICE etc.(as in the three existence of the Bouma model enables 4. The Bouma model as a predictor. turbidites of Fig. 3), and then add to the us to make one integrated interpretation Here, I shall show how the basic description any other features of of classical turbidites,rather than having hydrodynamic interpretation of the note. With the model as a framework, to propose different origins for each model, togetherwlth departures from the one is not only aware of the features different type of bed In Figure I,the norm, can be used on a predtctive basis. presented by any bed, but is also aware interpretation is considered in three Turbldile 1 (Fig. 3) begins with a thick of any features embodied in the model parts, Division A contains no sandy d~vision(A), and was deposited but missing in a particular bed. sedimentary structures except graded from a high velocity current. Turbldite 2 bedding. It represents very rapid settling (Fig. 3),by comparison with the norm. of grains from suspension, possibly in Geosciencecanada. Volume3, Number 1. February, 1976 27

does not contain division A. It begins with Bourna divlsion B, and was presumably deposited from a slower current. Turbidite 3 (Fig. 3) lacks divisions A and B. and presumably was deposited from an even slower current. In a caullous way, we can now make some predict~onsbased upon comparison wilh the norm, and uponthe hydrodynamic interprelations. A sequence of many tens of turbidites in which all of the beds are thick and begin with division A (Fig. 4, and, for example, the Cambrian Charny Sandstones in the St. Romuald road cut near Levis. Quebec) probably represents an environment where all of the turbidity currents were fast-flowing during deposition. Such an environment was probably close tothe source of the turbidity currents (proximal). By contrast (Fig. 5). a sequence of many tens of beds In which a1 the tdrolo tes oeg n ellher w~tnd~vlslon B or C rOroov~can Ut ca Formation at Montmorency Falls. Quebec) was deposited in an environment where all of the turbidity currents were flowing slowly during deposition. Such an environment was probably a long way from the source of the currents (distal). This conclusion will be slightly modifled below. This ideal proximal todistal scheme Figure4 applies only to "classical" turbidites. In ~ioiipof iowpara~~eisrdedlurb,dites.AE. AE, overfurned lop io r~ghl.Ordovrclan nature, variations in the size. sediment AE and AE, suggest,nglhaf the beds are Clorrdorme Formaf!on at Giande Vallee. load, and velocity of individual currents close fo lheir source (prox~mal).Beds slightly Ouebec. will blur the proximal to distal distinctions, whlch is why I Suggest taking the combined characteristics of a large number of beds before making environmental predictions. For example. if out of 250 beds. 70 per cent began wilh divls~onA, the environment could be characterized as relatively proximal. It follows from this application of the model that if one can work out the environment of deposition of a relatively large group of turbidites (let's say 300 beds - and a distal env~ronmenlis ind~cated),and one knows the general paleoflow direct~on,one can make prediclions as to what the same stratigraphic Interval will look like closer to source and in a specific geographic %&*% direction. The reader is now referredto ,qgure iron, lheirsourilridrsta!) Confrasf wrln Ftgure "A review of the geometry and facies Very r!nn furbidile sandsfones with fhfcker 4 Oidovicran Cioridorme Formalron. Grande organization of turbidites and turbidite- rnterbeddedshales Beds begrn wrlh Bouma Vall6e (near hsh cannery), Ouebec. bearing basins" (Walker, 1970). and, ~f div!srons B and C, andsuggesl deposifron lar straflgraphrc lop f0 left you are interested in the intimate details d lateral variability in classical turbidites, to an excellenl paper by Enos (1965) on the Ordovician Cloridorme coarse clastic faciesalso knownto have would be measured as A.A.A.A. using Formation in Quebec. been transported into very deep water the Bouma model. However. I would (as defined by bathyal and abyssal consider lhis to be a misapplication of Environments of Twbidite benthonic forams in interbedded the model, because its function as a Deposition shales). These facies can be listed as: norm. predictor, framework and basis lor Because a turbidite is simply thedeposit 1 ) massive sandstones hydrodynamic interpretation are all of a turbidity current, turbidites can be 2) pebbly sandstones seriously weakened to the point of found in any environment where turbidity 3) clast supported conglomerates uselessness if the beds only showan or density currents operate. These 4) chaotic -supported pebbly A.A.A.A. sequence. The massive environments include lakes and sandstones and conglomerates. sandstones are commonly not so reservoirs. delta fronts. continental This tacres list stems inillally from work parallel sided as the classical turbidites: shelves, and most importantly, the of Emiliano Mutti and his colleagues in channelling is more common, and one deeper ocean basins. However. to be Italy. and an English language version is flow may cut down and weld onto the preserved and recognized as a turbidite. available (Walker and Mutti. 1973). 1 now previous one ("amalgamation") giving the features imposed on the bed by the believe that the classification ol facies rise to a series of multiple sandstone current (ideally: sharp base with sole published by Walker and Mutti is beds. marks. graded bedding. Bouma unnecessarily subdivided (my opinion. The one common sedimentary d~wsions)must not be reworked by other not necessarily Mutti's), so I will stick to structure found in the masslve types of currents. Small turb~diteshave the simpler list above. sandstones is termed "dish" structure been preserved in quiet water glacial (Fig. 7), and is indicative of abundant lakes: thin prodeltalc turbidites can flow Massrve sandsfones This facies (Fig 6) fluid escape during deposition of the into water deep enough that agitation of conslsts of thick sandstone beds In sandstone. It indicates rapid deposition the bottom by storms is very rare (say. whlch graded bedding is normally poorly of a large amount of sand from a less than one storm in 500 years), but to developed Most of the dlvlsions of the "fluidized flow" (akin to a flowlng preserve a thick (hundreds or thousands Bouma sequence are mrsslng and quicksand). Thrs doesnot imply that the of metres) turbidite sequence. the most Interbedded shales tend to be very thin massive sandstone facies was likely environment is one that IS or absent A typical sequence of beds transported all the way from source into consistently deep and quiet Using the basin by a fluidized flow. However, it present day morphological terms. these environments would include the continental rise (made up of coalescing submarine fans) and abyssal plains it is important to emphasize that any sudden surge of sediment laden water can deposit a bed w~lhall the characteristics of a classical turbidite. A levee break in a river, and a rip current transporting sediment out across the continental shelf would be two examples of this Graded beds might be preserved In either situation, but the two environments would be characterized by the dominance of fluvial and shelf features, respectively. The presence of rare "turb~dites"would indicate the poss~bilityof density current activity, and would not condemn the entire sequences to deposition in great depths of water.

Other Facies Commonly Associated with Classical Turbldltes Classical turbidites can be characterized by three main features; first, the beds tend to be laterally extensive (hundreds of metres);second, rililddl Sl#ili(/i;iphtcIOP 10 feii Cimhia they tend to be parallel sided and vary M~SSIV~sandslone lac,es. Note lhrckness oi Ordovicidrl Cap Enlag6 Foinralion near St little in th~cknesslaterally (hundreds of bedsandabsenceolpel~Ocdrvrsron olBouma Sirnor,. Quebec metres): and third, it 1s reasonableto use Ihe Bouma model for this description and interpretation. However, along with classical turbidites there are other Geoscience Canada. Volume 3. Number 1. February. 1978 29

Figure 7 'Dish' sliuclures, lormed by rapld dewaler~ngp~pes(arrow on photo). dewarerlng 01 a massrve sandsfone. Some of Ordovrclan Cap Enrage Formation, near SI- the drsh edges curve upwardrnlo vertical Slrnon. Quebec.

does imply that a turbidity current. which normally maintains its sand load in suspension by fluid turbulence, can pass through a stage of fluidized flow Figure 8 during the final few seconds or minutes G,,~ilt,dbedof pebbly sandslone, followed of flow immediately preceding abruprly by a second bed wrlhour a pe1,Irc deposition. The massive sandstone d~vrsionSt-Damase Formalron (Ordovraan) facies is prominent in the Cambrian near Kamousaska, Ouebec Charny Formation around Ouebec City and Lgvis, and dish structures in massive sandstones are common in the Cambro-Ordovician Cap Enrage Formation near Rimouski. Quebec (Fig. 7).

Pebbly sandstones. The pebbly sandstone facies (Figs. 8.9) cannot be described using the Bouma model, nor does it have much in common w~ththe massive sandstone facies. Pebbly sandstones tend to be well graded (Fig. 8). and stratification is fa~rlyabundant. It can e~therbe a rather coarse, crude, horizontal strat~ftcation,or a well developed cross beddlng of the trough. or planar-tabular (Fig. 4) type. At present. there is no "Bouma-l~ke"model for the Internal structures of pebbly sandstones: the sequence of structures. and the11abundance and thickness has not yet been dlstllled into a general model Pebbly sandstone beds are commonly channelled and laterally Figure 9 discont~nuous.and Interbedded shales Peublv 13ndsI0nc. Idcies showmg m~d!urn from the Cambro Ordonaan Cap Enrage are rare. SLrile cross bedding in isolalron lhrs Formation/near Sl S!mon OuebecJ and6 Pholograph could easlly be conlused wrlh a lnlerbedded wilh lurb,d,les and graded Pholograph01 lluv,algravels but ul lac1,s pebbly sandstones It is clear that with abundant INVERSE - TO - channelling, and the presence of cross GRADED- GRADED-BED NORMALLY DISORGANIZED- beddina in pebbly sandstones, this STRATIFIED GRADED BED facies could eas/ly be confused with a coarsefluvial facies The differences are subtle and can be misleading to 7 sedimentologists - the safest way to approach the interpretation of pebbly sandstonesis to examinetheir context. If associated with. or interbedded with classical turbidites, the pebbly NO INVERSE NO INVERSC NO STRAT NO GRLlOlNG GRADING GRADlNG IMBRIChTEO NO INVERSE sandstone interpretation would bectear. STRhT. NO STRbT GRADlNG Similarly, if associated with non-marine CRUSS-STRbT IMBRICATED NO STRAT IMBRICATED lMBRiC RARE shales, root traces, caliche-likenodules. mud cracks, and other indicators of flood THESE THREE MODELS SHOWN IN SUGGESTED plain environments, the interpretation RELATIVE POSITIONS DOWNCURRENT would also be clear. This facies highlights the fact that environmental Figure 10 interpretations be based Won a Four models tor resedimented(deep wafer) graded-bed, andinverse-lo-normally graded "checklist" of features: the relative conglomerates. The graded-slratlhed, models are probably inlergradalional. abundance and type of features, in their stratigraphic context, must always be the bas6 of interpretation. Pebbly sandstones are particularly well exposed in the Cambro-Ordovician Cap Enrage Formation at St. Simon (near Rimousk~,Quebec). where grading, stratification and cross bedding are prominent. The facies is also abundant in the Cambrian St. Darnase Formation near Kamouraska, Quebec, and in the Cambrian St. Roch Formation at L'lslet Wharf (near St-Jean-Port-Joli, Quebec).

Clast supported conglornerales. Although volumetrically less abundant than class~calturbidites. conglomerates are an important facies in deep water environments. They are abundant in California and Oregon, and are particularly well exposed at many Flgure 11 sfraliliedaonglomeiale, very coarsr sandslone wrlh crude "dish"strucrureicen1ie localities in the Gasp6 Peninsula. Graded-sfral!lied conglomerale. Cambro- Ordovlcran Cap Enrag6 Formalron at Bic. oipholo) and !!nal!y inlo massive Sedimentologists to Quebec. Basal grades up info slruclureless sandstone (lop left) conolomerates, orobabtv because u . . without a facies model, there has been no framework to guide observations, and give rise to three models which are turbidites, the fabric isquite different: the hence the feeling of "not being quite probably intergradational, and a fourth tong axis isparallel to flow, and alsodips sure what to measure inthe field". I have (disorganized-bed) characterized only upstream to define the imbrication (Fig. recently proposed some generalized by the absence of discriptors. 12). This fabric is interpreted as "Bouma-like" modelsfor conglomerates One of the most important features of indicating no bedload rolling of clasts. (Walker. 1975). but because the models conglomerates is the type of labric they The only two reasonable alternatives are based upon less than thirty studies, possess. In fluvial situations, where involve mass movements (debris flows), they lack the universality andauthority of and cobbles are rolled on the or dispersion of the clasts in a fluid the Bouma model for classical turbidites. bed, the long (a-) axis is usually above the bed. Mass movements in The paper (Walker. 1975) discusses the transverse to flow direction, and the whichclasts are not freetomove relative models,their relationships,and howthey intermediate (b-) axis dips upstream. to each other do not produce abundant were established. In Figure 10, it can be characterizing the imbrication. However, graded bedding, stratification, and seen that the descriptors include the for most conglomerates associated with cross-stratification. so I suggest the type of grading (normal (Fig. 11) or inverse). stratification (Fig. 11), and fabric; in different combinations they Geoscience Canada. Volume 3. Number t . February. 1976

continental rise, Information on modern FABRIC fans is limited to short (1 -5 m) cores. surveys of surface morphology, and relatively little subsurface geophysical ~nformation.Ancient fans have been proposed on the basis of paleocurrent evidence, abundance of channels, and % distribution of facies. Two studies are ROLLING ON BE0 ABOUT FLOW NO ROLLING POSSIBLE FLOW outstandingly important - Normark's IN THIS ORIENTATION 0- (LONG) AXIS geophysical work and proposition of a fan growth model based exclusively Flgun 12 upon recent sediment work, and Muni Conlrasl &Ween cwglomerafe fabric and Ghibaudo's fan model based produced by rollrng clasls on fhe bed (long axis transverse to Vow) with typical labric in exclusively on ancient sediments. resedimenledconglomerates(norolling, long These two studies have been integrated axis DaraNel lo tlowl. into the review by Walker and Mutti (1973). Here. I will simply present the submarine fan - abyssal plaln model as clasts were supported above the bed in top of the bed. The deposit shows no it is currently understood (Fig. 13), fit the a turbulent flow. The support internal evldence of slumping. various facies into the various mechanism may have been partly fluid By contrast, the second type of morphological parts of the fan, and turbulence, and partly clast collisions. deposit commonly shows evidence of examine the stratigraphic Upon deposition, the clasts immediately slumping, and respresents the mixing of consequences of fan progradation. stopped moving (no rolling), and the sediment within the depositional basin Because of their generally parallel- fabric was "frozen" into the deposit. by post-depositional slumping. The sided nature, the classicalturb~d~tescan In the absence of experimental work deposits can range all the way fromvery be assigned to the smooth areas of the on cobbles and , the cohesive slumps involving many beds, fan - the outer suprafan lobes and the interpretation of the conglomerate to very watery slumps generated by the outer fan. The trend from proximal to models must be based largely on theory. deposition of coarse sediment on top of distal will develop most I suggest a downcurrent trend from the wet, poorly consolidated clays. The characteristically after the turbidites inverse-to-normally-gradedmodel, latter process gives rise to the classical have flowed beyond the confines of the through thegraded-bed model, into the pebbly . braided supralan channels. The graded-stratified model. This trend does Inasmuch as subaqueous debris massive sandstones and pebbly not necessarily exist in any one bed: flows, and slumps, require greater sandstones are less regularly bedded. rather.. de~osition, from a oarticular slooes than classical~ turbiditv~ ~~ currents. and the common presence of current in one of the three downstream thechaotic facies is most ab;ndant at channelling suggests that they be positions in Figure 10 will be of the type the foot of the slope into the basin, or in assigned to the braided suprafan indicated in the figure. the lnner Fan environment. Very few channels. As the channels become Clast supported conglomerates are examples have been described in plugged. and shln in position, a sand abundant in the Ordovician Grosses Canada. Large scale slumps are known body is gradually built up that consists of Roches Formation and Cambro- in Upper Ordovician turbidites in coalesced channels but no overbank Ordovician Cap Enrage Formation. northeastern Newfoundland (Helwig, deposits. In the absence of leveeson the Gasp4 Peninsula. Quebec, and also 1970).and pebbly mudstones are known suprafan, and with the lateral channel make up part of theCambrian St. Roch in several units in western shifting. any overbank fines that are Formation east of RiviBre-du-loup. Newioundland (Stevens. 1970) The deposited are rapidly eroded again. In Quebec. best described debris flows are nature, the gradual termination of the Devonian reef-margin examples suprafan channels is likely to result in a adjacent to the Ancient Wall. Miette and very gradual facies change across the sanosrones and conglomerates Thas So~tnes<-Ca~rnreel comp exes n s~pralanobes - some class cal

~~ ~ ~~ ~~ l~r0101lesm~qnl be preserved n vv oe lac es nc does-~ I*O ~-a~fterenl -~ tvoes, .~ of Aloena coo* el a1 , 1972 Sr~vastavaet deposit. First, there are conglomerates aL, 1972). shallow channels. and some unusually and pebbly sandstones that have large pebbly sandstone flows may spill abundant muddy matrlx, and possibly An Integrated Facies Model for out onto the smooth area of the show basal inverse grading and Turbldlter and Associated Coarse suprafan. preferred clast alignment. They Clastic Rocks Similarly. there is likely to be a s~milar represent the deposits of subaqueous The models discussed so far apply to facles change toward the feeder debris flows. Because the larger clasts relatively closely defined facies, and do channel, from pebbly sandstones Into inadebrisflowaremainta~nedabovethe not consider depositional environments, conglomerales (assuming that such bed by the strength of the Volumetrically, the turbid~tesand coarse clasts were available in the matrix, the deoosit commonlv has laroe associated clastics are most abundant source area). Conglomerates are blocks projecilng up above tketop ofihe in large submarme fans wh~chin many probably restrlctedto channels, malnly bed, or even restlng almost entlrely on areas have coalesced to form the the lnner fan channel, but alsoascoarse and pebbly sandstones as the braided FEEDER CHANNEL sLUw portion of the suprafan prograded The stratigraphically higher suprafan lobe sequences might therefore contain more massive and pebbly sandstones, CONGLCMERATES and fewer classical turbidites. The result of steady fan progradation RWI SSTS so far would be one thickening- and coarsening-upward sequence of YAOOlVE 5111 classical turbidites (outer fan), overlain by several thickening- and coarsening- upward sequences of classical turbidites, massive, and pebbly sandstones, representing several superimposed suprafan lobes that shined laterally and built on top of each CHLMNEL \ other during mid-fan progradation. The PlOXlYIL inner fan deposits would probably consist of one deep channel fill (Fig, 15). conglomeratic if coarse material were available at the source, and laterally equivalent to mudstones deposited on \ / OUTER FAN the channel levees and in the low areas behind the levees. It is DOSSible durina D(STAL NO mL.II"c %.LC il.LI0 - progadation, even in agenerally - aggrading situation, that the inner fan Figure 13 channel could cut intooneof the braided Submarrnelanenv~ronmenralmodelSeelexl suprafan lobes. lor IuNdrscuss~onD-8 is disorgan~zed-bed conglomerale model D F is debns flow Channel fill sequences, both in the inner fan and braided suprafan channels, may consist of "thinning- and lags in the bottoms of some suprafan a checklist to define environments - in fining-upward sequences" (Fig. 16). channels. The gradual downfan change th~scase, the abundance of CE beds Mutti and his ltaltan colleagues have from inverse-to-normally graded types and their facies relationships (with suggested that these sequences result tograded-stratifiedtypes is suggested in conglornerates, or with basin plain from progressive channel Figure 13, but this change is tentative muds) must be considered before an abandonment, depositing thinner and and IS ind~catedonly by theory, not by interpretation can be made. finer beds from smaller and smaller direct observation. The bottom of the flows in the channels. Thus an inner fan feeder channel and the foot ofthe slope Stratigraphic Aspects of Fan channel might have a conglomeratic are the most likely environments for Progradation basal fill. and pass upward into finer slumping and debris flows (D.F. in Fig. By comparison with a deltaic situation, conglornerates,and massive and pebbly 13) because of the steeper gradients. we can reasonably assume that sandstones. The disorganized-bed (D-B in Fig. 13) submarine fan progradation would result There are at leat two alternative conglomerates might also be assigned in a stratigraphic sequence passing stratigraphic records of submarine fans. here. from outer fan, through mid fan, into other than the steady progradation The inner fan levees are built up by inner fan deposits upwards in the discussed above. First, if supply for the flows which fill the channel and spills succession (Fig. 14). Progradationin the fan is cut off at source (or diverted onto the levees and the area behind the outer fan area would result in the elsewhere), the fan will be abandoned. levees. Sedlment consists only ofthe deposition of a sequence classical and will be covered by a rather uniform finest suspended material (silt andclay) turbidites that became more proximal in layer of hemipelagic mud. The but these may be sufficient current aspect upwards. This type of sequence previously active channels will also be strength to ripple the silt and produce is now termed thickening- and mud-filled. Abandoned mud-filled turbiditethat would be described as CE coarsening-upward". channels are known in the stratigraphic in the Bouma model. Hence although a The progradation of individual record, and include the Mississippi thick seqence of CE. BCE and C(D)E suprafan lobes might also be expected submarme channel (abandoned by beds probably does define a distal to result in thickening- and coarsening- post-Pleistocene rise of sea level), the

environment. a few~ siltv CE beds could u~wardseauences, but these mav not Rosedale Channel (Late Miocene. Great .~ , ~ also indicate levee or back-levee be restricted to classical turbidites. The Valley of ~aliforniajandthe oakum environments on the inner fan (a smooth, outer suprafan lobes would be Channel (Middle Eocene, Texas Gulf prox~malenvironment by any definition). represented by classical turbidites, but Coast). Again, I emphasize that one cannot use these would pass upward into massive Geoscience Canada. Volume 3, Number I. Februa~y.1976 33

(beds commonly begin with Bouma B FACIES SEQUENCE INTERPRETATION and C divisions) that appear more distal than proximal. The juxtaposition of - conglomerates in a channel, cutting into L relatively distal turbidites, suggests an SL. environment such as that labelled "incisedchannel" in Figure 13. INNER FAN D.F. F-U CHANNEL FILL Limitations of the Fan Model CGL The fan model presented here is based upon data from geophysical surveys of CGL \./ relatively small modern fans such as La 1 L Jolla. San Lucas, and the many other fans of the Southern California CHANNELLED Borderland. The model may not apply so PORTION MS. well to some larger fans (Monterey and F-U OF Astoria, off northern California-Oregon- SUPRAFAN Z LOBES a Washington: the Bengal Fan) because P.S. u they are characterized by major (CGL) cnannels wh ch cross tne entlre lengtn M.S. of the fan - Intnecaseolthe Benga. Fan. CHANNELLED 3 P.S. the channels are over 1000 km long. 0 cn However, the fan model as presented seems to be a useful framework for considering many small to medium C.T. C-U scale ancient basins. It cannot be SMOOTHI! Z applied to the long (hundreds of km) exogeosynclinal troughs in which the M.S. 2 a paleoflow pattern is dominantly parallel LZ 0 to the tectonic strike. Examples of C-u SMOOTH 2 C.T. PORTION V) turbidites in such troughs include the M. -- OF Ordovician Cloridorme Formation II SUPRAFAN (Gasp6 Peninsula) and its time LOBES equivalent in the Central Appalachians, C.T. c-U the Martinsburg Formation.Thedeposits -- consist dominantly of classical turbidites ir hundreds of metres thick, but showing no consistent proximal to distal change along the length of the trough in the C.T. c-U OUTER FAN downflow direction. It is commonly suggested that turbidity currents flowed downslope toward the trough axis. FININQ- OR perhaps constructing fans at the trough COARSENING- margin. However, at the trough axis the UPWARD flows turned and continued to flow Fiaure- 14 parallel tothe trough axis. The marginal n,p ,mcr ca s.or.,a, nc rho srrar yrdpn c otsi s 'cn .i,ron$ so.\ iricncnng- 3rro lans #err prcs~~nao/ ocstroyeo by )E )..once ~130~.o 0,. fan ..~rour3gar !,n ,'oarsen -7-.cn-ro .,cqxncrs lC-d s,o s.0scq~enl leclon cs an0 tne absence CT,classical lurbndite. MS.,massrve th~nning-and iin~ng-upwardsequences (F- of consistent proximal to distal changes sandstone, P.S.,pebbly sandstone; DF UJ See lext lor details along the trough axis is probably due to input from a whole series of fans along the trough margin. Thus any consistent Second, if the sediment supply coarser) being transported muchfarther changes developing from one source increases considerably, or the gradlent into the San Diego Trough. A poss~ble would be masked by inputfrom adjacent of the slope into the basin increases ancient example is the Cambrian St. sources up and down the trough. At (tectonically?),the fan channel may be Roch Formation at L'lslet Wharf (near St- present, there is no facies model that ~ncisedacross the entire fan, and all Jean-Port-Joli). Quebec, where a acts as a good predictor in this type of sed~menttransported much farther into thinn~ng-and fining-upwards sequence turbidite basin. the basin. This is the situation in the of conglomerates and pebbly modern La Jolla Fan (California), which sandstones rests in a channel (Fig. 17). has been entirely by-passed, with most The channel cuts into a thick sequence of the coarser sediment (sand and of relatively thinly bedded turbidites Henderson.J. 8.. 1972. of Archean turbidites at Yellowknife. Northwest Territories: Can. Jour Earth Sci., v. 9. p. 882-902. Turner. C. C. and R. G Walker. 1973. Sedimentology. strat~graphyand crustal evolutionof the Archeangreenstone belt near Sioux Lookout. Ontarto: Can. Jour. Earth Sci. v. 10, p. 81 7-845. Rousell. D. H.. 1972, The Chelmsford Formation of the Sudbury Basin - a Precambrian turbidite. m J. V. Guy-Bray. ed., New Developments in Sudbury Geology: Geol. Assoc. Can. Spec. Paper 10, p. 79-91 Cantin. R. and R G. Walker. 1972,Was the Sudbury Basin c~rcularduring deposition of the Chelmsford Formation7,in J. V. Guy-Bray, ed.. New Flgure 15 Developments In Sudbury Geology: Porllon of large channel cutting rnto shales Channel trll consals of drsorganrzed-bed conglomerates and lenlrcular sandstones, with an overallth,nnmg- andfmmg-upwardsequence Geol Assoc Can Spec. Paper 10, p 93- Ordovrcran Grosses Rocks, Ouebec, Apoalachrans 101.

2. Appalachian area Enos. P ,1969. Anatomy of aflysch. Jour. Sed Petrol.,v. 39, p. 680-723. (Note thts IS the classlc paper on the Cloridorme Formation.) Parkash. 8.. 1970. Downcurrent changes in sedimentary structures In Ordovtctan turbldtte greywackes Jour Sed Petrol.. v. 40, p. 572-590. Parkash. B. and G V Middleton. 1970. Downcurrent textural changes in Ordovician turbidite greywackes. Sedimentology. v. 14. p. 259-293 (Note: these two papers by Parkash are detalled studies of the Cloridorme Formation.) Sk~pper.K.. 1971. Antidune cross- stratillcation in a turbidite sequence Cloridorme Formation. Gaspe, Ouebec. Sedimentology. v. 17. p. 51 -68. (Seealso Flgure 16 cont,?rns larqe boulders which dre our upward d~scussionof this paper, Sedimentology. Ei?mole of a lhrnnlng and ltning upward 110wdrd~oplefll Cenlre of sequence s a v 18, p 135-138.) sesuence (see f cgure 141 from (he Cambro oebble conglomerate passing into pebbly Skipper. K. and G. V. Middleton. 1975, Ordovicran Cap Enrag6 Formatron near SI- sandslones (centre left) and lrnally rnlo The sed~mentarystructures and Simon The conglomerate (lower rrghll massrve sandstones (near water s edge) deposttional mechanics of certain Ordov~cianturb~dites. Clorldorme Canadlan Examples: Turbidltes and because they are important Formation.Gaspe, Quebec: Can. Jour. Associated Coarse Clastlcs contributions to a general understanding Earth Sci., v. 12, p. 1934-1952. The papers ltsted below do not of turb~dites Hubert, C.. J. Lajoie and M. A. Leonard, constitute a general set of readings w~th 1970, Deep sea sediments in the Lower respect to an introduction to theturbidite 1. Precambrian turbidites Paleozoic Ouebec Supergroup, in J. concept. Rather, they are significant Walker. R. G. and F. J. Pettijohn. 1971 Lajote. ed.. Flysch Sedimentology in contrtbutions to Canadian geology, Archean : analysis of the North America: Geol. Assoc. Can. Spec. either because they discussturbidites Minnitaki Basin, northwestern Ontario. Paper 7, p. 103-125. (Note: the main and the~rimportance to specific Canada: Geol. Soc Am. Bull.. v. 82. areas discussed in the paper are L'lslet problems of regional geology. or p. 2099-21 30. GeoscienceCanada.Volume 3, Number I.February. 1976 35

Canada, In J. Lajoie, ed.. Flysch Sedimentology in North America: Geol. Assoc. Can. Spec. Paper 7.p.13-35.

5. Western Canada Danner. W. R.. 1970. Western Cordilleran llysch sedimentation. southwestern British Columbia, Canada. and northwestern Washington and central Oregon, U.S.A.,~n J. Lajoie. Flysch Sedimentology in North America: Geol. Assoc. Can. Spec. Paper 7, p. 37-51. Cook. H. E.. P. N. McDaniel. E. Mountjoy and L. C. Pray. 1972. Allochthonous carbonate debris flows at Devonian bank ("reef") margins, Alberta, Canada: Bull. Can. Petrol. Geol.. v.20, p.439-497. Srivastava. P.. C. W. Stearn, and E. W. Mountiov. 1972. A Devonian Figure 17 loregrourirl (wilhgcnlogrsl lor scale), andcirll megabreccia at the margin of the Channel in Carnbrsan St Roch Formalron at ar lop right Note the graded-stralrlred Ancient Wall carbonate complex. L Isle1 Wharl Quebec Stralrgraphlc lop lo conqlornerate I~llmglowerparl 01 channel, Alberta: Bull. Can. Petrol. Geol.. v. 20. nghl Channel cuts tnlo classical lurbrd!les and passing up Into masswe sandstone p. 41 2-438. andcons~slsolal least two rnam porhons- (lower right) (Note: It seems astonishing that so little work has been published on the deep marine clastic sediments of the Western Wharf. and the Cap Enrage Formation in tectonics in west Newfoundland and Cordillera. The area should command the Bic - St. Fabien area. See also their possible bearing on a Proto- the immediate attention of Canadian Rocheleau and Lajoie, and Davies and Atlantic ocean, ;n J. Lajoie, ed.. Flysch sedimentologists. My own casual Walker, below.) Sedimentology in North America: Geol. observations on field trips suggest that Assoc. Can. Spec. Paper 7, p. 165-177. Rocheleau, M. and J. Lajoie. 1974. at least parts of the Miette Group Sedimentary structures in resedimented Horne. G. S. and J. Helwig, 1969. (Precambrian. Windemere) and Aldridge conglomerate of the Cambrian flysch, Ordovician stratigraphy of Notre Dame Formation (Precambrian, Lower Purcell) L ISel Oueocc Appalacn~ans-o~r See Bdy Newio~ndlana!n M

2. Modern submarine fans Other references cited in this article Normark. W. R.. 1974. Submarine Bouma, A. H., 1962, Sedimentology of canyons and fan valleys: lactors Some Flysch Deposits: Amsterdam. affecting growth panerns of deep sea Elsevier Publ. Co.. 168 p. Cited only as fans,in R. H. Dott. Jr. and R. H. Shaver, the first documentation 01 the now- eds., Modern and Ancient Geosynclinal accepted turbidite model. Sedimentation: Soc Econ. Paleont. Min. Kuenen, P. H. and C. I. Migliorini, 1950. Spec Publ. 19, p. 56-68. Turbidity currents as a cause of graded An updated version ol Normark's bedding: Jour. Geol., v. 58, p. 91-127. original (1970) discussion of fan growth Cited for historical reasons, as theflrst Nelson C H and L D Kulm. 1973. paper that directed geologists' atlention Subrnar~nefans and deep-sea tothe possibility of high density turbidity rhann~lsIn G V Mlddleton and A H current deposits in the geological Bo~maeds Tbrb 0~teSan0Dee~~alel recoro Tn~spaper represents one of tne Sedtmentat on Pac Itc Sect on Soc mnst.. ~ ~moortantlo~noallon stones ol r~ ~ - ~~ ~ Econ. Paleont. Min. Short Course Notes modern (post World War II) (Los Angeles), p. 39-78. sedimentology. Although emphasizing the N.W. Pacific. this revlew paper, with abundant relerences, is a good overall summary ol fan morphology and sedimentation.