I~-

1075

.-

JOURNAL OF SEDIMENTARYPETROLOGY,.VOL. 39, No. 3 ~,1074-11O6FIGS. 1-21, SEPTEMBER,1969

E!VKI'

,...'

JGRAIN SIZE DISTRIBUTIONS

tributed to provenance and to the hydraulics ofstream transport, but little environmental s}g-nificance WaS placed on the observations. Nogeneral hypothesis' was developed to explainwhy the same modes should appear both in flu-vial and mariJ1esediments. '

One of the, most significant of the early pa-pers on' texture was by Doeglas (1946). Heconcluded that grain size distribUtions follow anarithmetic probability law. Two majDr contribu-fions by Doeg:laswere that (l) grain siz~ dis-tributions are' mixtures of two or more compo-nent-,distributi~ris 'or popu!a!16ns,. and that" -(2)these distributions were produce<l by varyingtransport' conditions. From his analyses he de-veloped an ,empirical classification of curveshapes and related typeSDf curves to specificsediment~ry environments.. There were severalproblems in this type of analysIs: (1) a sedi-mentation balance was used for textural analy-sis which did riot provide sufficiently accurate orreproducible tesults; (2) cumulative distribu-tions were plotted on arithmetic probabilitypaper, which tended to ininimize the finegrained tail and strongly accentuated the coursefraction; (3) the mixing and truncation. ofcomponent' distributions was not observed; (4)curve shapes were not related to specific deposi~tional. processes. -

Regardless of these: limitations Doeglas' con-'tribution was norsufficient1yrecognized py sedi-mentologists, and this rather fruitful approachto the recognition of s.edimentary environmentswas not widely adoptedin this coqntry.

One of, the most significant papers. relatingsedimentation. dynamics to texture was pub-lishedby Inma"n (1949). He recognized th~tthere are three fundamental modes of tranSport,surface creep, saltation, and suspension (Inman,1949, p. 55), and Qe utilized the existing knowl-edge concerning fluid mechanics to analyze themooes of transport of .sedimentary 'particles.Much of the work in this area had been devel-oped by Gilbert (1914), Shields (1936) ,Rubey(1938), Bagno1d (1941), and Kalinske (1943).Many other workets aided in the developmentof these concepts, but the above writers relatedfluid mechanics directly to the problems obedi-ment transporLand deposition. ..

Preliminary' eonclusions concerning' sorting,skewness, and mean size were derived by Inman(1949). He did not; however, relate these pa-rameters to the total ,grain siz~ distributions orto the presence of individual populations, as hadbeen suggested earlier by Doeglas. Inman'swork formedtlie basis for the emphasisduringthe 1950's and 1960's on statistical measures ofthe grain size distribution and on the continued

I

I.GLENN S. VISHERj

University of Tulsa~Tulsa, Oklahoma

~

Ex'~"", 'OX,,"~!"i,d, oJ ""Ie "",,=~'~;::;;'::rt'" ..., ,,~~d'" ''''.,J.'' ;"., .,,,,,, ;.""jpretatjon of saqd texture. AnalysIs 1S based on rec~gnlzmg ,SUD~pOpUlat10ns wl~lIin iii~iviuual log-normalgrain size distributi?!1s. Each log-~.rmal sub-populatIOn '!lay ,be relate~ to a dlffer.ent _mode oiseiiimenttransport and depositIOn, thus providing a measure o.f their Importa.nce In the. genesIs ot a §and unit. Thethree modes of transport reflected are: (1) ~t,lspe';lslOn;. (?) salta~lOl1;; anf (3) .surface creep or roning.Each of --these is developed ,as a separate sub-populatIOn wlthm a, gram. Size d!stqbutl(>n, The number, amount,sizecrange, mixing, and sorting of these. populations v;;ry systematically m r~latlOn to. provenance, sedi--,mentary process,and'sedimentarydynamlcs. The analysIs of these parameters IS the basIs for detenniningthe process-response characteristics ~f individual san~ units., ,- ," '.. -.'

.A ,numoer of processes are Uluquely reflected ,m log-probablhty, curves of graIn size dlstnbutions of

sands and sandstones. These include: (I) current; (2) ~lVash and backwash;"(J) wave; (4) tidal channel;(5) fallout from suspension; (6) turbidity current; and (7) aeolian {june. The combination of two ormore of these processes also produce characteristic log-probability curve shapes.

'

" 'Ancient ~ands' show some differences from their modern analogues, but these are usually minor. Log- 'probabilitY plots .of ancient sands are directly comparable to thQse from mQdern sands. The principal limita-tion of this study is in comparing sands fonned under comparable conditions and obtaining an independel}tdetermination 0 f the processes of formation of ancient sands."

"-.

GRAIN SIZE DISTRIBUTIONS AND DEPOSITIONALPROCESSESI

INTRODUCTWN

Statement of Problem

For many years sedimentary petrographershave attempted to use, grain size to, determinesedimentary environments. A survey of the ex-tensive literature on this subj ect illustrates thesteady progress that has been made toward thisgoal. Many e..'<:ce\lent contributions have beenmade during the past tWenty to thirty years,each providing new approaches and insightsinto the nature and significance of grain sized{stributions., Only within the past' few years,however, have workers attempted to relate grain-siie distributions to the depQsitional proce,ssesres~onsibJe for their formation. This approachappears to be paTtic;ulai'ly frttitful, and it pro-vides the basis for the next step towards a truly'genetic classification, of sedimentary textures.One of the major problems in the analysis ofgrain sizedistribi.ttions is that the samesedi-

. mentary processes occur within a number of en-viromnents and the consequent

~textural re-

sponse is similar.' Now that there are many-physical criteriaavailable.to identify specificdepositional environments, the textural studies

- do not need to stand alone, but can provide aseparate line of evidence to aid in interpretingclastic deposits of unknown origin. ,,

In summary, the problem lays in the relation'

1 'Manuscript receh'ed October 10, 1,968; reyisedMay 2, 1969.

of sedimentary prqcesses to texturaL responses.If these can be related to~peCific deposition~lenvironments, then a powerful. tool' will beavailable for interpreting the genesis of ancientclastic deposits. -

Pie'llions WorkThe development' of a genetic approach to

clastic textures has been a long and difficult one.Many workers have provided 'informationillidfurthered the development towards thisgoa!lconsequently, it is nearly impossible tQ trace theorigin ,o£.~any of the ideas. Specific conceptssl,lch as thelog-.normality of grain size distribu.tion are very old, withe..'<:tensive treatment- oi

thisconceptj}yKrumbein (1937, 1938).Specn-lation~ concerning the reasons for this were eli,-c~tssedby Krumbein(1938), out no satisfactoryexplanations were giyen. From t.hat time to the

. present various appr-oaches togranulomctricanalysis have been proposed. Thosesignific~nrto the understanding of processes are summa. irized, so that the development of the essential iideas relating sedimentary processes to textural'responSes ,can be traced.

Din;elopl1lcnts During 1940's

. \Vork eyJ>ettijohn- (1949) indicated that a. number of modes. existed in grain sizedistribu-

tions, and that deficiencies occurred in thecoarse sand~fine granule size and in the coarsesilt size. These modes and deficiencies were at-

>

mathematical study of sediment transport andfluid mechanics.

De'llelop;ments Dnring 1950's,Studies by Eiristein (1950), Einstein and

Barbarasso (1952), and Einstein and Chien(195.3) involved the relation of sediment trans-port to stream characteristics. These papers,however, dealt with predicting the volume ofsediment transport rather than' with deposition.Papers by Bagnold (1954, 1956) dealt specifi-cally with t!1e transport mechanics of sediments,and these papers provided the theoretical basisfor tneinterpretation of the textures of sedi-ments. Papers by. Chien (1956), Sundborg(1956), Vanoniand Brooks (1957), and Brooks(1958) discussed -in detail the relations ofstream mechanics and sediment tFansp~rt. Thiswork in fluid mechanics was not applied specifi-cally to textures of the deposited sediments.Shapes of grain size distribution curVeS ot sedi-ments from' both modern and ancient environ-ments were described" by Sindowski (1958).He referenced' the pioneer stl,ldies 'by poeglas(1946 }c,.but deviated from that work in that heused log-probability plots of the grain-size in- ,.formatlon.Sindowski (1958, p. 239-240) em-pirically cl~ssified size distribution curves ac-cording to Seven different depositiona:! types:(1) relict, (2) strand, (3) tidal flat, (4) shelf,(5) tidal inlet, (6) minor tidal channel, and(7) fluvial. Many example'S are provided in hispaper from mor~than 5000 analyzed samPles.Sinclowski's work,' which generally has beenoverlooked in this cOl,lntry, provides the first.c~areful study of the reiationoi .sediment texturesfrom knDwn depositional environmentS to theshapes of grain size CUFves. It allows the envi-ronmentalldentiflcation of many types of sands

"from their textures. Sindowski, however, didnot try to relate the shapes 'of the grain sizecurves'to transport and depositional processesthat formed them. This step could nQt' be madewithOlllclose study of the fluid mechanics' ofsediment transport and deposition, as, was devel-oped by Bagnold (1956) and other workers.The vi'st step in the correlation of curve shapeswith proces.ses was publisnedbyA. John }foss(1902, 1963).

~~

~~

~

~

/ De'1/elopments Du'ring the 1960'sThe two papers by Moss represent, a major.

contribution toward an understariding of the re-lation/of grain size distrIbutions to depositionalprocesses. :Mossused shape and size of grains todistinguish subpopulations produced by, thethree means of sediment transport described byInman (1949) and Bagnold (1956): (1) sur"

.:I-'

1076 GLENN S. VISHE]?,

face creep, (2) saltation, and (3) suspension.He found that these three populations could beintermixed in the same sample. He discussed atlength the transportation of clastic particles andmechanisms of entrapment of . particles at thesedimentary interface. Moss also provided in-sight into the roll of shape and size in sedimentlamination and mixing, andirito the mechanismsby which fine or coarse-grained tails are incor-porated into size distribution curves of sedi-ments deposited from a traction carpet of sal-tating sand grains. His data illustrated the sub-division of three sub-populations, and showedthat the position of truncation, sorting, an~mean size of these populations were .different indifferent samples. The most. exactingly selectedparticles are the ones transported and depositedfrom the dense traction carpet' of saltatinggrains. Breaks or truncations occur between thepopulation of particles finer or coarser. grainedthan those. found in the saltation population orfraction. The fine particles transported in sus-pension usually have an upper size range 'of

! about .07 to .1 mm, but may be coarser. This.size provides an indication of velocity otJr,e~current clear of the- bed (Moss, 1963, p. 340),,

The coatsergrained particles appear to betransported into position at the depositional in-terface by sliding or rolling. This necessitatestransport over a bed of low. grain roughness;consequently, these particles are always coarserthan those transported by saltation. The, uppersize limit of saltation depends on the nature ofthe current and on the characteristics of the bed(Moss, 1963,p. 306):

Work on truncation points illustrated by log-probability plots was presented by Fuller(1961). He suggested that the break betweensaltation and rolling' populations in many inc,stances occurred near 2 phi,

.or the point of

junction between the Impac~and Stokes laws ofpartic;1e settling (Fuller, 1961, p. 260). Spencer(1963, p. 190) suggested £ro111analysis of datapresented by Krumbein and Aberdeen (1937)that: (1 ) all c1astic sediments are mixtures ofthree or less log-normallydisttibuted popula-tions ;anci(2) sorting is ameasure of the mi~-ing of these populatlons:'Il1termixture of thesepopulations causecCilie- variation .in mean andsorting values present within the group of re-lated samples from Barataria Bay, Louisiana.

Different populations in log-probability plotswere shown by Y:i~er (1965a) in a study of flu-vial sedimentation units in Oklahoma. Thisstudy using a ,factor analysis approach sug-gested that flow regime may control the range,of grain size of the saltation and suspensionpopulations and the approximate position of

truncation betWeen the two population$.,K1ovan (1966) applied a factor analysis to the

same data studied by Spencer (1963). He foundthat the degree of mbdng of the two fundamen.

,tal populations was environmentally sensitive.The environments separated by Klovan (1966,p. 12~) primarily reflected sedimentary process'and mcluded: (1) surf energy dominant; (2)current energy dominant; and (3) gravitationalenergy dominant. This illustrated the close asso-

- ciation of process to the mixing of suspension

. and sa1t~t1on populations -in Barataria,Bay.Other lines of textural evidence for environ-

mental identification have - been pursued during-the last ten years; the most significant are thestudies by Folkand Ward (1957), Mason andFolk (1958), Harris (1959), and Friedman(1961, 1967). These authors have used the sta~tisticalmeasures of mean, standard deviation,skewness, and kurtosis to separate beach, dune,aeolian flat,. and fluvial environments. . This ap-proach has been moderately successful in mod-ern environments but less successful in inter-preting the genesis of anCient sediments. WorkbyPassega and others (1957; 1967) hasIed tothe development of C/M plots. By using a num-ber of samples it is possible to distinguish sus-pension, traction, graded 'suspension, and othersedimentary processes. Analysis of many sam-ples by use of C/M plots when combined withother methods of textural analvsis should addadditional insight into the gene;is of individualsand units.

RELATION OF' SEDIMENT 'I'RANSPORT TO

GRAIN SIZE DISTRmUTIONS

The three modes, of sediment transport, sus-pension, saltation,and surface creep, have beenstudied in some detail from a theoretical andmathematical viewpoint. Some data are avail-able on the grain size ranges attributable to in.dividual modes of transport.

Transport by Suspension

True suspension caused by turbulence wherethere is no vertical change in grain size occursin the very fine grained sand rallge, typicallyless than.1 mm (Lane, 1938). Otherstudiessug-gest a size of ,0375 mm (U. S. Waterwa)'s Ex-,periment Station, 1939). The true value mlist de-pend upon tile intensity of the turbulence andpossibly could be coarser than .1 mm. The prob-lem is complicated by the interchange ofsus-pension and bedload transport in certain grainsizes. As shown in figtlre 1, this results in agraded suspension, with coarser suspension sed-iments increasing in concentration toward the

~

99 9~

;.~_jRAIN SIZE DISTRIBUTIONS1077

~

ic;

4.1CL.

>-c:4.1:)a:4.1

~30

r2

.5

2Phi Scale'

.25 125mm. Scale

.067

FIG. l.-Mississippi River sediment samples, U;S.Waterways Experi=t Stati~:m. The: s~ong sizegradation within 2 ft of the river bed IS Illustrated.

I

II

II

!j

bed (U. S. Waterways E::cper'iment Station,.1939) .'

The size of a sediment particle that may beheld in suspension is ,dependent. upon tur?u-lence; consequently, the break or truncatIOn

,p.oint between suspension and bedload t!,ansportm,a)' be highly variable and reflect physical con-,ditions at the time of deposition. In true suspen-sion no variation in concentration. of sedimentse.."<:ists'from the surface to the depositional in-terface.. There£ore,' some' of this material is- available for deposition with coarser material atthe sediment-water interface. A gradedsuspen"sion increases in grain size -downward towardsthe bed allowirig an interchange with the bed-load. Mixing between these two populations iscommon in' a sedimentary deposit. Most sedi-mentary laminae conta.in some of the .1mm andsmaller size fraction, which is directly deposite.dfrom the suspension mode of transport. Thissize material ,vas easily recognized as a sep-arate population in log-probability plots of. someRecent sediments (Visher, 1967a, X967b).

'The term graded suspension has be~n used by .anumber of writers but with little consistency. It ISuse<:!here to indi~ate coarse materials' (>.1 m.m)which are .part of the suspension or "c1ay"

populatIOn.

998

Saltation Transport .Very little information is available on the

grain size distribution of the m?ving b:dlayer,or traction carpet. The maXllllUmSIZe tilatmoves in this layer is unknown; but from stud.ies of the U. S. Waterways Experiment Station(1939), grains from .75 to 1.0 mm have beensampled moving within 2 ft of the bot~om(fig. 1). Grains of this size must be. deposited

by an interaction between the tractIOn carpetand the graded suspension. Log-probability plotsshow that grains of this size are the coarse endof a single population, and previous studies(Visher, 1965a) indicate these to be 1'!part of a

saltation population.Figure 2 shows a striking .similarity between

the shapes of. the log-probability curves of boththe bedload samples from the Mississippi River(U. S.. Waterways E2speriment Station,. 1939)and samples from fore-set beds of a model delta(Jopling, 1966). Jopling (1966) shows a com-parison to a theoretical distribution from aheavy fluid zone using Einstein's (1950) bed-load Jransport formula. The close agreementsuggests that the Recent bedload :",mp1e, .thegrain size distribution from a deposited laml~a,and the theoretical distribution from a movrnggrain layerare all measures of the .samefunda-mental distribution. The distributions shown in

98

90

o.~

0.1

0.01

99 9~

998

~;>.

i

10

98 -

90

70

50

c:::

uqJ

CL.

30

c:U::Ia:

05

'0 I

001

.5

2'Phi Scale

.is.

.125

mm. Scale

.067

FIG. 2.-C\lrves illustrate the similarity between cal-'C\llated bedload, sediment lamina, and bedload samples.

I-'~1078 GLENN S. VISHER

figure 2 all contain two or three populations andare similar to log-probability plots of fluvial de-posits described by Visher (1965a).

There appears to be a similar size distributionin the moving bed layer or traction carpet andin the resulting depositional laminae. This pro-vides .the opportunity to reconstruct from thegrain size distribution the physical forces pro-ducing a lamina. :More study of the conditionsresponsible for the development and character-istics of the traction carpet are needed; butfro~n prelimina.ry data t~econcentratio-n. and. ve-locityof the traction carpet appears to be di-rectly interpretable. The upper flow regime pro-duces a different shapecf log-probability plotthan does lower flow regime conditions (Visher,1965a) .

Surface Creep Transport

Most grain size distributions show a coarse-grained population with a different mean anddegree of sorting than the other two popula-tions. Certain fluvial deposits,' however, do not

99.99

99.9

RELATION OF PROBABILITY PLOTTO CUMULATIVE FREQUENCY ANDTO THE FREQUENCY HISTOGRAM

99

90

I-ZI.LIUa:::I.LIQ..

~ t; 70

~~>-::;1t:-o 50..JuJcao::<u-CDI.Ol

~ ~30Q.3

::I~::IU

10

.0101

1.5

FIG. 3.-Comparisons of grain size distribution curves. The log-probabilitycurve shows multiple curve segments and truncation points. .'

s~ow. this population, and the saltation popula-tio~ Includes the coarsest matcrial in the distri-butlO~. -r:he reason for this is unknown, but prob.ably It IS re~ated to ,removal of part of thecoars~~ fract~ans al1d t~ the strong shear at thedepositional. mterface m deposits formed b\.continuous currents.

.

ANALYSIS OF GRAIN SIZE CURVES

. The co~pari~on of grain size curves and themterpretatlon of separate populations is aidedby the use of .log-probability plots. Figure 3shows three different methods of plottinggrain size ~istr~bu!io~; .One curve shows the l~QLJh~..&,-rams~z\1', with frequency percent, '111-other wlth-.-tfiec.umulative frequency percent,and the thIrd ,:v~th thesumulati:vef.requencyp.ercent. pro~~Q!!!!I:. .This last type of plot, with

~ fewslm?le assumptIOns, is believed to be mean.mgful wIth regard to depositional processes.The first two curves are difficult to read and .in-terpret, and change in slope, amount of mixing

. truncation points, and other parameters canno;

2.25 .125

100

90

80

10 l-

i60 ~

3::I.2::I

50 u

40

30 I-Zu.I~a..

20 ~~::I

8'Q::

10

4.061

04-5 P1--u.SCAl£

MM SCALE-

GRAIN SIZE DISTRIBUTIONS 1079

. be. easily observed or compared. The striking as-pect of the log-probability plot is that:. (1) itnormalIy exhibits' two or three straight line seg-ments; and (2) the "tails" of the simple "S"shaped cumuiative frequency curves appear asstraight lines, allowing for easy comparisonsand measurements. :These straight line segmentshave been observed in nearly 2000 grain sizedistributions. The consistency of the position oftruncation points, slopes, and other characteris-tics suggests that meaningful relationships arereflected by log-probability plots:.

The most importai£asp~ct in analysis oftex~tural patterns is the recognition of straight linecurve segments. In figure 3 four such segmentsoccur on the log-probability curve, each definedby at least four control points. The interpreta-tion of this distribution is that it representsfour sepa.rate log-normal populations. Each pop-ulation is tru!1cated and joined with the next

population to form a single distribution. Thismea.ns that grain size distributions do not followa simple log-normal law, but are composed ofseveral log-normal populations each with a differ-ent mean and standard deviation. These sep-arate populations are readily identifiable on the10g-probabiJityplot, buLare very difficult to pre-cisely define on the other two curves (fig. 3);

Using the .fluid mechanics concepts summa"rized in the previous section, and the work byMoss on the different populations, the followinginterpretation and assumptions hiJ,ve been devel-

op'~d. Figure 4 ii1ustrates the anaI-ysls of a Siil-gle sample from the foreshore of a beach. Asuspension population has been defined whichrepresents less than 1 percent of the sample.Note that the point where it is truncated a.ndjoined to the next coarser distribution is closeto 100 microns. This is precisely the pointwhere fluvial hjrdrologists report marked

t- .zLIJ

W-'<.)4:0::(,)w(/)

Il.>-.w t-2=::;:jt-c:o«J~::>0:::Eo:::;)0-(.)

EXAMP-LE SHOWING FOURTRUNCI\TED LOG NORMALPOPULATIONS FROM LOWERSWASH ZONE

90

50

-.

$WASH-BACKWASH-SEpARATION ,

1090

0.1

I

1

0\

4 PH I SCALE.067 MM SCALE

FIG. 4.-Relation of sediment transport dynamics to popujations and truncation pointsin a grain size distribution.

.1080 GLENN S. VISHER

changes in the size distribution of the wash loador material transported dominantly in suspen-sion (Lane, 1938).

.

The grain size distribution curve (fig. 4)shows two very well sorted saltation populationswhich differ' only slightly as to mean-size andsorting. The high .degree of sorting of these two

. populations suggests very. exactingly selectedgrains that logically wouldpe deposited from the!TIovinggrain layer or. traction carpet. of saltat-ing grains. This, particular sample is from theforeshore of a beach where, swash, and back-wash represent two differing transport condi-tions and presumably produce two separate sal"tation populations in, opposite flow directions:Such a result, emphasizes that smalI changes incurrent velocity can modify a single,' detritalpopulation., .'

The truncation, of the saltation populationsoccurs near 2 phi,or 250 microns. This -,~Eej.khas been attributed' by some wOrkersa,sJhejunction between the Stokes and lmpact Lawformulae (Fuller, 1961). 'This might be inter-pretedas the size where inertial forces causerolling or sliding of particles rather than salta- .don. The coarser straight line segment repre-sents traCtion load or surface -cr"eep.

Inasmuch as the evidence suggests that grainsize distributions'arein reality mixtures of oneor more 199-normal "populations," an 'analysisof the numbe.r, degree of mixing, size range,pe.rcentage, and degree of sorting of each popu-lation should characterize a grain size distribu-tion. If.it is assumed that each traI)sportati°I1.procesL (surface. creep, saltation, andsuspen-sion}Js reflected'in a separate log-normal popu-lation within, a single grain size distribution, theproportion' of each population should be relatedto the relative importance 01 the correspondingprocess, in the formation of the whole" distribu-tion. InJ~ddition, t!J.t:sorting, s.ize range, degreeof mixing, and the points of truncation of thesepopulations can provide i!1sight into provenance,currents, waves, and rates of deposition. ,

An analysis of more than 1500 samples has-shown that these parameters, vary in a predict:able and systematic manner, and that they havesignificance in terms of transport and deposition.

CHARACTERISTIC CURVE SHAPES FROMMODERN.ENVIRONMENTS

Sampling Pro<;edflres

Samples coUected from known depositional.environments were analyzed in light of the con-siderations outlined' above. Sampled areas in-cluded a wide range of environments, physicalconditions, and provenance areas. The sampleswere collected to prOvide factual information on

the association of spL._,c sedimentary environ-ments with certain types of grain size distribu-tions. More than 500 samples' from modern ma-rine environments' were colIected to determinewhether there was a genetic association ofcurve'shapes with environment or depositionalprocesses. At each sample locality informationwas obtained on the physical aspects of thedepositional eiwironment, including tidal infor-mation, wave conditions" provenance, and geo-morphology of the depositional site. Special em-phasis was 'placed on sam'pling many differentgeographic areas" and presumably differingprovenances and physicaiconditions. In everyinstance effort was made to colIect sanwles rep-resentafiveof conditions at the time of the sam-pling. Most, samples were from the upperi110stbedding unit, and rarely did a sample extendmore than a few centimeters below the deposi~tional interface.

Bea<;/1,and Shallow Marine Sqmples

Samples were collected along profiles aaossthe strand line from the dune to several hun-dred yards offshore at more than 30 locali-ties. from Grand Isle, Louisiana to ,Cape Hat-teras, North Carolina (fig. 5). In addition, sam-ples were taken from beaches~ along the GulfCoast, from Brownsville, Texas, to Cameron,Louisiana. The nature of sampling is illustrated

. in figure 6 for each sample locality. Three to 25samples were collected at each locality, and spe-cific information on the physical and geo-morphic conditions was' recorded (fig. 6)"

Analysis of ,the size data from these samplesshowed that there were several different funda-mental log-probability curve shapes., The sam-ples could be dassifiedinto those deposited' by:(I) beach processes; (2), aeolian processes;(3)'wave action;. and (4) breaking waves.Each of these four shows characteristic log-probability plots.

Beach

Figure 7 illustrates the characteristics offoreshore sands from 11 different beaches, froina wide range of physical and provenance cQndi-tions. Each size distribution is plotted in identi-cal fashion, and all show three or four popula-tions, with two saltation populations. Manydifferences appear" however, and the 'position oftruncation points is highly variable. At MyrtleBeach (fig. 7f\J the 2-phi break occUrs at 1.0phi, and at Pensacola Beach (fig. 7D) no 2-phibreak is present.. The break in the saltation popu-lations riinges from about IS percent to near 80percent. Also slopes, or sorting, of the various'populations are highly variable. '

,

GRAIN SIZE DISTRIBUTIONS

SAMPLE LOCATIONS

FIG. 5.-Recent sediment sample localities.

SHORE -' PROFILETIDE GOING,OUT

1£.1

-<g>-......~en<CD

~

"fEET

°1 .2 3,

PH'SCAlE

~I

<0,en

~J

~II)0a::0..

1 .';

-, 0.1:0 1

°

flHI ~CAl~

4 ~ 0.1

- --.:.

'

...

'

, ,

,''''-SEALEVEL_

.

O

..:.

-...

..

1

..

.

~

"

3 ~,"

- i PH'-':CAlE

50

I.IJ0(:

~

~CD-<CD:

1081

/AvON BEACH

:--- CAPE HATT-ERAS

ArtANTIC SOUND

PAMLICO SOU-ND

BEAUFORT INLET

ATlANTICa.

SO'RF B.

lONG 8.

KURE 8,

MYRTLE B.

90

50

10 ~<1 ~

o.~ ~;!ID<ttI

~

LF

DUNE

90

50

10

I0.1

2 3PHI SCALE

HIGH TIDE

FOREST BEACH S. C.

FIG. 6.-Shore profile and sample distribution. Grain .size curves illustrate effe(:ts of envirO!llnenton texture.

10

10.1

1 2 3PH'SCAlE

I.IJ

<~~ 50

a:a...::CD

~10

1

0.1

1 2 3PHI SCALE

LOW nDE

GLENN S. VISHER1082 GRAIN SIZE DISTRIBUTIONS 1083

The similarity of each of the~e sands is in thedevelopment._of two saltation populations. Thereason -for this is believed to be related to swashand backwash in the foreshore zone,but otherpossibiIitiesmight include mixi.ng f~om s~parateprovenances or shape of particles of dlff.erentsize. These samples are related only by their oc"currence on the foreshore, and it is improbablethat the same break would occur in all samplesunless it is related to a specific process devel-oped orUheforeshore. In the hundreds of.ana-lyzed samples this particular curve shape IS al-"v~ys associated with th~ .foreshore of ,a beach.

lV/arille Sands from Wa",'e Zone

.More variability occurs in the shapes of. thelog-probability curves of marine sands than inthose previously described. The 12 samples plot-ted on figure 9, however,. were selected frommore than 100 samples of. sediment-s from. thelower tidal flat to a water depth of 17 feet. Thebasic simil~rity of all these samples is that theyare from the wave zone, and that the deposi"tional ..interface. was wave rippled at . nearlyevery sample. locality. In each. instance threedilferent populations. are. developed, and eachsamnle contains..a variabk amount of sHr andmud~ All show a poorly sort(:d coarse population\vhich remains after the shell material is re-moved by an acid leach.

Characteristics of these curves include: (1) apoorly sorted coarse.. sliding or rolling pop~la-tion; (2) a very well sortedsa~tation populaho~with a size range from approximately 2.0 to 3.:>phi; and (3) a variable percentage of the sus,-pension population. The amount of the suspen-sion population appears to be related to theproximity. of the depositional site to a source .offine clastics. The samples with the largest pro-pOrtion of this fraction came from either theSea Island area, Georgia, where a: number ofrivers draining the coastal marshes and. plainenter. the sea or from the Mississippi Deltaarea. Other ~amples.farther from a clasticsource, for example Avon Beach' near CapeHatteras, North Carolina. (fig: 9A)and Gulf-shores, Alabama (fig: 9C), have a very smallfraction of the suspension population. The rela-tions,how~ver, are complicated by localphysi-cal conditions, such as breaker height, shorelinegeometry, and sedimentation rates.

Certain characteristics -6f marine sedimentssuggest a correlation to wave processes. The os-cillation which prod1,1ceswave ripples may cause -winnowing that produces .the excelIeIlt sortingand the narrow size range of the.saltation popu~lation. The lack of strong currents prevents theremoval of the coarse bedload population or itstransport. 0)'. saltation. The. fine suspeIlsion pop-ulatioh is related to the.amount of the mate-rial in suspension and to the amount of win-nowing at the sediment water interface.

';';-9<;9999

//1 B!~

f11/'.1

I'.~l/~i.:'!.i-l-BEAUf08f INLET

!-,I'U~RTH CAROL~NAI , # SWASHZONE

/ ,'/

.i7/$/1

~" .'1/1/-Ii I.{ .

ti~:"J'~' .'

~~;~--.--l, ,

,.I,.,

I

~--1-

~

It0:

~9~99.8

;:

;:;:

;:

fi~r~MYRfLE8EACH

~~'i.

-SOOTHCAROLINA...u..ii:"CX- -.l~ i M:~~:"!.~'JJ't!:'!'"'O""NORTH CAROLINA. .~ iOf>P£f!;fORESHORE

../ ,

'B~ '38

]

~

1':'SURf BEACHN-(lRTHCAROLINA

'UPf>ERFORESHORE,

90

i_':'TO

50 DESTIN 8EACH, flORIDAUPPER fO~ESHORE

;:',

~0...i Dune Sands

Of the more than 100 dune samples analyzed,.12 are shown in figure 8. The samples are fromdune ridge~ adjacent to ,beaches. This locationinfluence~ the general shape of size dQributionsfor these samples, making them a highly selec-tive group of wind blown sand deposits. How"ever, certain characteristics are developed thatcan be. a~sodated with wind. processes,and thesecharacteristics serve to distinguish them fromsamples- from other environments closely asso-dated with beaches.

Of special significance. is that the two p.opuJa-tions found in the saltation range. of the beachforeshore have been resorted into onepopula"

- -. tion. A single saltation population is developedin all of the dune samples, and in each case itrepresentsriearly 98 percent of thedistributioJ1:The sorlingo:f this single population, as indi"cated by the slopes of the curves,is excellent,and. generally better than for bc=51chsamples.Also the truncation ofJhe coarse tractionpopu-lation o~curs between 1.0 and 2.0 phi in itIlsam-pIes. The percentage of the traction pop1.1lationwas found to be. very small, never more than 2percent. The presence of a suspensionpopula~tion and the truncation of the coarse populationaccount for the positive skewness characteristicof dune dep6sits. All these characteristic$serveto differentiate . dune sands f!'Om all other mod-'ern sands the writer has analyzed.

The importance of saltation in wind transportof sediment has long. been emphasized, and thedominance of this population in the samples an-alyzed suggests a genetic relationship; The gen-eral lack of competence. of wind. processes to

iIi-arille Sands from Zone 9fmove. a coarse populatIOn by surface creep .

Breaking Waves -(Bagnold, 194-1), accounts for the small per-. ..'. .centage of material in the coars~ population. Fi- ~

.. The final type of grain size distnbutlO? fro~nally, the addition of 1 to 2 percent in th~ sus- the near shore zone is a product of depos~tton mpension population above that present in the the surf zone. Twelve samples from this zonebeach foreshore samples suggests. that unidirec- are iIlnstrated in figure 10. The samples aretional winds are like fluvial transport, and that characterized by relati,'ely high percentages ofthe suspension materials are incorporated into material in the coarse slidin.g and r?lIir;g popu-the sediment at the depositional interface. lation.. The percentage of thIS matenalls depen-

30 ..

§

~io::ICI'" "-

r..-::O.~

0 I'.

001 -¥.

PhI Scale,25 125

mm. ScaJe ~

Phi Scale.2~ '125

mm'. Scale

06.7.067

, 9~=:S;,,JII'D

,j'.

".J,,I'I,

. ,II

I':--"'AYPORT II\I1.ET1 FLoRIOA-,

OPEN. SWASH ZONEI. .

I'.,I

/i1./,1

I .,6',

II,.;...'

;,,'i'II'" .

9999

~9~-.

~a

=- ~

cj)

J ::»

~

9')90en

-

110 ;'~

". i1

50 so:

g0...

j '0.:-.

c~

Q'"i~

iJ-<;

{)1

QOI :::0.,~

Phi Scale25

'12~

mm.SC..le

PhI Scalf:!S /

nun ~h..dl~

,,10;.' Ubi

- b . I'"

sa~ds.Sands formed ~ differing- wave and provenance conditions areFIG. 7.~Exa!ilples of e,lC1

lore>i~~~ded.,A.Il curves have a similar shape.

9999

99-8

98

90 ~,

10

J50

<,)

30 :::I..

10 :I-f7

I~,1084 GLENN S. VISHER

J ~

0.5

0.1

0,0"1

1/)

50

30

10

05

0 I

,001

9999.,_/8

'I~{,.

".I'1

-'II

II.,

II

IJ,

I,I

J.II,

1II

I,II

998

98

90

10

50

30

10

0.5

e;::

0..

fj

I"i

I;

/;i

!fj,

!i;

ii

Il.

ei. !..!J.-.ii,.~~'--j

;""'~: 'r., .0-1

(1.01

~

2~Phi Scale

.25 .125mm. Scale

.067 .5 -Phi Scale

.25 .125rom. Scale

.067

~ ~~~

"2:

..,' l.:;/', ~l c

",.,.

.'/ !~;,~, ;::I I ...".

/~If

I iI i, ,

I tI, iH

AMlll';"SlE ,-iftORIOA.

"i

I i, Il!

I iI i

CA,P-ESAH8I.A$ ' iFI.ORIOA I j

11-, ~

.Iift'f

I'

;:

/1/

/~~:..-lr-

"Ii"..F ~~..;;

99.99

99.8

~a:

.,I:

fl

IfIiI.Ij

I:If

IiIf;

MEXICO8EACH-r-f----HORID", , !I i

:'1OESTINeEACI1~i"hORIO", : I

I i'A

iI i

f'AHAMAC[n~/iflORIOA I _J

I .;J j, i

I ;I i

f f- IfS

CI...

~~

2-.Phi Scale.25 .125

mm.Scale067

98

90

so;;;

a... 30

::Ia 10

0-\

vOl

fiG. 8.-Examples from beach dune ridges. The similarity of curve shapes is evident but variation in meanand maximum grain sizes and amounts of individual populations are iIIu'strated.

Phi eScate25

.lZS

mm. Scale

0.67

GRAIN SIZE DISTRIBUTIONS

~

~

~Q:

1085

99.99

98

".'"I,.I,

AVON8EACH~JNORTH CAROliNA,'

115.0'WUfl".d"pttll . I

I.,I

I;.,,I~

0..SUltf 8EACHNOlttH CAROLIHA11.2.0'w.I",d"dlhlIIh""UH~dl

-S'AVANNAH RIVER INLET

GE'ORGIAlOWER TIDAL FLAT

!W."tnD"I"cll

999\1

::

~

Phi Scale.25 125

mm-. Scale.061

8

90.

~

&

~

-

j ](!

..

j

0..

cr

PhI Scal~~~ 12';

rom Scale(;67

50 50

)0. i JC

10::I

0..5

0. 1'~

1

0..01 vOl

Ptll Scale.25 125

m.m..Scale

0.67

9999 9999

998

."~!./;-~

""1- I1!~ 11- -I

~~~~~\lSlE~l' ITlDAliiAT ! IIWneroDPk;dii I

f (~ I

I 1

PONTEVERDA.fLO_RIOA~118,O' lt'd"DUiI J I

It,I

hf/~

LIMIT Of>50'\ .,,/!

SHEll DE8In~---,. ::'1. ~. t ...---- .

!ji

.-.....~-!f'

".."".r- - c:

998

,98 .98

~o ~ 90.

-;0 ..:::)

~

70

~o toC;

JO 1

0.0.1

3G

10.

(;

~

-~ORESleE"C'"~ SOUIH C~ROUhJ.

,111.0' I".d'0Illl

10

C5 C:~

0:

(;0.:

FIG. 9.-::'vVave zone ~al!d distributions. Similarity in the gener~1 form of these distributions can be seen. ThevanatlOn In water depth does not appear to aFiect the grain size curve shape.

PIliScale2~ In

mm Scale06::

99.99

998

98

90

~10

1so

30 Q..

c:

10:;)a

\--

1086 GLENN S. VISHER"

]

~~.

]c4,0~c:r

.1 .5Phi Scale

.25 125mm. Scale

~

~

pAHANA CITV, fLORIOA--'---I1.S'wiltcr.depthl

.(Plvn.~ !onel

a..1,,1c:;Ja

.5Phi Scale

.25 .125

mm. Scale

99-99

99.8

98

90

.'"

50

]0

10

0.5

0-:1

0.01

061

0.5

D.t

0-01

;067

~

99.99

B ,. ..J!';-..~f:'- "'~: ~...,I

..~~NYRTLEBEACH-ISOUTH CAROliNA .'12.0"w.ter depthl ,,PlullleIOne} It,.

11',,"",,{ .

i'-fOLlY BEACH,. i SOUrHCAROLINA

',- i.." !2.S'wntt depthI

" 1'''jPlun,elone'

...II'

.'

"

99.8

91J

110

50

0-.

7/

~..'~ fORESr.8EACH

l'SOUTHCAROLIN"

"r'U.5",utudepth)

~ ('lunCtlon".

i~10

30

C<U

='CT<U

0.5

0.1

,] 0.01 ~

Ph-Scale25 125

mm.- Scale

061

99.99

99.8

98

PENSACOLA BEACH, FlORtOA

CJ.S'wiltero;lepthl .

IPIv".elo".' 90 ~

10

50

tI~;.". ir-DESTINBEAC1t,flOIUOA

,."., ".."1 :~~~::lte:.~~:t::Uh' lonel

.."}.,/ '-

.,',./.~.:

,l,,1,,/ '

30

:0

]",

~"001

.5. Ph, Scale

25 125mm. Scale

061

FIG. lO.-Examples of sands from the surf zone. All samples show a high concentration of the coarser popu-lation. Comparison with beach sands (fig. 7) indicates the coarser population is more poorly sOrted.

--~-~---~ ~~--~ ~~---~--_.._-

GRAIN SIZE DISTRIBUTIONS 1087

dent upon the source area and wave conditions,but ranges from several percent to -~ percent ofthe grain size distribution. Thesaltationpopula-tion is added to this coarse material, but mixingoccurs between the two populations. The fineend of the saltation population is truncated, andlittle suspension population is present.

These characteristics appear to be consistentwith the prOcesses of waves interacting With.strong currents. Breaking waves keep the depo-sitional interface agitated, and stlspension mate-rial is winnowed out and tr<\lJsported seawardby currents. The traction carpet is of an inter-mittent nature, depending on the p~sition of thebreaker and on the direction and magnitude ofthe currents. These combine to allow mixing be-tween the saltation and the sliding or rollingpopulations.

Mississippi Delta Samples

Nature of sa1'rfptes.-Agroup oJ 30 samples was'taken from a series of short 2-3 ft cores col-lected by Jam~., Coleman and Sherwood Gagli-ano Of the Coastal Studies Institute of Loui-siana State University. These cores were takenin the area of the South.east Pass of the Missis-sippi River for the study of sedimentary struc-tures. The sampling and textural' analysis wereperformed by the writer to determine the rela-

,tionship of specific structures and environmentsto the shapes of log-probability curves.

Em/ironments sampled andn(lture of size distri-. butions.- The size distributions for this group

of samples appear to be fundamentally different£rom those described, for the beach and near-shore environments, but some, ,similarities exist.Three examples from each of four environ-ments are illustrated in figure 11. ,These in-clude: (1) strand-line deposits (fig.lIA); J2)distributary mouth bar (fig. lIB); (3) natural

-

levee (fig. lIC); and (4) channel deposits (fig.lID).

Distributary mouth bar samples are similar tothe shallow marine 'sands previol)sly described,but they contain ~n appreciable amount of thesnspension population' (fig: II B). These samplesare from shatIo\\' water. (less than 6 ft), butwave energy is not sufficient to remove the sus-pension population. The amount of this frac-tion is likely to be related to the high load ofsuspension material in transport in the watersof the Mississippi River.

The deposits from the natural levee (fig.lIC) along the distributary are different fromthose developed in other environments. Thecun'es show a single population, with a sizerange and sorting characteristic. of suspensiontransported dt'tritus. X atl1ral levee deposits are

formed by the rapid fallout of suspension mate-rial along the flanks of the distributary. This isa product of rapid change in current velo<;ityat the channel margin, and may account for thesingle population. SiJ.11ilar distributions wen~recognized by Klovan (1966) in the area ofrapid sediment fallout in the Barataria Bayarea, Louisiana,.

Sti'and linedeposits (fig, lIA) show the dualsaltation populations characteristic of foreshorebeach deposits. The major difference lies in thepresence of a .considerable silt and mud suspen-sion population. This is to be expected close tothe mouth of the Mississippi River with its highsuspension load, and also because of the mini~mum wave energy developed along strand linesassociated with the delta.

The distributary, channel deposits show' twomajor populations, one related to suspensionand the other to saltation (fig. liD) .The salta-tion population hasa size range o-f nearly 2 phi,and it is truncated at 3.0 phi in one sample andnear 3.75 phi in the other. The percentage ofthe suspension population ranges from 50 to 80percent. These two samples possibly reflect dif-fering currentyelocities.

Altamaha River Est1f.ary Samples

Nature of samples.-A series of 86 samples ofthe bottom sediment was collected from the Al--tamaha. River Estuary. The distribution of sam-ple stations. is shown in figure 12. At most lo-calities current, wave, tidal,depth, salinity, andturbidity measurements were taken at the timeof sampling. The environrnents sampled in-cluded: (1) marine tidal delta; (2) shoal areaseaward of inlet; (3) wave-current zone; (4)tidal inlet ; (5) low energy tidal channel andtidal riyer; and (6) major tidal channel abovethe zone of the salt wedge. In addditiotl to theseenvironments two sample stations (nos. 10 andS5) were occupied over a 14-hour tidalcyc!e,with sampling carried out at. 2-hour. intervals.These 12-hour sample stations. provided a mea"sure, of the response to high and low velocityconditions. In most insta'nces the sample .wasfrom the upper few centimet-ers of the bed: con-sequently it represented the physical conditiol1.sfor a short period of time prior to sampling. Inareas of strong velocity changes over short pe-riods of time, the sample represented an a\'er-age of the prior transport conditions. This is acharacteristic aspect of each environment, andthe individual samples develop characteristicsize distributions.

.

Characteristic size distributiOl1s.-Selected sizedistributions shown in figures 13 and 1-1-illus-trate different shaped loe--probabilitv Ctlrve,.

99 9~

998

98

,~

-

~or

soC;

~c '%>

1(;

~.:

(/ s

0-1

~001

067

99-99

C

998

STRAND-LINE

]-

j.~

":

::0cr

Phi Scale.25 125

mm. Scale

iL~L

d! t: Lj

NAruRAl LEVEE

.J'Ii'

. 1/

IIi

I'}'SUBAQUEOUSl"EVEE_. .

?~~ISusp:efts.onj

.,' }~

"If

. .I.. /f\ .CREST Of lEVEE . #. . ~-- C'JiEVASS OEPOSIT'R'l)pllltc,ou-~<ld:edl . , / Is..sptftS>Oftl

'SUSlltnS'o"<1ndcuttenll.".,..~-l'.'

Il~~~I'

.~~, .~~

,',/"t 8.' ~~

_.- . -- "'''

OISTRI8UTARY- MOUTH BARB

U"J.\.flJ..LV "';J./ ~L/l.J.L.J.\..J.LJL/:.t .J.V.!.V..) .1.VUJ

~-

"l-II-

98

]

~~ c:

,~-. 41 - !1,""'11" l 91)

,1'- ,".'/. I

SHAllOW ".l:RIN"[--:--I 1IW<1vllt rOMI I til

.

, ,l 1

".t

SHAllOW MARINE. ~~

IWullt lonel

"f~:

l l.1 I, j

J !

l, .1'.

~~,..f//l

.

...' ~,.,..~ - ~~.~

.5Phi Sea-Ie

25 125mm. Sc-ale

.067

9999

998U)

~

<

-'-'<ZQ

2j::

.!: ~ -II~ t) ~\0 cz:::r

-'<t- a... Q

~ ~ ~ Ji

rrPh-' Scale

2:' 12:'mm Seal!:

10

2.

05

01

O&J

067.

DISTRIBUTARY CHANN.EL D

98

~0(

~.XLUQZ

. ""......", "'.......

I.."./&

,~..;,'

!

-"A;OR eHAHN[L., ~1ISlroftccutreftn JI

.I.if

SHAllOW WATEfI--r---AOiAC[NifO f

NATURAll[VEEt~iWu.kCUtrtftl} l

~lJ

l/i-

Ii

I2-

. ~

Ph, Scale25 125

mm Scale

067

/~ .

J') P. 2. ~>~- .';f; .

-- $-0

Ew~-

,~::;:;;.

'J~6;--. t;

>.j:2

~~E.s

:;;::

(;~

g.:;:::::"'..~.""-'. N::.

~ ,<>

~ .0 (' .:~./..-c--'

,i~;

-r"'.',.<.".".-J

\;'...,"'",.....

.'-~_." ,...!::Q-..~~.>~:::. ,..,-f"';'":.;:;-.

""'.

- ..~"'"~:~'.,,(':.'- ...) ,:;~ " ...

...{~.~(~-..

a-e.......

.,

'''.''.'''---:.'f:

;..~.

"

_01t,.: ,/

/~(.It::"

£~OOt

99-99

99.8

~.

98

~o

'::IC.

30

10

0-5.

0 I

0.0 ~

FIG. ll.-Sands from associated :Mississippi RiVer Delta environments. The shape of the grain size curvesstrongly reflect environmental conditions. Transport processes are indicated for. ea~ illustrated curve.

-

0 !I

0 I

061

061

99-99

C

998

98

GLENN S. VIS HER.1090

The offshore marine sands in the area of thetidal delta at depths of 10 to 40 ft show thecharacteristic. shape described for the shallow'marine samples associated with beach deposits.The three examples illustrated in figure 13A arefrom different depths and positions within thlOtidal delta. They all show the well sorted salt~-tion' pop\1lationdeveloped within a very narrow

. size range. The break between saltation andsuspension populations was in the very fine saudrange, usually near 3.5 phi. The break betweenthe bedload population and the saltation popula-tion was also fine, generally near 2.5 phi. Thesecharacteristic.s are thought to be. typical of de-position by oscillation waves. The variation inshape oUhe curves appears to be related to po-sition on the delta and to - proximity to the

source of clastic detritus. :C1ase ~to the channdthe traction population is more abundant, andthere is less of the .suspension population. Thisappears to be related to stronger currents andshoaling action of the waves.

The ilioal.areas shoreward of the tidal delta(fig. 13B) reflect the action of breaking waves:They contain a well developed saltation popula~tion, truncated at the fine erid. Also a large bed-load population joins the saltation. popl1lationbenveen 2.0 awl 2.25 phi. These characteristicsare similar to those samples from the beachplunge zone (fig. 10), but the percentage Ofbedload is much greater. The three samples il-lustrated are from shallow water areas close tobreakers marginal to the channel which extends' -across the tidal delta. This environment is simi-lat physically to the plunge zOne adjacent tobeaches, and a,c similar log-prohabilitycurveshape is reflected.

The zone of interaction of waves and tidalcurrents (fig. 13C) produces a different shapeddistribution curve. Each of. the three curves il-lustrated cQntains three popUlations (fig. 13C).The saltation population is truncated Or! the fineend and has i restriCted size range. The coarseend is truncated between 2.5imd 3.5 phi, whichis relatively fine;yhen ~ompared. to Qther' types'of size. distributions. The coarse truJ;lcation'point is from J.O phi to nearly 2.0 phi. The sal-tation population is poorly sorted andha~ abroa( size range. Its size range and sorting is

.,unique 'when compared to any other distribu-tiort. ,The third population, truncated on thecoarse end, shows good sorting and extendsover a wide size range. ,

The mechanism for the. formation of this sizedistribution is unknown, but the line saltat;ionpopulation suggests \vinnowingby wave action.The poorly sorted intermediate population sug-gests dumping. from a highly turbulent gradedsusiiension-traction carpet, and the coarse popu;

lation s~ggests bedlWd transport by a strongcurrent. These conditions would be the result ofthe interaction of a strong bottom current withsurface w-aves within a tidal channel. Support-ing this interpretation, -the sample localitieswhere. these distributions

.were developed. were

cat thi margin between the tidal channel and theshoal area. . '

.

Samples from the tidal inlet. (fig. 13D) werecharacterized by three., well developed andmoderatelywelI sOfted populations. The' suspen-sian population comprised from 2 to 5 percentoUhe' distrioution alld ranged from near 2.0 phito 4.0 .phi. The saltation population' was welIsorted and occurred over a very narrow sIzerange from 1.5 phi 1'02.0-2.5 phi. Theb.edloador surface creep poptilation was also well sortedand repres~nted from 30 to 70 percent of thedi,stributioIL These two populations join withlittle mixing at about 1.5 phi,and, are truncatedon the coarse end riear -1.0 phi.

.

Strong turbulent.currents generated by dis-charge into the ocean combine to' produce a sus-pension population, a ~oarse truncation. pointbetween saltation andsuspe!lsion, and a largebedload frac.tion. The sorting of thebed]oadpopulation was directly related to the position in

.the channel inlet and the velocity of the bottomcurrent. The sample' from' station number 10(fig. 13D) shows only one. population, possiblyindicating that the coarse bedload population is..transported by sal~ation when current velodtyis high.. This is supported by data from the 12-hour sample station. at the inlet mouth~ which.indicates that during low~flow .coriditions threedist1n<;t populations are developed, and' -thatafter periods of. high flow the grain sizedistri-bution approaches

-a single population (fig.

14A).,Grain size distributions upstream' from th~

salt wedge (fig. 14B) are similar to those found.in the tnlet area. TheY generally contain les~ ofthesuspensionpopulMion, typic-ally from 0 tQless -than 2 percent. The 12-hour tidalstaHonexhibits the same relationsbipbetweenpoptila-tion discrimination and periods of maximum

'flow velocity (fig. I4e). The percentages ofthe other pop\1lations aresimiIar, and the .sort-ing and points of truncation between popula-tions are nearly identical. This suggests that thetidal action and currents are important in pro-ducing these log-probability curve shapes; andthat physical processes associated with the inlet,or trre...salt wedge, are not as important

The final group oidistributions is from 'areasin the Altamaha Rh-er where current velocity islower because of a reduced tidal range or a lo-cation in one of the less important tidal chan-ne]s (fig. 14D). Two types 'may be recogni:zed:

'..,,--'GRAIN SIZE DISTRIBUTIONS 1091

A

;;;<I)

~

4,10

a..

~.tr

Phi Sc.ale25 125

mm. ~ate

'"

j

'"Q.

::.<::T -10

'2Ph.Scale

25 I2S.

mm Scale

067

99'99

0:1

99,9~

99:8998

90 ~

IOFfSHORE BAR - .21 --::-tt32.0'w'l4ro-tpUll .I075kMtlbotzom("'i.nl~'

f,,.;j,"/'

II"~,- ,

".,

,~,.;t

.,...' .,.', 10

98

90

10 iso

~]0

CII>::J

o.~

O.Ot

Phi Scale25 - 125

~m. S~aJe

061

_99.99

,D

10

~

i

~#

.',TIOA1IH1ET-.tO~ "n6.0" ,o.pthl -

,(.4~A1Itll ~-,,

I'i','t-T.DA,l nWl£T-.t-4-, . IIS.G'wu o.pUlI,. 1.5uoUbotto=cu",,,UI .

I.I

I

90

998

98

(IOAll/t1.Er -117112 O'

l4t d,pCtIl{-, -,

I #,,')f

,-', !,;

. .',If'

!lOso

'j>-

a

30]0

10

oSos

0 I0 I

o'QI001

2Ph, Scale

;2S IZS

mm. Sca.!e

06]

FIG. 13,-Examples. of curves from inlet. and marme deIta.areas.. Sample numbers indicate. p~sition on in-dex .map (fig. ]2). The. variationss~o\:n Illustrate effects ot varYIng processes. \V ateI'

depth, and current

velocities itthe time or sampling are Indicated.

9999

A

998

98

90

70

50

]0

10

I1092

. .

~

INUT MOUTH - 12 HOUR TIDAL STATION ,l

l..i'l,i

AfTER hlAXIMOMFi..OW--

h~:'$

121.0' Wit" dtplllJ. .$lrnpMIf82- -~.,~

}--MARLOW TIDE

i 123.0'wlterclfPlhl

"Sampl.tt71

/-.--~

a:

..:l~~~/

./l"

,..i/4

~

,~ BOTTOM VELOCITY

(I

&..00

':MlOw

fide

Flood Hiih

Tide

!tooPM

Ebb lowHde

.5Ph,Scaie:

.25 125.mm. Scale

~-flEARlOWTlOE

112.0'w.lerdepl'"$11111'1'''58

~c:

c;

a....

eOTTOM-VEU)cITY{;T<I.>

(I

2:00

PMHigh

Ebb"Tide

2;00

AM

Flood He,,,

Tide

Phi Scale.25 125

m~.Scale

GLENN S. VIS HER

RANGE OF VELOCHY

FLUCTUATIONS

.067

c

aAHGE OFV(tOCITYfLUCTUATli:ms

~

B

067

, ~l"

,l ~

l' -'.'

,l ,;

-SOUTHCItANNEl- -52 .- - ~! ~ ,"'.0' ",.." ~

..

~~ -ISUftdll'C WIV"' .l I.-

-SOUTHCHM~NEL_-"4'~ 1240-"'&8,,,.pllll

,/:,1.f.35 kl101Sbomim cutrentl

if./

"'~/-SOUTHCH"'NHE~_-.49~.t:;"~1 - :.~4~~~:~b'::::~tUnU

,/ ...,.

"~ .,1'

",""~'l>- .~.'.~ If~'"

, ",;

!~~,

--D00:

j

0.5

0.1

001

Ph, Scale.25 125'

rnm. Scale.061

99.99

'39.8

l ,f /, ,

CClHEROfCHANNEI.-/ l,.

1.(".'-!-'!

l- 1MAlttCHANN"tL~ IU2.o'walerdeplh) j-:

! ,I ,ii,

1'#'~"',

'1j'~~.,f '

." "a.. . i~'~,".

~ ,,~i',.'~

~..ll.'

JO

AlTAMAflA RIVER0

98

90

<I>

(I)

.Q0

a: I~

i

.//

l~MAINCH.ANNEl,i. . "Nurwl!erlotvel,

if

A ~,

",

~"f;-I

,..." 05

50 ~()

30

10

0 I

001

Ph' Scale.l5125

mm. Scale067

FIG. 14.-S~mples ~rom t~e ~ltamaha River Estuary and River. Also, selected samplestrom two 12 hourtIdal statIOns. TIde mformation is plotted to ilIustrate effects' on grain size curve. shape.

''''''''-

GRAI:V SIZE DISTRIBUTIONS 1093

9999

(1) qistributions similar in sha,pe to those de-scribed for the main channel; and (2) distribu-tions with three well defined populations. Thislatter type is characterized by a highly variablepercentage of the suspension. population (fromless thin 1 to more than 10 percent), and bytruncation with the saltation population between2.5 and 3.5 phi. The saltation population of thistype extends over a range from 1.25 to 2.0 phi,is more poorly sorted, and the truncation withthe traction population, if present, occurs be-tween 1.0 and 1.5 phi. The amoim.t of the sur-face creep fraction ranges from 2 to 2S percent.The difference between these two types of dis-tributions appears to be related to positionwithin .the channel, with the second type foundin shallower water of a lower'eunent velocity..

The difference between these two types ofdistributions suggests that current velocity isthe controlling factor both for the position ofthe break between saltation and suspension andfor the. slope of the saltation population. Themaintaining of a bedload or surface creeppopu:-lation appears to be related. to the tidal actionrather than to current conditions. This popula-tion is thought to be concentrated in the estuaryby the alternating direction of the bottom cur-rent and may be an important textural criterionindicating tidal action.

9911

98

GO

}'J

~()

30

10

0..5

0.1

0.01

9999

.1-

IA

90

l-IL-L

. ;#' 98

"j,/. ;.f..,,'I-

}"'

.,',

' 'I ..,..

I,-,..,'"

h/O~",,,~...' .

, .l..A7

~~.RkA"SAS R."ER C~L.."O"'A

} .Cu"en!"DDle~:II

l :.

,Ii'''' I.1I";! /

,..,~' .1i~'. /I.

I'Ii

f-

- ~

'.'" t. AR"ANSAS,RIIIER,OlllAHOM.4 '--..-=

,POtfllT 8001 fCu,,~'nt I.Im'nal~"~

"I

4 r:lt ~:"

J;: r; r~

~~

U.ZOSI:UYUI, 'E,'XASPOlfll' B.I1}Cu"e"II~"'.'na!t,d!

Ph. Scale:l: I;tL

mm'Scale

:61

Modern Fluvial SamplesFluvial sample?, illustrated in figure 15, show

a distinctive pattern. They are characterized by:(1) _a well developed suspension population

. comprising up to 20 percent of the distribution;(2) the truncation between suspension and sal-tation occurring between . 2.75 and 3.5 phi; (3)the size ranging from 1.75 to 2.5 phi in the sal-.tation population; and (4) the saltation popula-tion having a slope or sorting intermediate be-tween deposits' formed by waves or reversingcurrents and those formed by suspension. The.slope of the saltation population is in the 60 to65 degree range, as compared to the high 60's or70's for wave deposited distributions or the 50'sfor suspension deposits. The bedload or surfacecreep population, if present, would be coarser

. than 1.0 phi. This is. strongly provenance con-trolled, and is developed most frequently in the-.deepest portion, of the channel. Because of va-riations in channel patterns and the' size ofmaterials in transport, an inclusive statementconcerning the shape of fluvial grain size distri-butions cannot be made.

The characteristics described above are par-tia.lly developed in some oJ the samples de-scribed from the ,Altamaha River Estuary (fig.14D) and the Mississippi River channel sands(fig. lID). A gradation between deltaic and flu-

9999 <:9;;;

B99.8 ....

I~ ,

. j .4"."'

J

";..;:

."

,

Ge

,

!

i.

//'

i~a:RAZOS RIVER TE::t.1,S

I ,rul~o..ao$$:'beo"edt

if

.il!j-

i

«i

"'-

~

ARkAfIISAS RIVEII. OkLAHOt.lA---

tFUloonCr(lss.bedjkdl

50

,:;,.

: ~z~

~~f.

.....

,'.

,. ("L

t ~ ....

.'~

,

FIG. 15.~Selected examples of modern channel sands. These curves illustrate variation in truncation pointsand slopes of individual populations.

Ph,Scale2~ ;2:-

mm. Scaie061

05-

0.1

0.01

067

99-9"

C99.8

98

r1094 GLENN S. VIS HER

vial. grain size curve shapes is indicated. thatmight make it possible to place individual sam-ples within the fluvial, upper deltaic plain, orlower deltaic plain environmental regimes.

RELATION OF MODERK DISTRIBUTIONS

TO ANCIENT SEDIMENTS

Possible Differences in Ctlrve Shapes

The primary purpose of the study of modernsediments' was to obtain information fromknown environments to aid in classifying distri-butions from ancient se'diments. Nearly 1000distributions from ancient rocks have been ob-tained, and a number of specific patterns can' berecognized. Specific shaped curves also werecorrelated with environments determined fromother pIlysical and paleontologic criteria, buttextural data from modern sediments providedthe basis for environmental comparisons.

The major difference observed between an-cient ~nd modern grain size distributions is. inthe amount of fines less than 44 microns thatoccurs in the ancient samples. The reason forthi's is probably multiple:. (I)' related to diage-netic addition of clays, (2) post-depositionalmixing, (3) sediment settling downwardthrough the pores, and (4) possible transport bymoving interstitial fluids. Each of these pro-cesses is described in the literature, but little in-formation has been published evaluating the rel-ative importance of each process. When grain

. size distributiorts of ancient sediments are inter-preted" the possibility of these processes modi-fying.the curve shape must be recognized.

Other changes might be related to solutionand pre<;ipitation of fine clastic particles and toenlargement of grains by precipitation of mate-rials on larger grains. In th~ size range from Imm to 44 microns such processes probably arenot quantitatively important for most sedimen-tary rocks, but in deeply buried or strongly de-formed clastics the possible effects cannot be ig-

,nored.

The mechanical. disaggregation of' consoli-dated rocks alters the grain size distributions tosome degree, but can be minimized if care istaken. Still, li.ttle hope can be held out for ob-taining' the original size distribution of mate-rials in the clay or fine silt size range. Problemsof flocculation, dispersion, crushing, recrystaIli-zation, and cleayage appear to be insurmOtmta-ble, Consequently, size analyses have not beencarried' finer than 44 microns.

Similarities in . Curve Shapes

The consistency of curve shapes from sampleto sample produced by similar processes andthat between ancient and modern analogues arenoteworthy. A comparison of curve shapes be-

tween ancient and modern examples showsthese similarities. Comparisons show the appli-cability of log-probability curves in the in-terpretation of depositional processes and envi-ronments.

Variations in the slopes or sorting ofindivid-ual populations, positions of truncation points,and amounts Of various populations are devel-oped. This is to be expected since similar varia~tionsoccur within modem environments, butthe general Cttrve shapes provide sufficient char-acter to recognize specific processes and envi-ronments.

.

Curve Shapes With No Modem AnalogueA number of curve shapes from ancient rocks

do not have a close analogue in the'modern.sed-iments included. in this study. All modern envi-ronments have not been thoroughly sampled,and these curve' shapes may be found whenmore' eJ{tensive sampling can be accomplished.In a number of instances the environmental in-formation from ancient sediments is sufficient todr;iw conclusions as to the origin of a particularsize distribution. In these instances interpreta-tions are suggested.

PATTERNS FROM ANCIENT ROCKS

Fluvial DepositsLog-probability curve shapes developed in an-

cient fluvial sands are similar to those describedfor modern environments. Similar saltation andsuspension populations are developed with atruncation point between 2.75 to 3.5 phi. The sus-pension population ranges from 2 to 30'percent,with the sliding or rolling poptilation generallyabsent. These' characteristics are shown in e:'C-amples of fluvial sands selected' from more than300 samples (fig. 16).

Samples from Missourian Series of Oklahoma.-Characteristic size curves are shown in figure16A illustrating the major types ()f fJttvial sandsclassified by a factor analysis' study of more.than 200 samples (Visher, 1965a). These sam-ples range from the base of a channel to the up-permost ripple cross-bedded unit. Sample varia-

. biliti is small compared to the rangeo! Curveshapes described from' modern enrironm.ents.but some variability exist~ in the position of thesaltation~suspension truncation p()mt and in the'slope of the saltation population. These mea-sures.probably are related to sedimentary struc-tures and to the position within the ~hannel.Grain size distributions of sands with smallscale cross-beds, found in the upper part of thechannel, showed a finer truncation point thanthe current"laminated or festoon cross-beddedunits; also, the saltation population was more

'--'GRAIN SIZE DISTRIBUTIONS

FLuviAL SANDS hlISSO~RIAN A.GE',OKLAHOMAA

~

1C

Q..

c:;

::Ia

ph, Scale25 .125

mOm, Scale

j ~

.,...-~.l '"

-i ~- /~DEGONIASANDSTOHE-:'-. .;p{."1"IlUNOIS8ASIN

-t

"'J - 'MISSISSIPPIAN AGE

" - l-NE'A,SEC.27.1IS,tW

J JiI I

I ~~

~~~:I~I:~A~~ONE7J /~~~A£RSTTl~;V~~~~;~NDSTON[30

PENNSYLV,l.NIANAG£I l-'OKLAtiO'U., ~,,

"/4 ifJ /

Ii1/

I II ~

.I'I~ J

I lI. I

I

-.0~00:

0....

::I

99-99

998

10

70

50

10

05

Gi

001

1095

'19 9~

B

99.8

<.I/J

. t'i la:

.-- .,",,,,,.,",,"'°",,,--,Is..bSU'.'Clcoul- .'I,

I,,."",~',

I(,,-,,

I, .

!>o

I::)

c

-",

:0

1,5

:~,I

r,Ol

.5PhI Scale

.25- 12Smm. Scale

067

~9.99

0

998

~8

~

1-'0

0:'0

g

0....

::0

c::

aiO

(0 ~

(:J

:,1.;1

Phi Scale2~ 11'::

mm Scale

\'

arl'et,

'

ot'

an, cien , t fluvial sandstones. These sa-ndstones show \"ery similarFIG. 16.-Examples from a wide - shaped grain size cur\"es.

Pht Scale2~

~}n

- mm.,Scale"f,7

067

5.'n

-.0)1} Q

0Q..

tlO

",0 !c41

10

.1.V;1V lrL1:!.lV 1V .). V 1Sl-J.!:5H

poorly sorted. These factors suggest that cur- (fig. 13), indil.-.ng .that physical conditionsrent velocity and depth control the saltatian-sus- may have been similar and that low current Ve-pension truncation point as~well as the slope of lacity and high suspended laadalso character-the saltation papulatian. ized the Bluejacket-Bartlesville delta. In addi-

The sliding or rolling papulation was. nat tion, the high concentration and poor sorting .ofcammonly developed in these size distributions. the surface creeppopulatian in'these Pennsylva-The .only samples that shawed thi~ papulatian nian sandstanes suggest dumping .of the caarse\vere ripple cross-beds at the very tap .of the flu- fractian, possibly as a result .of a large tidal Ivial sequence (nat illustrated). The absence

. of ra:nge similar to that of the Altahama Estuary. :a tractian papulation appears ta be characteris- itic of many fluvial sand depo~ts. Cretaceous deltaic sands.-Samples. from the'

, . Almand and Lance Formations show a differentMid-Continimt Pennsylvanian channel sands.- shaped lag-probability curve shape (figs. l7BSandstanes from. Mississippian and. Pennsylva- and C).. These sands have been described byn!an channel depcsits-~vefe-:analyzedJ and r€:pn:--~\Veimer (i965) -ana are jnterpn~tedby- him assentative examples fram the. Arkoma basin, the being .of deltaic .origin. The curves show a smallIIlinais Dasin, and the Oklahama shelf are iIIus- maderately well sorted surface creep populationtrated in figures 16B' and l6C. The sandstanes (figs. l7B-C). The saltationpopulationtanges ifrom the Illinois basin (fig. l6B) are described from abaut 2.5 to 3.5 phi with moderate sorting!by Patter (1963), who provides detailed .dec (slope from 60 to 68 degrees). Some .of these I

. scriptians of the channel sequences and geaine- curves are similar ta the fluvial <;urves, excepttries, The sands from the Oklasoma shelf (figs. that they contain a surface creep papulation.16B' and C) are described by Saitta (l968) and These curves are interbedded with fluvial type

~a,

e

,

rt,_,

ie,

r

_,,

~:II~h,

~

"

e~,

.

,

:~~v,

.~,

'a

,

~,

l,

1

,_

1,

i

,

;k,

oo~ath

,

;a,

sB

,

i,_

~u~~a

,

m,_

c~7:~1~,,

~v

,

n~~

-

'~~,

d;,,

~_

~t~~_

.

~

,

',

~

,

11:_s~

,

~~;~

,

.e;

,

h

_"

~,

n

,_,,

l~,,

:_P

-

~,

~,

~:~:-

~=

I(fig: l6B) are from a .large channel in the ity 'curves without a madern analague have aAtoka Formation near..Ozark, Arkansas, de- well s.orted saltation populatian (usually with ascribed by Hendricks (1950).. . . slope above 70.degrees) and a poarly satted

All examples shaw the same characteristic suspensian population (fig. l7c.:;). The paint ofsaltath;m and suspension populatians, Differ- junctian of the twa populatians usually .occursences between these sands and madern examples between 2.0 and 275 phi.

~

are slight, supporting a fluvial .origin. These curves are thoughtc ta be produced by)trang tidal currents/in an area where the sur-face creeppapulatian has been removed"possi-bly in shallow water or an bars in the tidalchannel. The clase assaciatian .of three types .ofcurves-(1) fluvial type, (2) fluvial 'with a sur-face creep papulatian, and (3f truncated andwinnawed saltatian populatian with a large sus-pension papulation~suggests a genetic associa-tian and possibly teflects a distributary with.only a small tidal range. This wQuld accaunt farthe absence of a large surface creep papulationand far the clase assaciatian with fluvial type'distributians.

Cr~taceous fluviql sand~R(}ck Springs Area,Wyomi1tg.~The.sands illustrated (fig.16D) arefram channels infhe Almond and Lance For- .

.matians. The channel .origin .of these sands isbase<i an the work by Weimer (1965). l'neshapes .of. the lag-probability plots are nearlyidentical ta' thase described fram' other fluvialdeposits, thus supporting Weimer's interpreta-tian, .

Deltaic Distributary Sands.

Bluejacket-Bartlesville.::-Sandstones from thePennsylvanian Bluejacket-Bart1esville delta .ofthe Oklahama shelf (fig. 17A) are described bySaitta (1968) and are similar . ta thosedevel-aped in the north-ch<tnnel .of the AltamahaRiver Estuary. (fig.f14D).Three populatiansare present, with .thep'oorly sarted surfa,cecreep'papulatian ranging framl5 ta 35 percent.of the distribution. The moderately-well sartedsaltatian populatian ranges from about 2.0 ta 3.0phi. The Suspen$ion population is poorly sorted,with nearly 10 percent .of the distribution lessthan 44 micrans in size.

The same characteristics are shawn by the.

sands fram the modern Altamaha tidal channel

.-',-,.'.

- ---,-~"'-

Ataka. deltaic SCInds.-The log probability plotsare of .outcrop samples callected during outcrop

Jstudies by the writer (Visher, 1965b). The end-

Ironmental interpretatian is based. .on sedimen-tar

,

y st

,

r

,

uctu

_

res,

a,

n

,

d

,

: v,

ert,

iC

,

a

_

1 se,

q

,_uen

,

c

.

es,'

S,

im

, ',

ya

"

r

,I

shaped curves are developed as w!=re describedfor the Cretaceous sands of the Rack Springs jarea (fig.. 17D) ,and a similar origin is sug- Igested. . I

1

Log probability plots-deltaic sauds.-These ex' iamples illustrate the types .of distributions de- !velaped. Additianal textural studies from mad- 1

"]

tc':

lf

-4>

~

~Ii:

J

~.:...

GRAIN SIZE DISTRIBUTIONS

COR£SANP(ES -BLUEJACKET "":SARTlESVlllE SANDSTONE. A

-.,I'>.. .,..',:; .

."'7 ".~9Q,

"'- .""" '.

8[RRYHlll161'8'~I'. ~............

I'e.JACK'l.90S.-J-.I."

/ .II '

I I.

"i

;.-1

"~."",1'(...4

~~..';i!.~$4e,...!

'1:;'f

IilU

2Phi Scale,

;,.2':. Inmm Scale

'.)6/

1097

..q 99So999

UPPER CRETACE°!JS SANDS-. ROCK SeRINGS UPLifTC

2Ph. Scale

25 125rrim Scale

067

998

JH

91:

ALMOND fORMATION - NORTH Of SUPERIOR, WYOMING8

;98#

98

]::

5(,;

c.... 30

~ WHlTf SA~DSTONE

SASE Of SECTION

~ 10

0505

1)1 ;)1

001 OO,!

Phi Scale~., rL~

mm 'Scale067

9999,

999,9

AJOKA FORMAflON, ARKOMAaASIN. ARKANSAS0

.998

.-i"',

I.,

I Mh( S MOIJIHAlNaURG~1us ROUHlI' I

l'.:.;

:::1

0-5 ..' ~

01".

001 ~,o:.i

5-

2Ph,Scale

25 12~mm Scale

067

. .FIG. 17.-Deltaic.distributary sandstane curve. shapes. Thesee..'..c~mplesshaw a wide variation in cun'e shape

and poSSlbly reflect strangly cantrastmg delta types and positions within the~delta cample..x.

Ij

1- !1m GLENN S. VISHER

CC

fI

i-'

AALMONDJ"{)f!MATION-

u.s. ROUTE 30

fAST OF ROCK SPRINGS WYOMING

~-~..,

~0

.:i:

~....i:;:;

~

2Phl,Stale

.25 .125min, Scate

:0;67

~9.99

PENNSYLVANIAN SANDSTONES - NORTHEASTERN. OKLAHOMA,C 99.8

~-

~'0

0:

;;

0...

~

~

0.1

0.012

Phi Scalen :~s

mm S.cale061

9999

99-?

98

--;..n

i.:i:

- ~cL

itS

01

001

l:.-,

~£ -;;

::1..

i

~

UPPER CRETACEOUS SANDSTONES"-ROCK SPRINGS UPLIfT, WYOMING 8

-

AlMONO fORNATlON-"';'-STATE ftOUTE 430

'

~~

AUION-D fORMAtiON.NORTH-or SUPERIOR41'-CARBONATE

t.ANCE fORMATION

NORTH or SUPERfOR

Phi Scale.2"5

..125

-mm.-Scale067

ATOKA FORMATlON,ARKOMA BASIN, ARKANSAS

D

~~

'"l-

'''''

~

~."", ;r ,-..

.P'. . "". -;1/";-'

~l."-,.;l l,i' :, 1--"11

.5PhJ$cale

.25 .125mm: Scale

(167

FIG. l8.-Sand~tones from probable marineenvironments. Each sample Occurred within a section withdemo?strable mart!lecharacteristics. .AII examj:)les are suggestive of Mississippi Delta area samples (fig. 11)«r of ~lta,,!aha River Inlet and manne delta samples (fig, 13). These samples may be related to marine por-tions ot anqent deltas rather than to nearshore environmerits associated' with beaches.

'vl\.hli'V")lL.;~ L/1Jl1\.lLJU.L l.VJ.VJ

9999

ern and ancient deltas)r!e needed before therange of the variation can be determined. Moreprecise in formation concerning the processes re-sponsible for the formation o~ indivi~ual curve

-

shapes is needed beforespectfic environmentalinterpretations are possible. Sufficient informa-tion: however, is available for the identificationof deltaic type curve shapes:

Shallow 1na~inesallds

Burrowed and wave-rippled sandstones werecollected from many different iock units, includ-ing the Cretaceous, Almond and Lance Forma-tions, Pennsylvanian Sands from northeasternOklahoma, and. the Pennsylvanian .A.taka For-mation from the Arkoma basin (fig. 18) :Threedistinctive characteristics are' common to thesesands :( 1) the bedload population when present15 poorly sorted, and truncatidngenerally isfiner than 2.0 phi; (2) the size range of the saletationpopulation is from 1.0 to 1.5 phi; and (3)the suspension population is well sorted andusually truncated at a size finer than 3.5, phi.This population typically ranges from 5 percentto as mUj,;h as 80 percent of the distribution.

These sands differ from tidal-channeldistri-<hutary sand~in the degree of sortirig of thebedloa.d population and the position of trunca-tion of the saltation population. They differfrom fluvial sands in the sorting of the suspen~sian population and the position-of truncationof the saltation population. These curves aresimilar to those described for modern environ-ments, and this characteristic curve shape ap-pears to reflect wave pwcesses.

No systematic study of -shallo,,, marine sandshas been 'carried' out;. consequtmtly, other

.shaped distributions are possible. Relict sands,thoseprodticed by transgressions, or shdf sandsall are a product of wave processes, but theirgrain size, distributions might show differentcharacteristics. The shallow marine sands' de-scribed are probably .similar- only to -the nearshore ,sands developed adjacent to beaches anddeltas included in this study. A more compre-hensive sampling of mode'rn manne-shelf envi"ronments is needed to detetmine the range {)fpossible log-probability curVe shapes.

<}'98

I

Ii!

~ I

05

.01

0.0:

99.99

'3J.~

j

j

iI

!- ,98

so

Ie

0.5"

0.1 Beach Deposits

Modern bea.chdeposits have a particularlycharacteristic curve shape, and their identifica-tion should, be posssible in ancient rocks. Exam-ination ,of all log-probability plots available ofancierit sands indicate onlv a fewthat have<the

: characteristic two saltati6n populations.. Curve

1'

-

'

,

'

,

_,

shapes

...

0

'

f P

,

OSslh,

I

,

e beaChd

,

e

,

posits"

are

,

ill

_

U

,

st

,

'

,

r

,

a

,

'

,

te

,

d(fig. 19).

. One ,deposit from the Cretaceous Castlegate

0:01

J..V7:7

Formation from the Book Cliffs, Utah, wasidentified as a beach by Spieker (1949). Thedistribution curve for this sample contains t~osaltation populations, and is similar to modernbeach curves (fig. 7). The sample contains 5percent suspension population, 'but this may re-flect 50UrCe area or diagenetic effects.

A distriQution curve of the Ordovician Ber-gel1'Sandstone; equivalent to the St. PeterSandston'e, from northeastern Oklahoma isnearly ideriticaliri shape to mod~rn beach distri-butions. The St. Petet; is interprei:ed}VideIyasbeing of a beach origin, butlittieevidence hasbeen presented. The Bergen outcrop sampleddoes' --riot contain sedimentary structures cr- ~vertical sequence, which would indicate a beach\!.eposit, hut other interpretations are eqljallyambiguous:

.

A v~rtical profile from a marine sequence ofPennsylvanian age near' Tulsa, Oklahoma showsan upward, progression from a shale to a sand-stone-shale interbedded urtit, followed, by asandstone containing brachiopods. This unit iscapped by a para;lfel bedded sand unit. Log-probabIlity curves of sands from this unit con-hiin two saltation populations. The position inthe sequence is that of a strand-line deposit, butinsufficient evidence exists to call it a beachfrom other environmental criteria. The- sh~pe ofthe log-probability curve, however, is distinctiveenough to suggest this possibility (fig. 19). '

These -are' the only examples of pus'sible, beaches found in this study pf ancient sands.This suggests either that beaches are rarely pre-served as' ancient. sands,.or that post-deposi-tional processes have altered the distrihutions sothat beac1tes cannot be identified.

~

This, how-ever, is unlikely since characteristic cUrveshapes, as' indicated by previous comparisons,are commonly preserved.

Turbidity' Cur'rent DepositsOne of the, most ch-aracteristic shaped size

distribution curves is from tUrhiditycrn:rent de-posits .(fig. 20). No modern analogue is avail-able for, these sands, but' the environmentalcriteria are i;vell developed and allow for easyrecognition. k large range occurs in the shapes

, "of the, Jog-probability plots of density or turbi-dity current deppsits, possibly because thesecurrents are highly variable in velocity, density,grain size oftt;ansportedmaterials,and thick-

- ness.The distinguishing characteristic of these de-

posits is the development of a large, poorlysorted, suspension population, which includesgrain sizes from clay and silt to 1 mm. Evencoarser materials may be transport~d in the sus-pension mode by some currents, but precise data

llOO GLENN S. VISHER

~~

. are limited. The truncation point of the suspenesion population can be a.s coarse as 1.5 to 0.0phi, and a coarser population may be present.This population is better sorted and may repre-sent the saltation population; but the physicalcharacteristics of particle transport in deI;lse sUSepensions is unknown, and whether there is salta-tion or surface creep transport has not been as-certained.

Ventttra Basin graded bed.-Four grain size ~distributions are plotted from a' single' Pliocene. .;:turbidity current bed from the Ventura basin

~(fig, 20A), The upward decrease in grain siie,..the change of the saltation-supsension trunca-

tion point toward finer sizes upward in the bed,and the increase in percentage a f the suspensionpopul;ition toward the top of the bed are notablefeatures. Environmental information concerningVentura basin turbidity currents and informa-tion concerning' this bed are discussed by Cro-weIl and. others (1966). The interpretation ofthe log-probabilitycurves supports the conceptthat the graded suspension mode of transport ispredominant, and that a dense suspension cantransport coarse detritus.

Other e.'~amples .of turbidity current deposits.-Figure 20 iIlustrates examples from the Penn-sylvanian Atoka Formation (fig. 20B), theHudson River submarine fan (Kuenen; 1964),and the Delaware basin (fig. 20C), and miscel-laneous eXq,mples from several areas and ages(Dzulynski and Walton, 1965) (fig. 20D). Theturbidity current o-rigin of the Atoka samples isdiscussed by Briggs and Cline (1967)~ Thesecurves are similar in shape and verticalprogression to those from the Ventura basin.The curves from grain size analyses reported byDzulynski and Walton are, more varia.ble (fig.20D), but do show the poor sorting of both thesuspension and.the saltation population (slopeusua.Ily less than 50 degrees). The position ofthe truncation point between the saltation andsuspension populations appears to range widelyand probably is dependent upon the density ofthe turbidity current.

The Delaware ba.sin curve shapesdand ~hosereported by Kuei1en (1964) are different fromthe other examples (fig. 20C), and certainlymust reflect a different type of depositing cur-rent. Turbidite sedimentary structures are lack-ing in inost sandstones from the Bell CanyonFormation, stlggesting that they may have beendeposited by a different process than those de-scribed above. The distributionc!lrvesof sandsfrom the Hudson submarine fan (Kuenen.1964) are similar, suggesting a submarine fanorigin for these Delaware basinsands. Environ-

~

~

~_J.$!!.!;!:A.!! SJ.~!:S!::-;:!

THOMPSON CANYON

800~ CUffS. UTAH

.5Phi -Scale

.25 .125mm. Scale

-.061.

FIG. 19.-=Exarpples of sandstone with grain sizecurve shapes similar to modern beach sands. A beachorigin for these sandstones 'is inferred from theircurve shape.

mental information on the origin of the BellCanyon Formation sands has been presented byHull ( 1957), but little detailed work is pubelished. The, term fluxoturbidites (Stanley, 1963;Dill, 1964) has' been proposed for deposits- pro-duced by sand transport down .thecontinentalslope and across subsea fans. A welt sorted saletation population and. the mixing of a gradedsuspension population at the point of truncationare characteristic of these curves. This wouldfit the processes of sand transport described formodern examples (Dill, 1964).

Significanc[ of turbidity current curve shapes.-The curve shape of the log-probability plotsof turbidity current deposits provides a new'ar"proach to theinterpretatibn of their transportand deposition. The vertical gradations in size,the truncation point between suspension and sal-tation populations, the slope of the suspensionpopulation, and the amouI;lt of detritus less th;in44 microns aU suggest that a turbidity currentbed is a single genetic unit. The variationsfound iI;l the bed suggest that a similar vertical\'ariation mav occur in theturbiditv current.The prepond~rance of poorly sorted ~uspensiondetritus in turbidite units of the classical type is

§'i'S5

GRAIN SIZE DISTRIBUTIONS'---'

lIOI'

PliO-PICD fORMATION - VENTURA eASIN, CALIfORNIA A[Stop.4, 1967AI. P.G.MII.los A"ltlnl

;

10

I f1! eI ;

'I ~ ..-;'1

J[ .~.;''''

j ~ .

I ~I ...

SlftGlE SECTION fROM ONE

s.oroor TURBIOtTE8ED

50

30

10

05

0:1

~

0.01

Phi Scale25 125

mm.$dle061

!jii

I j

c

!Q..~c.

"~

HUOSON-SUBSEA f_AN./.:>

LAMONT OBSERVATORY ./ICofedepthO.4clII.1 ,;:."'1'/.

/~.i~-

/$'

BfLlCAlHONFORMATION 0.5

.5"

Phi Scale.25 .1"25

mm. Scale067

9999 9999

qq$j f!r-I-

"'JOkAFORMATION~~G~lg~~~~~T~f

~~OG(H, OI(L"'HOM~ 8qq

~

10

~

1~

1

..

'::

] 70

'/~50

-,.,~ , ,., ""- - ,~~ 30

'.~' ..',"'..""," ..: ..~.,

BASE OfUI'UT-,- -.--,~~..

,/ ;'f"~

'"

1..5'A80V[-. ~~~..".. BASE

..-",-"~...i'-3.5'A80V(BASE

,..., ,

/.."'"

,.

50Q;

30!l.

100"

10

05 05

0 I 01

001 001 ~

PhI Scale

2':. 12';;

mm Scale

067

9999 99-99

D99:-8 998-

98

.~ 90

t~'

...".,_..,..~~-;'J-:""

OROOVICIAN--:..~-

,.1"SCOTlA:; ""~".'

",--,'

"",""-,'-PEIUoIIAN JAPAN

,,,... l.d" -

,

/""- .HY' ,..r

."..."" ,,/I

"0- ~

! 30

c::

~. H.i

O}.fA fRGN OJUl'1'1'1$111ANO WAt ION

0 I 01

0.01 001,0

'-1,Phi Scale

.25 12~

m-m. Scale

.5 067

FIG. iO.'-:"Examples of turbidity current deposited .s~dstone~. Il!sets A, B, and ~ iIlustratethe variab~ityfound in turbidity current deposits. Inset C shows similar gramsl:ze curve shapes trom modern and anCIentsal1ds..A subsea fan origin is suggested for the Bell Canyon Formation.

1102 GLENN S. VISHER

'1999, . I

r

AMODIFIED SUSPENSION DEPOSITS

]

ia:

,~ ..,-"~-,.......BlUEJACj(£T-8ARHESVILLE~..' 'BEtlRYHlll1114' .,,:.

III""."'";' ..l '"

"...,

;'"...~~::f

..#' ,.;0'

f..,' ..".'7 i{

Iif';>

/ l,"', ~

I li, 'II ,r !I . l, !

0

Q,;

c~

~

.5

2Phi Scale

.25 .125

mm. Scale.067

REWORKED SANDS

998B

98 .

90

7-0

~-

, . .,.

;""'"ALfdONDfOR"'AII0r"~,-«ORfHOFSUPERIOR

i' ,."WYOMINC*::;1 ",

.,\;;.:,;'1

.--A~~:,"j".':":»~ .; ~ /

ATOI(AfORMATION-~'SOUTHOF ./'~il(~~~~~aURG ,..

./"..

I

V)

~~

50;

30

~~

0.5

0.1

001

2Phi Scale

.25 125. ,- mm. Scale

FIG. 21.-Examples of ancient sandstones not observed in modem sands. The 'ppssible mode of origin isindieated from curve shape and the relation to other samples and data. .

supporting evidence for the concept of a denseturbid c1oudmoving rapidly down a slope, Theturbidite unit from the. sole. marks tq the upperlaminated zone, therefore, would all be a prod-uct of the same flow. Alternative modes of de-position would be reflected by abrupt changes inthe shapes of the 10g-proba,biIity plots andshould be easily recognized.

Miscellaneous CUr'"ueShapes, The study of any grouP of log-probability dis-tribution curves from an ancient sand providesmany unanswered questions concerning a num-ber of curve shapes. Many variations and .un-usual curve shapes are present, and their expla"nation necessitates a re-study. of all the physicaland biologic aspects of the sedimenta,ry se-quence. Some of the anomalous curve shapesmay be due to composite'samples (those repre-senting more thana single depositional unit),but others may represent unusual processes andhave real significance in interpretation of thegenesis of the, sand body. These curve shapesusualIy can be related to the vertical sequenceof sedimentary units or to their position withinthe environmental framework, and their signifi"cance can thus be properly evaluated. $ome ofthe interpretations suggested for specific typesof grain size distribution curves have been de-veloped in this manner; but their true signifi-

.5 .067

cance must await detailed analyses of samplesfrom known modern environments.

The most important group of grain- size dis.tribution curve shapes that have not been de-scribed an; those developed by deposition fromsuspension (fig. 21A)c ,Many. density currentand slump deposit~ occur at relatively shallowwater depths, These curves usually 'shmv apoorly developed saltation population,' or strongmixing between surface creep and suspensiontransport populations. These characteristics also

. can ~e developed by post-depositional reworkingproduced by burrowing organisms or secondar~'processes (fig. 21B), and care must be takennot to confuse these curve shapes with onesproduced by primary depositional, processes.These curve shapes are inCluded to show thehazards of attempting to interpret every distri"bution found within a sand unit.

DYNA:xnCRESPONSE OF cURvE SHAPES TOENVIRONMENTAL CHARACTERISTICS

The analysis of grain size distribution cun'eshapes from both modern and ancient environ-ments . has provided information concerningcharacteristics of the log-probability cun'e:,.Most of these characteristics were suggested bythe association of a specific property of an em'i-ronment to a unique characteristic of one ormore of the stJbordinate j>opulations of the size

. GRAIN SIZE DISTRIBUTIONS' "

-. '.~ Jistribution,. The characteristics indicate certain!eneral hypotheses concerning cause and effect~eiationships between sedimentary processes and'textural responses, These relationships are out-iined below, but with the precautionary note[hat they are only empirical and are not based\Ipon quantitative hydraulic studies.. They areoutlined here to provide a basis for the morequantitative work that is needed to support the~eneral thesis that log~proba'bility curves do re-flect sedimentary processes.

Characteristics Reflected by theSUsPei-;,sion .r.C>opi,f,lation

'~

".

C 5.

The suspension population reflects thecondi-!ions above the depositional interface. A Closeassociation exists between a large suspensionpopulation, a high concentration of suspendedsediment in the fluid, ane! rapid sedimentationrates. Relations concerning sorting of t)1ispopu-lation and, mixing with the saltation popuhitionare ambiguous, but appear to reflect turbulencein the overlying fluid and the presence of aboundary. layer. Strong currents produce aboundary layer and 'restrict both. the amountand the sorting of the suspension population in-'eluded in theclistribution. Strong mixing be"tweeri the suspension and saltation populationappears to be related to highly variable energyconditions which result in the partial destruc-tion of the boundary layer.

Characteristics Reflected by theSaltation Population

The saltation population is a product of themoving grain layer. The forces active in thetransport of sediments within this zone arepoorly' understood. The range of grain sizes,sorting, and points of truncatio.n of the popula-tion are highly variable, but they do s1.iggestcertain interpretations,'

Samples with good sorting of. the saltationpopulation appear to reflect reworking or win"hawing by wave, tide, or swash and backwtlsh.The higher the velocity of the opposing cur-rents and the slower the rate of sedimentation,the better is the sorting and therefore the~teeper is the slope of this part of the distribu-tion curve. When opposing c,urrents each formseparate laminae, two distinct saltation popula-tions may be developed as described for beachforeshor~ deposits. The position of the finetruncation point may reflect turbulent energy atthe depositional interface. High turbulent en-ergy would produce tnmcation at a coursePOine,anc!JQ~v:tu!,~ulenE energy at a 'finer trun-.~ion point. The coarse truncation point wouldreflect- the-. shear at the depositional interface,I';ith high shear produced b)' high bed layer ve~

01

001

1103

locities. The amount of the saltation populationdepends upon the stability of the moving bealayer and the rate 0 f deposition.

Characteristics Reflected by Surface. Creep Population

The amount of the surface creep populationis largely provenance controlled. A large perccentage of this .population necessitates the re-moval of finer graiiJ sizes. This can' occur bythe selective removal of finer materials by win-nowing. The slope of this population reflects thecompetenc," of the transporting currents. Themaximum size may indicare a provenance con"trol; or a limit related to current velocity. Manydistributions are truncated at the coarse end,which suggests there is a, mechanism limitingthe coarsest size material in transport.

CLASSIFICATION OF GRAIN SIZE CURVES

The characteristics of the individual grainsize distribution curves provide a oasis for anenvironmental classIfication.' Any attempt to de-fine' precise limits for the slopes, truncationpoints, and percentages of each of the- threebasic populations for individual environmentsprobably is impossible.. Certain guidelines; how-ever, may be based on the samples that wereavailable for this study. Because of variations inprovenance, post-depositional processes, and im~proper sampling, any single grain size distribu-tion curve may not. fit into a unique category.Also improper Classification, is possible if theguidelines are taken too rigidly. With these limeitations as a guide, a proposed Classification ispresented in table 1. Only a few sedimentaryenvironments are included, but others may beadded as 'more information is obtained.

CONCLUSIONS.The determination of the depositional envi-

ronments of an ancient sand is a difficult prob-lem, arid ,in most. instances physical, biological,and chemical criteria are needed before a firminterpretation is possible. The textural criteria

. outlined in this paper should properly be only~nother set of criteria to be used in conjunctionwith many others. Together with other informa-tion such as sedimentary structures, position insequence, fauna,. and mineralogy, the texturalinformation may provide new insight or possi-bly the confirming data needed for environmen"tal interpretation.

The emphasis .of this paper has been in de\-el-oping the background material fOr anew ap"proach to textural analysis. Sufficient datahave been presented to indicate that this ap-proach has possibilities. Rigid application of theprQPosed classification, or specific genetic in-

-Saltation .population ~uspension pop~lation Surface creep population!

lV1ix~C.T. F.T.

IPercent

Mixing F.T;" C.T.Sorting Phi.

,Phi. Sorting A&B Phi. Percent Sorting Phi. A&CSand type Percent

i:.{ttl;--1.5 '. 2.15 2 Poor Little >4',5 Varies Poor N(fLimitFluvial 65 Fa~r3.50 3598 ,~I.O

-Fait 2.0 2.0- 60- Poor Much >4.5 0- None-Natural levee 0-.\.0 3.5 100 5-30

-20 ,Good . 1.5-' 1.5- 0 Poor Much 35. 0- Fair- -0.5 AverageTidal channel2.0 3.5 20 Good >4.5. 10. Good -1.580

Good 1.25 2.0- 2 Fair A~rage 3.5 30', Fair- -'0.5 Ave~geTidal inlet I' 30-4.0 10 Good No Limit65, 1.15 2.5 5 Good .to

2 .Populations .5 3.0 0 Fair- Little 3,5 0- Fair -1.0- Ave~ieBeach 50->4.5 50 No--Limit99 Excellent 2.0 4.25 10 G90d

Good 1.5 3.0 0- Good Much 3.0- 10 Fair- No Limit AveragePlunge zone 20-'-, 2.5 4.25 2 >4.5 90 Poor.90

Good 2.00 3.5- 0-' Poor- Little 3.5- 5 Fair- 0.0-, 'MuchShoal area 30-2.15 >4.5 2 Fair >4.5 10 Poor -2.0'95

Good~ 2.00 3.0 5- Fair Much 3.15 '0- Poor. 0.'0- Littlet Wave zone 35->4'.5 10 Poor-

' \(. >4.5 10 No Limit90 Excellent 3.00

1.0- 3:0 I Fair ~verage 4.0- 0-' Poor \.0- -LittleDune 91- Excellent>4.5 Z 0.099 2.0 4.0 ,, 3

'>

Fair 1.0 0.0-, 30 Poot Much >4.5 0- Fair. No-Limit MuchTurbidity 0-2..> 3.5 100 40 Poorcurrent 70 Poor

.'-

1'::-:1104 GLENN S. VISHER

~"~-

TABLE 1.-Key: C. T. = Coarse Truncation point; F. T. = Fine Truncation point;

4.=Saltation population; B = Suspension popufation; C = Surface creep population.

terpreta~ions,probablYiS unwarranted at thistime, but the approach has been successfully ap-plied to a number of study are.~' ..' .. The analysis'of log-probability gram ~Ize dis-tribution curves appears to bl: a fruitful methodfor studying sedimentary dy~ami~s. If morl:

'teJCtural data were presented m this manner abasis would exist for comparing textures of clas-tic rocks. More infonnation is believed to be ob-tainable from this type of plot than for anyother method of presenting the data, .and for"this reason alone such curves should be mcluded,

as a part of the petrographic description of, Clastic rocks. '

ACKNOWLEDGEMENTS

Support during the early phases of this studywas furnished by Sinclair Research, Inc. Laterthe National Science ,Foundation, under Grant

Nos. GA-997 and'GA-1635, and The University.

of Tulsa provided direct aid. Without the finan-cial assistance of these organizations this studywould not have' been possible. In additiQn, aportion of the publicatiQn costs have been pro-vided by The. University of, Tulsa and the- Na-tional Science Foundation: Ideas expressed inthis' paper' were developed, during discussio~swith' colleagues at Sinclair Research and tnclasses at The-'University of Tulsa. Drs. JamesDavis, Phillips Petroleum Company" and Roder-ick Tillman, Sinclair Research" Inc., offeredsuggestions for revision of the manuscriptwhich have been helpful indari£ying. the ideasexpressed. The careful editing" o~ the m~nu-script by Dr. Roy Graves, Information Services.University of Tulsa, is gratefully acknowledged.Responsibility for all errors and interpretations.however, must rest with the writer.

REFERENCESR A 1941 Ph sics of blown sand and d~ert dUPes. Methuen and~o., London, ?65 p: . -.BAGN~~954, Experiinenfs on a gravjty~free dispersion of large spheres In a Newto~!an fluid under stres>r

, Phil. Trans. Roy. Soc. Londo!!, v. 225, p'.4tJ-?3.. .' s

"

d,

249 ' '23- -29i.19-6 The flow of coheslOnless grams m flUids: PhIL.Trans. Roy. oc. Lon on, v,. ,p: :J- .B~' G

:J ,IT ANDCLI ~~ L M 1967 Paleocurrents

.

and source ar.

eas.

of late Pa.

le.

ozolc S.

edlment.s otRIGGS ARRE, ..~, . . .," J Sed

' P t I 37 985 1000th~ Ouachita Mountains, southeastern Oklahoma: our. lentary e ro ogy, v'. .'p.

'. - .BN ' H 1958 Mechanics of streams with movable beds of fine sand: Am. ClVll Engmeers Tra,ns.,ROOKS, o_.t ,- -,

v. 123, p. 526-594."

"'. . ' A S C

. .1E ." TransC' N N 1956 The present stratus of research on sediment transport: m. oc. IVI ngmeers'."

HIp;er'Z824 '.833-844. ""

'.,', , .

'".CRowdL, J. c.: ~-\NDOTHERS,1966, Deep water s.edlm~ntary structures-Pliocene Plio PICOFo~atlOn Sant:1

Paula Creek, Ventura basin, Californ~a:, Calif; Dly:MIn.es GeoL, SP~c. Rept. 89,.40 p.D, The.is.DILL, R F" 1964, Contemporarysubmanne erosion ,m Scnpps submanne Canyon. Unpub. Ph..

-Univ. of Calif. San Diego, 299 p. .

-/GRAIN SIZE DISTRIBUTIONS. 1105[JOEGLAS,D. T., 1946, Interpretation of the results of mechanical analyses: Jour. Sedimentary Petrology.

v. 16, p. 19-40."','DtULYSKI, STANISLAW, AND WALT.°N, E. !:C.,.1965, Developments in Sedimentology, 7, Sedimentary features

or flysch and greywacke. Elsevier Puhhshmg Co., New York, 274 p.1::r~STEIN H; A., 1950, The bed-load function for sediment transportation in open channel flows: U. S.

: Dept.'Agriculture, Tech. Bull. 1026,71 p. ,"-, AND BARBAROSSA,N. L, '1952, River channel roughness: Am.' Soc. Civil -Engineers Trans., v. 117,p.121l-1l46.

" - . -. '

"-..'~..c--, AND CHIEN, N,; 19J3, Transport of sedunent mixtures with large ranges of gram size: Untv. Calif.

Inst.Eng. Research,M. R.D. Sed.Series,No.2, 49p, ,- ,

.FOLK,R. L., AND WARD, W. c., 1957, Brazos River bar: a study in the significance of grain-size parameters:

Jour. Sedimentary Petrology, v. 27, p:3-26.'

"F~EDMAN,G: M.:' 1961"Distinction between dune, beach" and river sands from the textural characteristics:~'. Jonr. Sedimentary Petrology, v. 31, p. 514-529. ,- 1967, Dynamic processes and statistical parameters compared for size frequency distributions of

be~ch and river sancl,s: J oUF. Sedimentary Petrology, v. 37, p, 327-354.FULi:.Ert,-'A. 0., 1961~ Siz-e characteristics of shallow r£tarine san,ds- from Cape oiGood Hope, South Africa:Jour, Sedimentary Petrology, v. 31, p. 256-61.

GILBERT,G. K.; 1914" The transportation of debris by running water: U. S. Geol. Survey Prof. Paper 86.

ffA~~~, Pi;.. A.,J959, The mechanical composition of some intertidal sands : Jour. Sedimentary Petrology,v,29, P. 412-424. ", ,

H£."i"DRICKS,T. A., 1950, Geology of Ft. Smith District, Arkansas: U. S. Geol. Survey Prof. Paper 221E,p.67-94.

" '

.

HULL,]. P. D., JR., 1957, Petrogenesis of PennianDelaware Mountain sandstone Texas, New Mexico:Am. Assoc. Petroleum Geologists, v. 41,p. 278-307.

-I:OIAN, D. L, 1949, Sorting of sediment in 1ighto£ fluvial mechanics: Jour. Sedimentary Petrology, v. 19,p. 51-70.' '

".'10PLlNG,A.V., 1?66, Some pri~ciples and techniques used in,reconstructing the IiydauIicparameters of a

paleo-flowreglme: Jour. Sedimentary Petrology, v. 36, p: 5-49.KALINSKI,A. A., .1943, Turbulence and the transport otsand and silt by wind: Ann. N. Y: Acad. Sci.,v. 44;

Art. I, p. 41-54. ,. .KLOVAl1,J. E., 1966, The use of factor analysis in determining depositional environments from grain-size

distributions: Jour. Sedimentary PetrologyJ v.36, p. .l1J-..125.KRUMBEIN,W. C,1937, Sediments and exponential curves: Jour, GeoL, v. 45~ p.

577-601.~,'1938, Size frequency distributions 'and ):he normal phi curye: Jour. Sedimentary Petrology', v. 8, p.84-90.~ AND ABERDEEN, E. J., 1937, The sediments of Baratana Bay (La.): Jour. Sedimentary Petrology

"v. '-T, p. 3-17."

,

"'.

, ,'KUEN'EN,PH. H:, 1964, TIeep-seasands and ancient turbidites, p.3-J3 in Bouma, A. H., and. Brower, A.,

eds., Developments.in SedimentOk)gy, 3, .Turbidites. Elsevier Pub. Co., New York, 264 p.1...\NEE. W.,'1938, Notes on thefoxmation of sand: Am. Geophys. Union Trans., v. 19, p.

50S-50S.\lASON,C C, AND FOLK, R. L, 1958, Different!ation of beach, dune, and aeoliap flat environments by sizeanalysis, Mustang Island, Texas: Jour. Sedimentary Petrology, v. 28, p. 211~226.

'!foss, A. J.. 1962, The physical nature of common sandy and pebbly deposits. Part I: Am. Jour. Sci. v. 260,

p. 337-373.,'~, 1963, The- physical nature of common sandy and pebbly deposits. Part II: Am. Jour. Sci. v. ~6I,

p. 297-343. ,.,

.' '?;\SSEGA,R., 1957, Texture as characteristic of clastic depositiQn: Aril. Assoc. Petroleum Geologists

Bull"v. 41, p. 1952-1984., , -., ."

~, RrZZINI, A., AND BORGHETTI,G" 1967" .Transport of sediments by waves, Adriatic coastal shelf,

'

Italy: Am, Assoc. Petroleum Geologists, v. 51, p. 1304-1319.?ETrIJOHN, F, J., 1949, Sedimentary Rocks. Harper and Bros., New York, 526 p.?OtTER,P. E., 1963, Late Paleozoic sandstones of. the Illinois basin: III, State Geo!. Survey, Rept.of Hivest.

217, 92 p. ",' ','

", .

",-~UBEY,W. W., 1938, The force required to move particles on a stream bed :.u. S. Geo!. Survey Prof. Paper,. 189E; p.121-14L

,

,', ,,',-

'.\i1'rA, SANDRO,'1968, Bluejacket Fonnation~a subsurface study in, northeastern Oklahoma: Lnpub. :M.S.. Thesis, Univ. of Tulsa, 142p.

,

"

''f!IELDS,

A"1936, Anwendung der AehnIichkeitsmechanik und der Turbulenzforschung an die Geschiebe-

. bewegung: Mitt. Preuss. V ersuchsanst. vVass~rblau und Schiffbau., Heft 26, Berlin.\1)(IJO\vSKI,K. R., 1958, Diesynoptische Methode des Korkt1ryen-Vetgl~iches Zltr Aussenzug iossiler Sedi-. Il1entationsrij,ume: G"eo!.]ahrb.,v. 73, p. 235-275.

,',,.

'?t);"CER,D.W., 1963, The interpretation of grain size distribution curves of clastic sediments : Jour.,Sedimentary Petrology, v. 33, p, 180-190,

'IiEKER, E.' M., 1949, Sedimel).tary facies and assodateddiastrophism in the Upper Cretaceous of centraleastern Utah, p. 55-81 in Longwell, C. R., chairmal}, Sedimentary facies ih geologic histor\", Geoi. Soc.America );fern. 39, 171 p. , .

.r~!;LEY,D. L 1963, Non-turbidites in flysch-like sequences: G,eol.Soc. America, Spec. Paper 76, p. 155-156.fJXDBORG,,AKE, 1956; The River Klarah'en-a study of tju\"ial processes: Geografiska Annlaer,

\'. 38, p. 127-. 316.'S. Waterways Experiment Station, 1939, Study of materials in suspension, Mississippi River: Tech:

Memo. 122-1, Vicksburg, La., 27p.

I }!

..

1106 GLENN S. VISHER\ .

VANONI, V,A.,AND.BROoI\:S, N. H., 19$7, Laboratory stUdies of the roughness and suspended load of allU.vial streams: Cahf.lnst. Tech. Sed. Lap., Rept. E-68, 121 p..

VIS<j:ER,g. S., 1965a, Fluvial pr~esses as .interpreted from ancient an~ Recent fluvi:¥ 4eposits; p.. llIH32<n Mlddletonr G. V., ed;, Primary sedmlentary structures and theIr hydrodynamic interpretatIOn. SOt.Econ. Paleontologists and Mineralogists, Spec. Pub. No. 12,265 p.. . .

.

-, 1965b, Use of the vertical profile in environmental reconstruction: Am. Assoc. Petroleum Geologists.v. 49, p. 41-62.

-, 1967a, Grain size distributions and depositional processes: Pre-print VII International Sedimentologi.Congress, Re"ding and Erinburg, England, 4 p. . ..'.

\

~, 1967b. The relation of grain size to sedimentary processes (Abst.) : Am. Assoc. Petroleum GeoLBullv. 51, p. 484. . '. .,"

WEIMER, R J.. 1965, Late Cretaceous deltas Rocky Mountain region (Abst.): Am. Assoc. Petroleum Ge.ologists, v. 49, p. 363. '

,

'"'.

JOURNA / SEDIMENTARY PETROLOGY,VOL. 39, No.3, P; 1107~11l7FIGS. 1-17, SEPTHIBER, 1969

,"\

EXPERIMENTAL INVESTIGATION OF PRESSURESOLUTION' OF QUARTZl

f J. RENTON, M. T. HEALD, AND £. B. CECIL'West Virginia University, Morgantown, West Virginia

ABSTRACT

Experimental pressure solution of quartz was conducfed in hydrothermal reactors. Loads ranged' fr0m2500 psi .to12,Ooo psi, with temperatures ranging from 270°C to 550°(, Pressure solution occurred in distilledwater as well as in solutions of. NaOH, Na,CO., NaC!, and naturaL brines. Clearly defined pressure solutionpits. were readily observed on. faces of quartz crystals which had been surrounded by small zircon grains an<Jsubjected to load. . . .'.

The rate of compaction of fine sand was much greater than that of coarse sand resulting in a pore spacereduction of 70 percent in fine sand compared to a 45 percent re<luction in. coarse ~"nd. The rate of compac"tion of fine angular sand was approximately 2.3 times that of fine round sand. As a result of pressure solutionand growth, the. appearan~eof !he angular grains was lit!le ,different from that of the. round' grains aftercomparable pressure solutIOn. Simultaneous. pressure solutfin. and quartz growth .in sand samples producedaggreg«tes which were considerably stronger th"n those resulting from cementation "lone.

.

..~amplescomposed of gmins of chert responded to pressure solution much more rapidly th"n mOnocrys-tallme qu"rtz. The chert did not completely recryst"lIize, but gmin

.

b.

oun&ries'iJecame ver)! indistinct and theresulting product resembled" solid mass of chert.

'The experiment"l studies show iliat"s"

result of pressure solution, initi,,1 differences in texture and com-position of natural sands may lead to striking differences in fiua.l porosity. .

INTRODUCTION

The phenomenon of pressure solution refers. to the solution of quartz at the point of grain

cQntact as a result of stress, generally from loadpressure. The effects of pressure ~olution'havebeen' widely observed and reported from petro-graphic studies of sedimentary rocks. Probablythe most dramatic expression of pressuresolu-tion takes the form of stylolite seams. Heald(1955) described the formation of stylolites insandstones and stated that.the silica producedby stylolitic solution could be an importantSource of silica cement in s'ediments.ln someparts of the Simpson and St. Peter sandstones,the entire volume of secondary silkacementmay have been derived by pressure solution(Heald,1956).

The possibility of clay pr6moting. the processof pressure' solution has been proposed by anumber of workers. Heald (1956) suggestedthat the clay simply aCted as a catalyst. Weyl(1959) advanced the idea that solution wits fa-Vored by the greater diffusion through day be-tween grains. Thom§on (1959) theorized that theclay promoted the process of pressure solutionby providing a microenvironment of high pH atgrain contaCts. As the silica migrated awayfrom the points of grain contact into the porespaces, the silica \vouldbe deposited as secon-

:Manuscript received October 16, 1968.

-- Present address: Colgate University, Hamilton,

~e\V YOrk.

dary quartzo1Tergrowths on the quartz grains asa result of a decrease in pH.

.

,METHODS OF INVESTIGATION

Recognizing the potential importance of pres-sure solution as a process of sandstonelithifica-tion, the present writers initiated" series of eX-.periments .designed to produce. pressure solutionuni:!er laboratory conditions. Fine grained nactural quartz sands have been used in most pre.vi~:JUsstudies of compaction (Maxwell,' 1960;Ernst and Blatt, 1963). Modifications develop-ing on these grains during the eariy stages ofcompaCtion are difficult to distinguish. fromoriginalsur£ace irregularities, and mechanicaleffects cannot be clearly distinguished fromso-'lution effects. In an attempt to increase the pos-sibility of detecting solution effects at the con-tacts, grains. were polished by air abrading withfine grit. The treatment was partiaUy successful.in that Some small pits were ()bserved on the'polished grains at the cOmpletion of the pres-sure solution experiments. However, the pitswere so small. in the early stages of sofutionthat quantitative measurements were difficult tocarry'.out and mechanicaleffeCtscould not be .

easily recognized. .

It was found that the best method of observ-ing incipient pressure solution wa's to use crvs-

. talsor cut plates of quartz surrourided by grainsurider load.' In transmitted light under a petro-graphi'C microscope, solution pits o'f extremdy