Journal of Jounal of Sedimentary Research, 2006, v.76, 731-770 Sedimentary Research Article Ok 10.2110/).2006.080 Research WIDTH AND THICKNESS OF FLUVIAL CHANNEL BODIES AND VALLEY FILLS IN THE GEOLOGICAL RECORD: A LITERATURE COMPILATION AND CLASSIFICATION MARTIN R, GIBLING! "Deparennt of Eta Scie, alone Ui, Mali, Now Seto DSU 398, Coma ‘sail: mailing Ansiacr: ‘The three-dimensional geometry of fluvial channel bodies and valley fills has received much less atten thr internal structure, despite the fact tiat many subsurface analyses draw upon the geometry of si Although eliannel-body geometry has heen widely linked to base-level change and aecommodation, few stuies geometry, and present a literat width (SV) and thiekness (T) are recorded. Twelve types of channel bodies 2 gcomoephie setting, geometry, and internal structure, and log-log plots of W against Tare presented for each type. Narrow and Iiroal ribbons (WIT < § and 5-15, respectively) and arrow, broad, and very broad sheets (WIT 15-100, 100-1000, > 1000, respect rl The dataset allows an informed selection of analogues for subsurface appli spreadsheets and graphs can be downloaded from a data repasitory. Mobite-channel belts are mainly the deposits of braided and low-sinuosity rivers, which may exceed { km in composite thickness and 1300 km ughout geological time reflects their link to tect activity, exhumation events, and high sediment supply. Some deposits that rest on Mla-lying bedrock unconformitis cover areas > 70,000 km*, In contrast, meandering river bodies in the dataset are < 38 m thick and <° 13 km wid, and the organized flow conditions necessuty for their development may have been unusual, They do not appear to have built basin-scale deposits. Fie channels and poorly channelfzed systems are divided into distibutary systems (channels on megafan, deltas, and di alluvial fans, and in crevasse systems and arulsion deposits), through-going rivers, and channels in eolian settings. Because ‘widelimasimuin depth of many modern alluvial channels is between 5 and 15, these bodies probably record an initial aspect ratio followed by tor avulsion. The narrow form (WIT typically < 18) e ce and rapid filing, although some are associated with base-evel rise, Exceptionally narrow bodies (WIT locally ionaly relect unusually deep incision, compactional thickening, filing by muss-flow deposits. balanced ager of natural levees and channels, thawing of frozen substrates, nel reaceupation. alley fills rest on older bedrock or representa brief hiatus within marine and alluvial successions. Many bedrock valley have WIT < 20 due to deep incision along tectonic lineaments and stacking along faults. Within marine and allu tupper Paleozoie valley fills appear larger than Mesoznic examples, possibly reflecting the influence of large glacioeustatic sin the Paleozoic, Valley fils i ‘and proglacial settings are relatively narrow (W/T as low as 2.5) de to incision from catastrophic meltwater lows. The overlap in dimensions between channel hodies and valley fills, as identified by the original authors, suggests that many braided and meandering channel bodies inthe rock record occupy paleovalleys. ‘Maieling has emphasized the importance of avulsion frequency, sedimentation rate, and the ratio of Aoodplain width in deter J-hody connectedness, Although these controls strongly influence mobile channel belts, they are ess effective in fixed-channel systems, for which many database examples testify t0 the influence of oeal geomorphic factors that include bank strength and channel aggradation. The dataset contains few examples of despite their abundance in many formations. Whereas accommodation is paramou ated through geomorphic factors, thus complicating inferences about base-level con ssronvcri0N. bounding surfaces (Mia 1988, 1996; Jordan and Pryor 1992; Lunt et al Riverchanne and river-valley deposits are prominent in the goolosieal 2004), thus emphasizing the internal heterogeneity. chat common! record, where they range from the smallest oodplain channels to the eonrols water and hydrocarbon flow through the channel fils epost of continentabscale rivers that dominated ther landscapes and In contrast, only a few accounts (Krynine 198; Poter 1967: Priend fled entre basins, Over the past 20 yrs, studies of ancient and modern 1983; Fielding and Crane 1987; Reynolds 1999) have dealt compen Ahoial channel deposits have Focused ngayon thei internal orgamza- sively with the dimensions aud 3D form—or exteraal geonsiry of 1 following the approach of identifying architectural clemients and eharinel deposits and valley Fill, Such information is topical fr several Cope 2006, SEPM Sosy for Ssbimotry Geology) 127. 1AMNOATESISISONm MAR GIBLING JSR Tye |= Qualtative ters used to deveribe satchel Bodies aad flava fs ‘Guahlave tame Flava anal boy Ser Maltsory Matiaterat Simpl ad complx bodies Central boo Wing ‘Swossson dominated reson dominatad Conse il Aspmmetisfi Azwradatonindee Dens Bots (Channessnsided Sheers Link Cumchbend scour (or valybemd riba junction sour regular scour Avchiteswal demons Bounding surices Definition of Terns That Describe Chal Bois Tvedlinensioval said form composed of unconsolied or Fife seen, generated by iva harm proses troup tims. Muy eopresent an individ chal bay oF may bea composite of vo or mate canal bodies. The tenn nelads channels, which represen the ing of and without hange in its prince (banks and sal $url), for example the fil ofan shandond chars Erosonlly based componcat of s chanel bod Sand body of ene ejel is supsrinponst epon on oF oe carl sand bods. Somsines wed o indict sera Stacking of stories Gert lem for chinn Bos ith more tha one story. Literally cotessent sd bodies, Singlestory and muistory bodes, respectively. Maia part ofehanel bod ‘Thin marginal prt f channel bods, distinguished fon ‘cet body where bas cour shows a distin iatiection Point, Usually composed of ewe andor erevasespay epost Stris represent the amalgamate, relatively complet fs ‘of csc cans Storiss represent smaamated bees spurte hy scours gzneated by shorcist evens pel foods, within hanna Scour surface that unde sor. Infilig of veavely nur, single-story cane active ‘orabandand) by deposition on ie Moor and sions Danks, progressely reducing the erswaseuna te, Infting relatively maton. singe stony and ative ‘anal by aba that ast Tatra ore eapily chan ‘Me chanel bak reese Parameter that dire ate of climb of lita aston Serle. spproehing tfesstt bask. Dlied as VIL for {wo sucssive pots of manu cca it of inn stein saraces(V ~ hight of sorta gestion = ‘stance of lateral aces), Branded, elongate nl ypclly sinuous channel ads ‘ith buat or dsebtary puters CCalessece of channel bods tbons amd denroi to form composite bods, commonly trough ral migration Mattson Body ia which iid soi have wid ‘ks ais fs than 20 Malistory body in wich idol stories ave width thikxs rato greater th 20 = Region ofa chancel ody Been b Region of unl dep scour a the ase of a chanel or ‘alles oy Rondrng et Regio of usually deep scour atte use of a chanel or "alley body atthe confuense ofthe ain chun st it tbr: Region of unusily deep scour atthe ose ofa alley body where ay ars Region of unustly deep scour atthe base of & chanel or “ally body tat shows no apn aio to meats oF th drainage network such 36 ban o ibis, (Component of deposond! yen equalon i i: 1, oF Siler than, ene il, args ham i tis uit Surfaces tht bound bods of sats nldng architectural ‘ments, commonly develop in hitarchy fom cal 0 ahi eet sn i ‘Author Pour 167, Raigon 1958 Feoflba 198, cd in Bote 1967 Feoflota 1984 Gite in Pate 1967 Bridge and Mackey 19934 Pour 1967 end etal. 1979 Boner 1988 Denier 1958, ‘This paper This paper Fried ta, 1979 ‘pki 1985; Kirschbaum and MeCabe ‘Hopkins 1985; Kinchbaom aud MeCake 1992 Gibling and us 1990 se lo Casas Govalo 19S Pej eta 1972 Petjobn et a. 1972 Cone 1978 CCoar1978 onon 1935 Arie etal. 2002 Anis eal 202 Andis x 2002 This paper Mat 1985, 1996 Mil 1985, 1996 reasons. The discipline of sequence stratigraphy requires a strong understanding of the styles and dimensions of Muvialchannel and Fuvilvaley fils becuse alluvial kudseapes may respond rapidly so forcing factors such as buse-level change and elimate (Blam and ‘Tomngvist 2000), Thros-dimensional seismic surveys are providing remarkable subsurface images of channel bodies and ally fill andthe ssomorphic surfaces to which they are related (Posamenter 2001). The ‘modeling of sediment transport systems (Paola 2000) and thee response to forcing fietors requires a fuller knowledge of Quaternary channel systems and geomorphology. Furthennors, aia channel deposits formFLUVIAL “HANNEL BODIES AND VALLEY FILLS By Plan form, Longitudinal form Bed configuration ‘bed form height + spacing Cross-section form ‘wit, depth + aroa Frio, L-Geometi dessripton of made chanel fora, No see “edified fom Knighton (1998, 2auifers and hydrocarbon reservoirs, hast economic minerals, constitite of und floor rocks in coal mines, and are associated with important it stes. Many subsurface investigations of channel deposits whore 2 are sparse depend on the application of suitable analogies with 2 known range of dimensions “The pesemt per explores the external geometry of chanel bois and i fills preserved in the geologial recor. Such an enterprise requires ‘only an understanding of the channel bodies themselves but also = consideration of thee broader geomorphic setting: the relationship of channel to is floodplain: vial interaction with other depositional sestems sch as deltas, eolian dunefields, and glaciers: and the erucial kage between alluvial hasins and river courses in eroding uplands re Nuvi deposits may accumulate 3s unconformity-hase valley fils Paper sets out terminology use to deseribe channel bodies and valley °s and dravs together large database of case stds (heeafler termed. e dase) from the literature, mostly published over the past 30 years, +5 provide precise information on their dimensions. In order to use the of geometie data most effectively, the channel boies and valley classified on the basis of internal constitution and geomorphic supplemented by a consideration of their width : thickness -ssipuion, Finally the paper discusses the factors that control the form * channel Boles and aley fils inorder to explore how ehannel ystems ‘iliar wus in modern lnnseapes generate channel bodies wth the snsions and form that we observe in the rock record, as wells _exeloting the use of the dataset fr modeling and subsurface applications. ‘QUALIENTIVE TERMS To DFSCRUME CHANNEL Chane Boies Fuvial chamel deposits comprise a suite of widely reognizal cserpaneats. A st of bedforms, sich a8 dunes and Fipple. type crane ito bars and bed sheets, which within camels (Bide "53: Lunt etl. 200), Channel banks tomo migrate tral the sesacent floodplain deposits are erode with concomitant later -S:vation of bats within the hana, and the channel base may ineise ‘5 the underyng oodphin depos, resuing in stacked bar depois ‘ee blond shows In these case, the evoltion of the channel generates chanel Bad (Table 1) that larger than the igi (nstantaneo0s) funnel dimensions. A special ype of channel body involves the filing of 3 without change in its perimeter to generate acne! fl. For fsrple large mast of landslide derived sediment may suddenly ill an sce chanel [Keser 199), or repeated Noo events may saul ill aw absdoned channel. fo the case of channel ills, the dimtsions of the fil approximate the instantaneous chant! dimensions Individual channel bodies commoniy amalgamate to form a eompsite ‘channel body when relocation (avulsion) of the river channel justaposes ‘younger and older channel bodies. The juxtaposed sagments may have been deposited by the same river, typically over # shore period, or may fepresent the emplacement on the floodplain of a dilfernt river. perhaps alter long period. Ptiohn etal, (1972) used the term dell to deseribe coalesced smaller bodies, typically formed by lateral migration of channel, and this term has been widely used as a synonym for channel body. However. the implication that belts involve coalescence of channel bodies makes the term inapplicable to single-story bodies and channel fs, and the more goneral term “channel body" is pelerred here ‘The majority of the channel bodes and valley Fills compe! in the dataset are Hess than 60m thick. but-no upper thickness limit was presribed because no natural breakin thickness was apparent, Hens; composite arrays of amalgamated channel deposits may reach hundreds of meters to more than a kilometer in thiekness- on the scale of “bist Fill” ~and ae included inthe analysis A flail camel body can be defined as a theeestimensional form composed of unconsolidated or lithe sediment, generated by Nuvi channel processes through time. This understanding follows Potter (1967), who defined a sund body as a "single, interconnected mappable body of sand.” He included the term “interconnected” to take account of the branghing pattes of many bodies and the superposition of sand bods of different cycles, and he included the term “mappable” te Liscinguish them from mast single bods. Knighton (1998) defined athe: dimensional slid form parameterized at some appropriate stage suchas tbankful. Modern channels can be described in tetms of thsit cto sectional form (widhh, depth. cross-sectional area, wetted perimeter, hydraulic radins), planform, and longitudinal form (Fig. 1), and constitute a general starting point for describing ancient Nuvialhodics ‘The time required to generate a channel body may range from decades in the ce of the most frequently avulsing systems (Sinha etal. 2005) tens of million of years in the case of thick valley fils that revord the prolonged transfer of saioent From seve orogens (Vineent 2001) Practical difficulties arise in defining 1 “vhannel body.” Tobe recognized as such, channol sediment must be embedded within extn channel material typically ineerained floodplain sediment. However, disetete channel bodies amalgamate partly or completely to form ‘composite bodies with varied degrees of connectedness, and describing their dimensions becomes a matter of julyment. Settings that particularly promote concentration of channel bodies include sites of drainage entrenchment (valleys), restricted entry points for drainage into. un confined plains (alluvial fans and aprons), and basins with differentia subsidence that preferentially draws channels into certain areas, The degree of connectedness of channel boule has been widely studied using compiilerhased models (North 1996: Bridge 2008), Criteria toe the ‘ntification of valley fills are discussed below. Terminology for 2D Analysis Terms used to deseriby channel bodies are shown in Table F and Figures 2 and 3. Channel bodies in cross-sectional view eat he divided into singlestors bodies and maltistory bodes (simple and complex bodies of Friend et a. 1979). Many bodiss. especially single-story bods, ‘comprise a cemralbad and was, The central body represents the main topographic low and may be symmetric oF asymmetric with the zone of mavinmom thickness near one margin (eg, Torngyist et al. 1993 Hampson et al, 1999a), The ings may represent a relatively wide topmast story. in which eae they ate part of the chanel ody. or may represent natural levee and erevasse splay deposits connected to but sistnet from the channel fill, Channel-body and adjacent strata™ ISR MULTISTORY CHANNEL BODIES eg wats ee Suocossion- dominated Sa Mubateral Erosion - dominated utitory SSS EES ‘SINGLE STORY CHANNEL BODIES AND INDIVIDUAL STORIES OD Vw ‘Symmes Asyrmaiis——_Coveantio SS: Aograstenindon © t amma intergtate in some cases (Hill 1989; Nadon 1994, implying coordinated aggradation. Because central Bodie and wings generally differ in agiler and reervoir characteristics (thology, permeability), the width and thickness of the wings, as well as the central body, may be important (Fie. 2, Povter (1967) distinguished molistory bos and muller hols, bso on vertical stacking and lateral coalescence of stories, respectively {In weality. story arrangement commonly combines veri! and lateral Positioning. and the term multory ean be used to describe ods with several stores, however disposed (Bridge and Mackey 1992a). Bodies ean bee termed sucessonedominared where the stories tepeesent reasonaoly complete channel fll, with only modest erosion of topmost strata beneath subsequent story scours (eg, Hampson etal, 19993). However, many bodies contain abundant erosional surfaces that could reprssemt sori-term evens such as Noods; these bodies are ene termed erestan- ‘laminate. This sinction may’ be dificult to draw, Many bodies contin lenses of fine-grained material, and such permeability bartiers exert an important influence on Muid flow (Mall 1988: Robinson and McCabe 199 ‘ollowing concepts set out by Allen (1983), Misll (1988, 1996) described) a hierarchy of hounding surfaces for fhuvial bods that, although not elaborated hete, constitute erueil part of channe-baly analysis. Some surfaces separate distinctive balsels and barforms (rchvectuatelenunt), whereas higher-order bounding surfaces delineate Foo, Terminology for dseribng th «oss sstonl genet of chanel Bois, ‘Tee rin ofthe ems is nota a Table entire channel bodies and stores within multistory bodies. Holbrook (Q001) snd Mil! and Jones (2003) presented good examples of the use of this hierarchy in studying complex channel bodies. Singlestory bodies and stones within multstory bodies san be described in terms of their overall symmetry and fll geometry (Fig 2). Asymmetric ils form in channel bends whewe a bunkeatached bat aotrtes laterally due to cuttank erosion and deposition of sediment ‘wansportd ftom upstream. I some cases, the bar migrated more rapidly than the channel Bank rita progressive rise of the aceretion surfaces can be represented sumercally By the aggradation index Ganceniric fils © progresiveiling of « channel (active of abandoned) banks, pro- aresively crosssectional area; modern ephemeral iver yield examples of this filing style Sehumm 1960; Taylor and Woodger 1978; Schuman 1989: Gibling eta. 198). Termisology for 3D Analysis Few terms ate available re mensional form. Pesos sre channel bodies in thie fll ures ‘al. (1972) identified deni bodies ically sinuous with eibuzary and disurary form, Some additional mi ae i als betwsen branches ean be termed fink, bs analogy with modern drainage networks (Horton 1945), Exceptionally thick zones represent reaches whore the thaweg wasJSR i.— Nalley-constriction =soour Convergent w7 branches (tributary) Divergent branches (distributary) ‘Slope of upper [/ ortower sutace Fo. 3 Terminology for desing the thre mensional form of channel ‘bodies. The diagram is based on valle il with dotnet branches and uibuaies (eg Feld ot al. 1995; Phin 2002), but th terms may aso be applicable to broad channel bodis with complex history and internal prone exceptionally deep, and include confluence scours, channelbend seous, and valleyansirition scours (Ardies et al. 2002), as well as iregular -scous—the later showing no apparent relationship to channel or valley form. Salter (1993) noted that “scour” results from processes of erosion intrin to fuvialchanne's, especially at bends, confluences, and control points (structural elements, clay plugs, buf), and Best and Ashworth (21997) noted that scour depth in conluenees and bends may be iv times greater than mean channel depth. Zones of unusual widih (@.., Pint 2002) may’ reflect the influence of control points such as tributary and Gistriburary postions Salter 1993). If branches are present, drainage potters (ellis, dendrite), convergent systems, and divergent systems can be identified (Thomas and Anderson 1994; Feldman etal. 1995; Ye eta. 1999; Posamenier 2001; Arde et al. 2002; Pint 2002, _FLUVIAL CHANNEL BODIES AND VALLEY FILLS _ ns DIMENSIONS OF CHANNEL BODIES AND VALLEY LAS Previous Classifications ary casifiations for channel-body dimensions (Rich 123; Table 2) focused on lngthivith—a combination of longitudinal and eros sectional measures. Krynine (1948) recognized the uty of with! thickness (hereafter abbreviated to W/T), with wicth meastied perpen dicular othe local clongation ofthe sdiment body (Fie et. 1979) or tothe measured paleofow dein. Maximum thickness be widely used, and tepresets the site of maximum thickness af a single-story body orof amalgamated striesin a mulistory body. Comparison of msan and mmaximam yalues may be meaningful but hes rarely beew upp. In wo casi papers, Friend etal. (1979) and Friend (1983) used a WPT valu of 15t0 divide chan! bodies into ribbon and sheets Table 2) Tis division hasbeen generally accepted and, os discussed ltr, accords with the aspect rato of modern channels some workers have retired ubigher ‘boundary value Blakey and Gutitos (1984) separated narow and broad sbets a 8 WT value of 100, and Krynie (1948) defined tabular Does (QV7T 50-100) and blankets (WIT > 1000), Poter (1967) used shots and blankets to describe ese equiimensional bodies width and lngth).Feend (1983) drew an important Fnkage between channse body form and channel behavior, distinguishing mobile chat bel, ‘fixed chamels, and sheefood seins with ited ehannaiztion; tis division is developed inthe present stad. Based on the dataset, Table 3 resets a revised casifiation of channel bodies in terms of width, thickess, WT, and ara Previous Approackes to Compilation Many approaches exist for investigating and predicting channel-body seometry (se eviews by Martin 1993, North 1996, and Bridge 2003) The pprosch taken here insoles the ploting of W/T data on a log-log scale fora lage suite of Quaternary and older bedrock examples. A. crucial problem in using examples from the ancient record is the sparsity of ‘lable data on channel-body width (Tye 200) Several previous studies have prodoeed compilation. To test the lateral extent of sand bodies generated from highsinuosity Nuvial systems, Colinson (1978) plotted the dimensions of anciont chaunel bodies in comparison with a statistical relationship between channel depth and meanderbelt width for modem rivers. Fielding and Crane (1987) pow channel-belt wih azainst depth or thickness fra lace suite of modern rivers and ancient bodies. Their data set spans examples 0.3 to 0) m dep or thick and 1 m to 20 km wide, with a WPT range of 3 to 2000, Superimposed on the plot in onder of progzesively increasing WT were (0) an upper bounding line, aso delineating incised, straight and non rnigrating channels, (2) an upper bounding ine for meandering channels (G)Collnson’s statistical relationship for Tully meandering riers, and (8) 8 lower bounding tne, abo delineating braided systems. However, their datasources were not recorded, and the mixing f data from modem river channels and ancient channel bodies fs problamatic. Using average WIT vas and range bars for datasets, Cw 1991) and Robinson and McCabe (1997) presented logog plots of channel body dimensions for several Suvi groups, in eaton to « WAT value of 1S. They distinguished fixed ribbons, meandering ribbons, sheets deposited from low-sinuosty, braided or shetlood s}stems, and sheets deposited from highly siavous and meandering systems. Some kg: composite sets had WIT values up to 20,000, Dreyer (1999) presented compilations on linear plots. Reynolds (1999) ploted data for 409 channel and valley bodies in parle sctings, identified in outrop and stisurface studies, ona logog plo, Although this isa large dataset, no information was provided about the localities, seuimentary features, and agnostic features ofeach type, and stacked sand bodies were excluded. Reynolds generated eumlative fequeney plots for sand-ody width, as136 MAR GIBLING ISR ‘Tame 2.—Geomerie measures nse to fine fail-channel odes Author ‘Channel Boay Dinnsions Ron Lemp wed od nine 948 Potter 1962, 1967 Pen et al. 1972 McGugan 1968 Couer 1978 Fiend ot al 197% rind 1983 Blaby and Gubitoss 1984 ine “hoosring ml Cength >> i) ‘Width sn thickness noted for some bodies Classified sedimentary bodies ofall ype ‘Wats are > 8 km, scar $0 kn, sll < 8 km Thickness thick > 1801, dian 30-180 hi <8 m9 Wilcke ares Leng Volume ge > $0, seen 1-500, soul < Length wed to distinguish (2) suidimensioal boss hss or blankets. ~ I) 8) —inegudimsnsonsl bose (pode — 31, ribbon or shootings > J, demas with ‘wanching frm, bts with complex pater of cose). Persigene Factores etenverage tMckaess Hoth tems measured ia Same uns, but uso ‘em and m, respectively are wed ere. No categories deine. Widinthcknass of stores within a multory (braided) body: chamell-brded < 2, ses ‘td 2, Widtthicknes (or eighty sibbon < SE, sheet > 153 Wihivibickmes ribbon < 15:1, arrow shox 15-10, broad set > 100 Atkinson 1983, ced in Alexandr 19323 Nadon 199 Fendt a 2001 Wikis hake sibbonshet boundary revised vo 25:1 Widitthickness: ribbons boundary red to 3 Miecbodies = 1.2 m bck, minor sheets 26 tin eps 6-12. hick mapas (or Stalk Group exposures) 1 guide 10 most probable widths, and related vial syle to systems tracts. These studies have outlined the range of width thickness, and WIT of ‘channel bodies in the rock rssord. They have alo shown that channel bodies form & coutinvum in W-T space, with mobile channel bets (braided and meandering) at larger WIT values than fixed channel bodies. The studies have tended to represent channel style in terms of planform and the 15 W/T boundary. However, theres scope foram approach that combines quantitative information with a qualitative, geomorphic ara Compilation and Analysis For the present study U compiled iterature inthe English language on channel bodies and valley fills from the Quaternary and older bedrock records, for which the authors provided measursments of width, thickness, WT, and (less commonly) area and length. This information is either explicitly stated or can be calculated from diagrams and maps. ‘Channe! bodies less than Im thick were excluded but no upper size limit vas imposed. Data suites include exact width and thickness for single bods, width and thickness rages for suites of bodies, width ranges for a single thickness, and thickness ranges fora single width. Where the available data include ranges of width, thickness, and W/T, a more restricted distribution of width and thickness could be estimated than was possible for width and thickness ranges alone. ‘The inclusion of such general information seduced peesision but allowed a moch larger suite of examples to be considered. This helps to ‘overcome a bias towards smal and low W/T bodies accesible in smal foulrops, and is especially important for extensive shots, the full dimensions of whieh are unknotvn, Some of these shets may represent “>big rivers” comparable tothe largest modern channels~a group that is underrepresented inthe literature (Poter 1978; Mall an Jones 2003), The dataset represents examples from about 185 individual state graphic units from all continents exeept Antarctica, and from Archean to Holocene, though the bulk of the examples are Devonian or younger. The numberof channel bodies is dificult to asses, inasmuch as examph ange from single valley fils and amalgamated bodiss of basinal scale 10 formations with hundreds of small bodies. Adtionally, some studies provide goneralizad estimates for Muvial bodies hundreds of Kilometers Jong, or quote ranges of dimensions for suites of bois but do not state the numberof bodics studied. However conservative estimate pus the total number of discrete channel bodes at well over 1500. “Many excellent facies studies were excluded because suitable dimensional data could not be obtained from the available outeops or wells oF wer not rovorded by the authors, Although subsurface datasets commonly provide ‘good isopach and length information, fuvial style may be diffi o assess From limited core, muddy fils are dificult 10 idem, and limited well intersections preclude accurate width assssment (Lorenz eal, 1985 Bridge and Tye 2000; Tye 2004), Consequently, most of the selected subsurface studies include some outerop information, A compilation sue ‘Taaue 3—Clusifaton of fitch bois an fava ills according to size al form base othe present sty Wid oy Thickness (H) Wath Takka Arca im Ney Wie > 10000 Very Take >a Very Broad Shes > 100 Ve Las > Toa Wide 5190 Thick Sis Broad Shots > 100 Lane > Low Median ti Medium as Natt Shee 315 Metin Sau Narrow 210 Ton 31 ‘Broad Ridbos Smal a Vary Narrow < 10 Very Thin <1 Narrow Ribbons es Very Stat <0ISR FLUVIAL CHANNEL BODIES AND VALLEY FILLS somone a t 1 Ra or ay MOBILE, ee narrow sheets| i canner |_| 2 Meandetng mers ‘Sony bend sens as cane Tost 7 ay omens Ce ene x Sodan we] | Seats | CHANNELS [~) —p.Crevasso channa's and ‘he AND POORLY avulsion deposits = fracas faa m2] | 2 rane 2 Pon rt 1 Chr cnt 2 or ee a Secs aad 2 Valloy fs within alluvial and ‘eos Fi, 4—Ciasfiation of fui channel ne we sce a cinnches noite aes ‘3 Vallay fils in subglacial and architecture, See Table 3 for form definitions and wane EASE cians in cet sos 4 this depends on the accuracy of description and the quality of ‘aczrpetation of the original authors, although the present author is sally fiir with some 45 of te statiraphic units represented. The set represents the reported dimensions of bodies of rock, and original Socknesses may have been grcater prior to compaction. Because of the biily of datatype and precision, no statistial analysis was attempted Tae larg dataset allows a comprehensive elasification ofthe channel ‘ocies and valley fill, based collectively on their geomorphic sting, and internal characterises (rain size, bedforms, architectural (Fig, 4, Table 4). W/T plots were constructed using information ‘arid preison, and individual datasets were plotted as poims, lnc, 224 polygons (Fig. 5). Tight envelopes were constructed around the collective data (dashed lin in Fig. 5). The diagrams (Figs 6-10) represent se most common dimensional tendency for channel bodies and valley ‘is of recognied types, bu the large amount of data precludes notation rofe than a few key examples on the graphs. To make the dataset accesible, the SEPM Data Repository (see Acknowledgments seston) contains a mich fuller account: & reference list of literature ples under each category ofthe elassifiation (Appendix 1); ase of spreadshects and W/T graphs in Exee!® format that ean be vsnloaded, slong with an account of how to ereate and modify the aphs Appendix 2); and ase of photos of some well exposed examples Appendices 3-9). Thus, interested readers can identify individual tases on the graphs and select analogues that suit their purpose, as sell as plotting their own data. A checklist that sets out information ded for a full assessment of channcl-body geometry is included Appendix 10), as well as «key diagram (Appendix 11) to assist in asifing an individual Nuva! body under investigation. The graphs will be updated periodically on the website . ‘The data are presented as log-log plos primarily because of the large ‘ange in dimensions -more than five orders of magnitude for width and ‘more than three orders of magnitude for thickness. One disadvantage of ‘he log-log plots that readers neod to bear in mind is that sight dierences in position at high thickness and width represent major changes in body dimensions. However, loglog plots are aso appropriate for hydraulic ‘and geomarphic reasons. Channel-forming discharges have a log-normal isibution, with progressively fewer events towards the high-ischange cnd (Yu and Wolman 1987), and river systems tend 10 contain many sill and few very large channel reaches. For the rack record, the present ‘dataset tends to confirm the abundance of small channel bodies and relative scarcity of very large bodies. Thus, ehannels—and by extension ‘channel bodies~may be suitably represented using log plots (Robinson and MeCabe 1997), However, the superimposition of individual channel ‘deposits through time is key to preclude simple dimensional trends “The dataset represets single and multstory channe! bodies and valley from | 10 1400 m in thicknes, from 2 m to 130) mn in ‘WAT from less than | 19 more than 15,000, Because no history sand-bed channel bodies that lla large proportion of their parent basins. The length of chan boss rarely reported except in some subsurface studies where channel systems have bsen teaced for sore than 330 km (Pint 2002), In view of this sparse information, is dificult to assess, ith the largest bodics exceeding st upparenly much less than 1000 ki’ Evident from the dataset isthe convergence of channel bodies and valley fils from a wide range of settings into broud ribbons o narrow sheets (N/T 5 t0 100). Whete many channel bodies were studied within ‘narrow stratigraphic interval, itis apparent that a WYT spectrum exis, with no sharp distinetion between ribbons and narrow sheets (¢.. Olsen 193; Fried et a. 2001), The plots for most grovps show « proportional increase of wide eelstve to thickness as seale increases. This result in population belts that cut obliquely across W/T lines (as noted by Fielding and Crane 1987) Although this trend may reflect in pact te ictease i ‘wid relative to dapth observed in modern alluvial channel as discharge increases (Church 1992), the larger channel bodies are mainly thase of ‘meandering and braided rivers, which tend to generate wide sheetsMR GIBLING 78 “yaar asyeou0N wy spoHy Ay FURAN BRE fue vous > 0 sw $2 aes uounuoa 3 9} > seu pac gt Bog aapowoa pan eosf oamsounoa pa nis dydrwwes1 FLUVIAL CHANNEL BODIES AND VALLEY FILLS Telagoqu 29 feu 043 ee aa‘akan WA \e wee thickness (m) 1 10 1 wiath (my "°° (CLASSIFICATION OF CHANNEL BODIES AND VALLEY FILLS Previous Classification Systems Few studs of Puvial channel deposits ave dealt with wide enough range of channel bodies to allow a comprehensive clisifcation. As noted earlier, Friend (1983) set outa tripartite casifiation of fhvial systems into mobile chet Bells, ix chamuels, and poorly chanel? sens, whieh has been widely adopted along with his division of channel bodies into ribbons and shews. The most compehensve system, svaluble is that of Miall (1996, Chapter §), set out in Table 5, ‘Miall yoognized 16 common fluvial styles, cach essentially a facies model, and describe each style on the basis of modern and ancient ‘amples. He drew also ona classiiation of floodplain by Nanson and Croke (1992), Mials classification bas a strong basis in modern rivers, and the majority of styles are variants of braided and meandering ‘tet, implying some degre: of linkage between planform and channel- body isle although partisulr emphasis was given to the observed assemblage of architectural elements, Miall did not include allviafan posits. and no distinction was attempted betwoen valley ills and ‘chinnel bodies, Although the extemal geometsy of the channel bodies ‘say not included explicly in his elasifeation, geometry was frequently mentioned in the descriptions Tamer § MAR GIBLING we Stinoen T 1000 ‘iinun / Fic, Construction methods for Hh ant width= tikes plots, Note ha otal anges (Guuomeep ‘Sf widh and thiknes report by autors a icra ‘represented by poljgoas, bones, and tangs. Stinger However. wid and thickness show abroad ontlton. and wpe ft and lower right arcs fof shapes are probably aot occupied. Ts Toundingeavlopes forthe dass may eat though these areas they projet beyond other data pins. 10000 Reynolds (1999) distinguished three types of channel body (distribu. tary, crevasse, and “Muvial") and incisedevaley fils, and provided dimensional data for each type (Table 6), His data ranges are compared ‘with those From the present study in the subsequent tet. ‘As the present compilation of channel-body dimensions progressed, it ‘became apparent thatthe large amount of information could not be ploted effectively on a single diagram. Moreover, many examples could ‘be grouped bascd on their intemal constitution and interpreted. geo morphie setting (Table 4), and some groups tended to occupy & stint part of WIT space. Consequently, WIT plots (Figs. 6-10) were ‘onstructed separately foreach category. with line drawings constructed For some representative examples (Figs. 11-13). The cusiieation system and its basis is presented below Valley Fills The dataset yields a division of fuval bodies into chanel dodies and suis fils based on the interpretations of the authors (Fig. 4, Table 4) Because strong interest in valley fils is relatively recent (Dalrymple eal 1994), many examples desribed as “channel bodies” in the dataset may lie within valleys, and only examples eypitly interpreted by the ezginl authors or later workers as valley fills ate s0 categorized. Although the widely wsed term “valley fil” is retained here, the perimeters of valleys Connon Asal styles, based on modern and interpreted ancient examples. From Mial (1986, Table 8.3). 1. Graeldominaid ives Grivel Brand wth sedinentgravity Nowe Shallow rae! aid (Scot pe") Deep gate braided Donk 7°) Grae wandering 2 Sandominte higksinwosiy ners Gris meimderng GGraelsind meandering (‘coars-raned meandering") Sandy mendes (lie meandzting") Ephemect sandy meandering Finsgtined meandering Antone! 5, Sund-domimuel lowsinusity Hers Highencey Sheed dis Lowsinuosity adedsmeanderi Shallow pore braided ( Pte wpe Deep pre braided S Saskatch nbd baie braided lahy phone setlood ("Bijou Ciek ye")Table 6, Dinwnsions of fluva-channel bodies and valley fils in examples from the anciew record, fram Table 3 and figure 6 of Reynolds (1999) Thicke (1) Mees NaN Mea wr Namie = 2S 7 3-100 @ Seem Chicaes 0 7 18 +60 Pr nap a 6-90 4 om sma wo? a3 15-30 st ssmecas we ere provide, change their postion through time, and sediment volumes linked 10 valleyexight more appropriately be termed “valley bodies.” ‘adn et al. (1954) defined an incisedevalley system as a “vially- cseiaez. Songate topographic low that is typically larger than «single Sua Sem" (se also Schumm and Ethridge 1994), Because incision ‘Suuricsuss the majority of fluvial conduits Salter 1993), the existence of'imcerasional margin alone cannot be considered diagnostic of valleys, specially 22<:0s the prominence of erosional features may be largely safimction of the most recent major food (Wolman and Miller 1960; Nasib 1994) of confluence dynamics (Best and Ashworth 1997), For ‘channel 5:50 to be identified as valley fills, Posamemtier (2001) noted ‘must have cut into the loodplan suficiently that, even at sated From figs 6. The data represent maximum wlth and Ticks vals ford aot Hacked) chanel dis. No as lformaton ‘ood stage, Now does not overtop the banks, and he noted that incised tributary valleys and gullies may be important in recognizing incised systems (asin the planview of Fig. 138), Within continental stings, Mucwations in discharge of water and sediment as a result of climate, testonis, avulsion, and river capture may result in periods of incision and ‘aggradation (Goodbred 2003), so that “channels” may be transformed into “valleys” (and vie vers) over periods of thousand to tens of thousands of| years, a in parts ofthe Himalayan Foreland Basin (Gibling etal 2005; Tandon et alin press). Terraces are widely represented within modern valleys eg, Blum etal. 1994) but are rarely identified within the dataset. Valley fills idemitied in the dataset are mainly incised into bedrock oF into coastal and marine strata, and some contsin marine units that Braided & Low-sinuosity Rivers [ qo 1000 10000, width (m) 100000 ‘Meandering Rivers thickness (m) ‘1000000 10000000 Fo, 6 Width: thicknes plots for Mobile ‘Can ls, Stas wed forthe plots ae sted in Appendix I Diagram format is Plaine in igure ) Drala and lowsinaosity Tiers, C = Cadarin Formation; CH = Cy: pr His Formation, C8 ~ Castlegate Sand Hone; E = Escala Group H = Hskesury Sandstone | = Ishak Sandstone: M = Mest Rca Formation; N= Neveaste Goal Measures (G8 bod O-= ternary, Riverina, Aur ‘Group, SB = South Bar Formation, T = Tos ‘arora Formation. B) Meandering ies, ‘Beaufort Group: G = German Crick oggis Formation, Spain (Muriel rato, Indonesian Cenozoic; R= Rangal Coal “Measures (lid squares} S = Sealy Forma ‘100000 10000Me A Channels on Megetans 100 - —— eS = $0 z 1 Z 1 10 100 ‘width (m) B Delta Distributaries thickness (m) 1000 MER GIBLING SR Fic, 2.—Wigth «tikes plo forse ry stems, Sdin ued forthe plots re Te i Appendix | Diagram format i xe ised in Figte 8. A) Chanson megafans, 5 = Sunnema Fomution, SD = Sela Dei Group: U = Cnsstlo Formation B) Dea i) and tvo delasnouth bodies Goll square, G = Geman Creck Formation; K = Koote- ‘si Formation: § = Salvick Formation 5 width (m) implicate sealevel ucwation in valley cutting and filing. Valley fills incisd into vial deposits (alluvium-on-allvium contaets) ae frequent- ly subdle and more difficult 10 identify, and it may be dificult 10 ‘singuish local deep seours from regional, valley-base scours (Best nd Ashworth 1997}. Many valley fills eantain discrete channel bodies ani 1984; Vincent 2001), Valley recognition is in past a scale problem: within the dataset, eroionally based elements within smaller bodies tend to be described by the authors as stories rather than as ‘channels within valleys, eventhough many modern valleys are very small, Miall (1988) recognized “channels within channels” through his CH architectural element. In view of these issues, Fielding and Gibling 2005) suggested three iagnostic criteria for valley fills (1) the basal erosion surface and correlative surfaces in extra-channel deposits ean be traced widely, in some cases throughout the basin and between basins; (2) the dimensions ‘of the overall Muvil body are an order of magnitude larger than those of ‘other channel forms in the system; and (3) the sale of erosional reli on the basal surface is several times the depth of scour evident from ‘component channel fils. Many of the examples grouped here as valley fills accord with ll three of these eritera, although not all have been sescribed fully enough tobe certain la cases of very small valley fils such as a Kansas valley fill? m thick and $0 m wide desriped by Feldman etal 2005), only the first of thes criteria is applicable, and the valley bods C) Distal alluvial Fans ad aprons. 1) Grease channels and avubion deposits; W = avulsion deposits of the Wateco! Formation. assesment is based om the correlation between the fluvial body and an extensive iterine paleosol. The distinction of channel bodies and valley fills inthe dataset was based on interpretations provided by the orginal authors. However, some examples dacribed in th literature as channel Dovies may cevupy pabeoralieys, and future rescarch might change tit attribution. Valley fills are dived bere ito tree types (Fig 4, Table 4) based on the material into which they are incised and the processes of incision, Valley fils on bedrock seconformities are commonly angular and represent 3 lengthy period of bedrock erosion prior 10. sediment aceumulation—iypically 2 ealozcal period or longer, Valley fs within liuval ont marine surasa record 2 short period oferosion—commonly ‘one glaioeustatic qe ia upper Paleozoic and Quaternary example. This grouping requires the datzcton of “bedrock” (fully lihitid) from “sediment” (usconsotidies ce semiconsolidated) at the time of de- Position. This is not alazys eas for example, some valleys transect marine carbonates only sight oider than tbe valley fll (Fig. (3B; Feldman et al 199% Feldsran et al. 2005) or are incised into tropical alluvium and paleosls dhs ware Lckified atthe time of valley cutting (Nanson et si, 2005). Valiey ls above a lengthy hiatus (representing at least a geological period but ypically much longer) are included in the bedrock category: The two groups yield overlapping but substantially sitferetcistioutions ox W-T picts (Figs. 9, 10, A distinction based onFLUVLAL CHANNEL BODIES AND VALLEY FILLS mB Ota Atal Fane seone z 3 Z width (m) D Channels and Avulsion Deposits © erg E i i ~ i | t 4 a ; a wwe ras width (rn) ‘Se nature ofthe ill Quvial versus estuarine and marine) did mot produce Sstnctive WAT groupings. Valley fin subglacial and proglacal stings incipally comprise tamed valleys formed in subglactal stings, Chanel Bodies Nine channebbody types (Fig. 4, Table 4) are divided between «wo .zoups mobile channel belts, and fixed channels and poorly channelized ‘ostems (following Friend 1983). For mile channel belis, systematic sal migration of channel banks and avulsive channel behavior smoparted a high degree of mobility to the sjstom, Braided au low ‘seaesty river depsisin the dataset include a range of sandy and gravelly ar deposits and bedlod sheets that were inferred by the orignal authors so represent braided (mult-channe) systems or sight sinuous snklbed The parent rivers lcked systematic lateral migration, a indicated bp the scarcity of lateral-acretion deposits. Most ofthe examples belong so the satepories of graveldominated rivers and sand-dominated low ssavosity rivers recognized by Mill (1996; Table 5 of this paper), who provided detailed descriptions of thee faces. {In contrast, meandering-river deposits in the dataset show evidence that the parent channels migrated systematically through cutbank erosion at bends and concomitant point-bar migration. This process generated Aistinctive lateral-seeretion sets with paleoflow predominantly along Fao, 7—Cominad, strike of the aseretion suriess, and such sets typically extend for tens to hundreds of meters in sections normal to paleoflow (Fig. 118), Deposits ‘with prominent lateral accretion deposits are grouped in this eategory ‘Examples fll in the sand-dominated high-sinuosity rive group of Miall (1996), Although the prominence of latetal-acerction deposits is a key identifying feature, such a distinction may be dificult vo make even where paleoflow data are available: many meandering-iver point bars show components of downstream aeration, especialy at their downtiow ends (Sundborg 1956; Jackson 1976), and bars within braided rivers common show componeat of lateral acretion (Bristow 1987; Lunt etl, 2002). In such cases, interpretation of the channel bodies may need to rely on ‘range of criteria (see Mil 1996). Many moder braided, tow-sinvosity, and meandering rivers expeds ence frequent avulsion, resulting in th justaposition of deposits from different courses, This tendency is well represented in the dataset, where ‘many deposits contin lange numberof stores and are mulateal. The mobility ofthese river systems has resulled in relatively high WT values, ‘typically inthe range of narrow sheets to very broud shots ‘The deposits of fixed chamels and poorly channelized systems are Lived into seven types. Four types can be distinguished as dstibutary in style: channel deposits formed on megojans on deta, on sta alana fans and aprons, avd in crevasse charnels and audsion deposits. Ths types cam be distinguished on the bass oftheir host facies and distinctive1 MR. GIBLING A Fixed River Systems é we # thickness (m) 100 width (m) B Eolian Systems 100 thickness (m) 1000 JSR S Ro 10000 Fig. 8.—With = thicknss pots for A) Fixed river systems Quaterary deposits fom Rh Meuse sem (ope rangle) and Colubis 100 with (m) geomorphic contexts (Table 4): some of the most complete examples in latst are from extensively preserved megafan and dela deposits for ‘which the landscape setting is well documented. In contrast, deposits \ributed 10 fixed river systems yield litle evidence for a distributary style, and were interpreted by the original authors as the deposits of through-going rivers, in some eases with inferred anastomosing plate forms. They also belong to the sind-dominated high-sinuosty river group ‘of Mall (1996). Where information is limited, the distinction between Listrbutary and non-distributary ixedchannelspstems may be dificult to draw Tivo other types are Ks well represented in the dataset, The deposits of floodplain channels are generally smallscae (les than a few meters in ‘width and thickness) and are interpreted based on their intimate association with Moodplain deposits; most were probably not part of basinal drainage networks. Channel deposits in eolian settings are f distinctive group for which interaction of channel flow with ‘noncohesive sand imparts some unusual propertics ‘The dataset was examined closely 10 soe whether poorly channelized stems could be distinguished as a separate group. However, numerous sudies of megafan and distal alluviafan deposits noted that upitow regions had fixed channels whereas Jownflow regions had mote poorly channelized systems, associated with a high proportion of sandy +1000 River (oi ingly) B)Channes in eis setings. Sti asd or the plots ae ste ‘Appendix I. Diagram formats expned in Figure 5 10000 sheeflood deposits. These observations suggest that the depositional systems experienced transmission losses downstream, and preclude an easy separation of fined snd poorly channelized bois Friend (1983) characterized fixed channels as laterally stable between episodes of abrupt switching. However, nny suites of “fixed” chan} bods in the dataset include some bodies with laterabaccretion ses, although the sets typically ean be raced laterally for only a few meters to afew fens of meters (Fig. 12A). The term “Fixed” is wed hereto imply tha the development ofthe channel body mainly took place within a non mobile perimeter of loodplan deposits, with only modest bank erosion Although these systems were aso avulsive, ther is litle indication that avulsion resulted in frequent juntaposition of channel deposits. Channe!- ‘body W/T is mostly inthe range of ribbons to nartow sheets. The deposits of the (wo groups (mobile channel bels; fixed channels snd. poorly channelized systems) overlap in W-T space (Fig. 10) and donot correspond precisely with a division imo ribbons and sheets An example ofthis approach to separating fixed and mabile channel bodies eames from the Joggins Formation of Nova Scotia (Rygel 2005) ‘This formation was deposited in a rapidly subsiding extensional basin where much of the orignal gsomorphie diversity ofthe drainage network hs been preserved. The majority of 82 channel bodies (mostly single story) were cased as fixed channels with WIT Iss than 20 and verteJSR FLUVIAL CHANNEL BODIES AND VALLEY FILLS A Valley Fills on Bedrock Uncontormities Sate i io i ] 3 ™ en ga 2s] il - af 10 “00 ‘1000 “000 700000 width (m) 5 E a 3 2 + Ora 1 100 1000 +0000 00000 1000000 width (m) Valley Fills tn © Subglacial and Proglacal Settings 1000 210 3 Subatiieous 3 Shentelsvateys 3 0 lee 3 . is aw a 100 1000 10000 width (m) as Fic, 9.—Width : thickness los for salley Gils, Studies sed fr the plo ae lst in Append. Diagram format is expt in Figure 8) Valley ils on beroek uncon mies O = Osy Forastion okt squares} M = Miocene Miocene of fly and Fran MG_= Manmeile Group 8 = Sis Conloner ste; T = Mose of Tuan B) Vali ils ‘ovat ith alluvial and marie sta. C) Vat tls in subglacial al prolactin D> tunnel vals ofsore DenmsrkA Channel Fills 10000 000 a = Beene 8) 100 2 Braised 8 2 Eretbety Bo aes 10 100 ©1000» 10000 100000 108 107 os width (m) B Channel Fills too " - Channals y 8.” Fat Fiver Ealian Settings & Fixed Fi € | orvasse = Gnannals & Cranes on| 3 10 | Baposts Mestane 3 & sta a Pane one iL : 1 70 100 1000 ecco width (m) © Valley Fills 10080 ‘alo Filson ras Speers 1000 E g 8 100 z MAR. GIBLING 100 1000 ‘10000 "700000 width (rm) 10.— With: thickness emvslopss for ps fom Fgurs 69CHANNEL BODIES A mat CHANNEL BODIES : MOBILE CHANNEL BELTS A. Braided and Low-sinuosity River Deposits Castiegate Sandstone latoral aooraton unit bar 40) ~ > paaotiow B. Meandering-River Deposits Beaufort Group 3 sutace or local channel scour orosional surtacas (Chanel scour) rato ba (downstream andlor oblique Secretion) Bolnt bar with ‘accrotionary ridges (sevol bar remnants) sccrtion predominant. Although a small proportion contain Interal scretion sets, these onlap concave-up channel margin, indicating & high aggradation index (Table 1) and a relatively stable channel perimeter. [Numerous sibbon tiers (Table 1) indicate the presence of multiple coexisting channels, Bodies in redbed, dryland parts of the Joggins Formation were attributed to throughgoing fixed rivers (probably with anastomosin planforms), witereas those in greybed, wetland parts ofthe formation with standing tees and marine incursions were attributed to 400:m represent large meandeting rivers in coastal \wetlands—peobably the main deainage systems ofthe basi, Comments on the Classification System ‘Several points about the clasifcation system shouldbe en |The casification sets out groups and types that can be recognized within the proserved record offal depois. Because ll examples jn the dataset have precise dimensional information, dimensions and WIT values could be used as supporting criteria for classification. In particular, the general distinction between mobile ‘channel bodies wth W/T commonly > 50 nd fixed-chanel bois748 MR cuBLING ISR CHANNEL BODIES : FIXED CHANNELS & POORLY CHANNELIZED SYSTEMS A. Channels on Megafans Huesca System, Ebro Basin = SS = 2 : Slama mizaton ‘Bron seat —— m eat miraton boas shot muta seal ezton =z Sel Palsow man towards vowr ving (owe depots) ena bay cose sug, bon rita ie) wrre's 2 Se tora eotonsuracos tinal oar postan {concentric fil) snes SSE ; wisis \ Pa. 2am fed camo amalgamate compen nee 8m pth antl ae SMe B. Delta Distributaries Joggins Formation [siacked, aggrading channel tls fee Hess megan stern, Ebro Basin, Spin (Gist 1991). Upper patel shows ld exangl with about oF channel bois by re, and Tower panel shows hari tyes of ody in ‘he megsfin, Amalgamated comple lover ana) are present locally in oer parts ofthe ‘ouzrop bet B)Jogsins Forwation, Peasy nian, Nova Seni (Rysel 2005) C) Waters Core Formation, Pennsylvanian, Nov Sota (Gilg and Rust 1990, D) Quaternary of he (Ganga Pins, Inn (Gibing etal. 208). The han body Sm hick with WIT of 6 1B) Page Sandstone, furs, Utah Jones ad genlly dipping loves deposits "gry, hetorohic bay fis vith WIT commonly < 30s apparent, as is the great range in WIT for valley fills, with many high values. The classification makes it clear that certain types of channel deposits and valley fill recur repeatedly inthe geological record Thus, although every cae study js diferent and is to some degree its own model, a useful level of generalization is posible, ‘The classification is gonevie rather than descriptive, although the types have distinctive features (Table 4). This approach isin accord with the comments of Potter (1967), who noted that most temninology apple to sand bodiesincldes a mixture of descriptive and genetic terms, and noted that genetic terms are commonly landform names, efeting the close connection between sand-body origin and geomorphology. Some aspects of channel planform (for example, meandering systems) are included in the elasification. However, the orignal ‘channel planform can ouly rarely be observed (See Fig. 118 for an example), and faces are seldom diagnostic of planform type ata Bley 199. (Brierley 1999; Brierley and Hickin 1991) Thus inferred planform ‘was nota primary criterion, Although the great majority of examples were categorized without difcuty small number of channel bodies were problematic. For example, a few channel bodies atibuted to the Fxed-ver (ype hhave prominent lateral acreton sets, steep margins, and fills hat contain slump blocks of cemented paleosol material (Gibling and Rust 1990); they were attributed to fixed channels because resistant Danks precluded a freely meandering condition, as indicated by «high aggradation index (Fg. 2. Megafans and fixed river systems commonly include low W/T deposits of shallow sand-bed and sgravel-bed sircams (eg, North and Taylor 199) ‘Cassfcaton drew heavily upon the most extensive and completly preseved examples inthe literature. Channel bodies within a poorly nown subsurface setting or incomplete outcrop exposure may be dificult wo casi; this is especialy key for some fxed-chuanel bodies, for which information about the goographic setting is