SEDIMENTARY ARCHITECTURE IN MEANDERS OF A …basin.earth.ncu.edu.tw/download/courses/seminar_MSc/2010/0106-1... · SEDIMENTARY ARCHITECTURE IN MEANDERS OF A ... Sinuous deep-water

Journal of Sedimentary Research, 2010, v. 80, 852–866

Research Article

DOI: 10.2110/jsr.2010.078

SEDIMENTARY ARCHITECTURE IN MEANDERS OF A SUBMARINE CHANNEL: DETAILED STUDY OF THEPRESENT CONGO TURBIDITE CHANNEL (ZAIANGO PROJECT)

N. BABONNEAU,1 B. SAVOYE,*2 M. CREMER,3 AND M. BEZ4

1Universite de Brest, CNRS UMR 6538 Domaines Oceaniques, IUEM, Place Nicolas Copernic, 29280 Plouzane, France2IFREMER, GM-LES, BP70, 29280 Plouzane, France

3Universite Bordeaux I, CNRS UMR 5805 EPOC, Avenue des Facultes, 33405 Talence, France4TOTAL, CSTJF, Avenue Larribau, 64000 Pau, France

e-mail: [email protected]

ABSTRACT: Sinuous deep-water channels are recognized in most large deep-sea fans in the world. They present a particularinterest to oil companies, since they are significant hydrocarbon reservoirs in deep offshore environments. The understanding oftheir geometries and their internal sedimentary architecture is necessary to better characterize reservoir heterogeneity ofsinuous submarine channels. Therefore, numerous studies have been undertaken recently to better understand the behavior andsedimentary architecture of deep-water channels.

The aim of this paper is to present our results concerning the development of the meandering channel of the present Congoturbidite system (or Zaire turbidite system). The study is based on high-resolution data including multibeam bathymetry,seismic lines, echosounder profiles, high-resolution side-scan sonar images, and gravity cores, collected by IFREMER along thesubmarine Congo channel between 1994 and 2000, during Guiness and ZaiAngo surveys.

The present Congo turbidite channel is a long incised turbidite channel. It is presently active. It has been built gradually byprogradation of the distal depositional area. The most distal part of the channel is the youngest part and shows an immaturemorphology: the channel presents a low incision and a low sinuosity. In contrast, the upper part of the channel has undergone along evolutionary history. Its pathway is mature and complex, with numerous abandoned meanders visible in the morphology.

This paper presents evidence of progressive channel migration and meander development of the Congo channel. It describesand explains the presence of terraces inside the channel. The detailed characterization of channel morphology and migrationgeometry shows that the evolution of the channel path is very similar to fluvial meandering systems with (1) lateral meanderextension or growing, (2) downstream translation of the thalweg, and (3) meander cutoff.

Seismic and 3.5 kHz echosounder profiles show that the terraces, which are visible in the seafloor morphology, are not theimprints of incisional processes. Terraces are true depositional units infilling the channel. They are built during and after thelateral migration of the channel. They are composed of (1) point-bar deposits and (2) inner-levee deposits aggrading above thepoint bar deposits. Point-bar deposits are characterized by low-angle oblique reflectors forming deposits with a sigmoidalshape. They seem very similar to those observed in fluvial systems. The similarity between fluvial and turbidite point barssuggests that the basal part of the turbidity currents flowing in this channel can be considered as very similar to river flow.

With the high-resolution dataset collected in a present Congo turbidite channel, we provide a new description of the channelmorphology and evolution, at a ‘‘reservoir’’ scale, intermediate between outcrop observations and 2D and 3D seismic data. Thedetailed interpretation of intrachannel sedimentation, associated with lateral channel migration, also provides new data forinterpretation of flow dynamics in submarine meandering channels.

INTRODUCTION

Meandering deep-water channels have been observed at the surfaces ofmost of large deep-sea fans in the world: the Amazon Fan (Damuth et al.1983; Flood and Damuth 1987; Pirmez and Flood 1995), the MississippiFan (Kastens and Shor 1986; Pickering et al. 1986), the Bengal Fan(Hubscher et al. 1997; Schwenk et al. 2003), the Indus Fan (Kenyon et al.1995; Kolla and Coumes 1987), the Nile Fan (Loncke et al. 2002), and the

Rhone Fan (Droz and Bellaiche 1985; Torres et al. 1997). Submarinemeandering channels present a particular interest to oil companies, sincethey are significant hydrocarbon reservoirs in deep offshore environ-ments. The understanding of their geometries and their internalsedimentary architecture is necessary to better characterize reservoirheterogeneity. Therefore, numerous works have recently been carried outby both industrial and academic researchers to understand the behaviorand sedimentary architecture of sinuous deep-water channels, includingdifferent approaches: study of modern deep-water channels, outcropanalysis (Arnott 2007; Labourdette et al. 2008), 3D seismic studies (Abreuet al. 2003; Deptuck et al. 2007; Kolla 2007), and numerical and

* Bruno Savoye, the second author, died in august 2008. This article is

dedicated to his memory.

Copyright E 2010, SEPM (Society for Sedimentary Geology) 1527-1404/10/080-852/$03.00

experimental modeling (Corney et al. 2006; Das et al. 2004; Imran et al.2007; Kassem and Imran 2004; Keevil et al. 2006; Peakall et al. 2000;Straub et al. 2008).

The aim of this paper is to present our results concerning thedevelopment of the meandering channel of the present Congo turbiditechannel. Based on high-resolution data acquired on the presently activeCongo channel, we describe the detailed sedimentary architecture linkedto meander development and intrachannel deposition, and analyze thepossible channel behavior and flow processes. Finally, we propose amodel explaining the evolution of channel morphology through time,focused on meander development.

Our work results from the study of the Congo (or Zaire) modern deep-sea fan, conducted within the ZAIANGO project, in collaborationbetween IFREMER and TOTAL. The study of meandering morphologyof the present Congo turbidite channel was performed using a completeset of high-resolution data: bathymetry, backscatter sonar image,echosounder profiler, and gravity piston cores (Babonneau et al. 2002).The Congo channel is presently active. The meandering and incisedchannel morphology is dynamic and presently continues to evolve.Babonneau et al. (2002) described the channel morphology and thegeneral seismic structure of the turbidite system from the head of thecanyon down to the channel-mouth lobe. The work presented in thispaper focuses on the detailed architecture of several meanders with high-resolution tools to address the sedimentary structures linked to meandermigration and intrachannel deposits. The high-resolution dataset includeshigh-resolution bathymetric data (EM300), side-scan sonar, seismic andechosounder data, and piston cores. It provides a new description of theevolution of channel morphology, at a scale intermediate betweenoutcrop observations and seismic data.

Submarine meandering channels are often compared to fluvial channelsfor their similar planform shape. However, deep-sea channels are knownto exhibit several differences with rivers in terms of channel behavior asmigration, aggradation, incision, or levee development (Abreu et al. 2003;Kolla et al. 2007). Based on high-resolution data, the characterization ofthe Congo turbidite channel provides new information for comparisonwith fluvial channels, existence of lateral-accretion deposits (point-bardeposits), and interpretation of flow processes near the channel bottom ofturbidite channels.

GEOLOGICAL SETTING AND PREVIOUS WORK

The Congo–Angolan Margin and the Congo Deep-Sea Fan

The Congo Deep-Sea Fan is located on the Congo–Angolan margin,which is a mature, passive continental margin resulting from the EarlyCretaceous opening of the South Atlantic Ocean (130 Ma) (Jansen et al.1985; Marton et al. 2000). The depositional limits of the fan extend fromthe base of the continental slope, at the outlet of the Congo Canyon, tothe abyssal plain. The submarine Congo Canyon marks the limit betweenthe Congo and the Angola margins. It is strongly incised across the shelfand the continental slope (Ferry et al. 2004). The Congo Canyon isdirectly connected to the Congo River mouth (Fig. 1). The gravitycurrents generated in the canyon head transfer most of river sedimentdischarge towards the deep-sea fan at more than 3000 m water depth(Babonneau et al. 2002; Savoye et al. 2009).

The deep-sea fan displays a complex paleochannel network extendingover 300,000 km2 (Savoye et al. 2000). Most of the paleochannels still visibleat the present basin seafloor are Quaternary in age (Droz et al. 2003). Thepresent turbidity currents arriving from the Congo Canyon flow along a

FIG. 1.—Shaded bathymetry on the Congo deep-sea fan (EM12 multibeam data), showing the network of meandering submarine channels diverging from the canyon.The present turbidite system is underlined: the present channel in black, levees and lobe in gray.

DETAILED SEDIMENTARY ARCHITECTURE OF THE MEANDERS OF THE CONGO TURBIDITE CHANNEL 853J S R

unique active channel located in the central part of the fan (called the AxialFan by Droz et al. 2003) and end up at more than 5000 m depth.

The Active Turbidite System

The present Congo Channel is located in the axial part of Congo Fan,in the direct prolongation of the Congo Canyon. The Congo Channel iscontinuous and clearly visible on the bathymetric data. It is deeply incisedbelow the base of its associated levees (Babonneau et al. 2002). It extendsas far as 750 km from the African coast, and ends in a relatively flat areain the distal lobe complex (Fig. 1). The presently active Congo turbiditesystem is composed of four morphological zones (Babonneau et al. 2002):(1) an erosional submarine canyon, (2) a transitional zone from canyon to

channel–levee system corresponding to the upper-fan valley, (3) thechannel–levee system, and (4) the distal lobes. Each zone is characterizedby turbidity-flow behavior in terms of erosion, transport, and sedimen-tation.

The main peculiarity of the Congo turbidite system is the continuousactivity of turbidity currents through the last sea-level rise and during thepresent high sea-level stand (Savoye et al. 2009). This continuous activityis recorded in the cores taken along the present Congo turbidite system(channel floor, levees, and lobes). Core studies show high sedimentationrates and record turbidites during the Holocene (Van Weering and VanIperen 1984). Moreover, the activity of turbidity currents is documentedby submarine-cable breaks observed between 500 and 2300 m water depth(Heezen et al. 1964) and direct observations of gravity currents by a

FIG. 2.—A) Longitudinal depth profile of the Congo canyon and channel. B) Sinuosity of the present Congo thalweg. C) Simplified map of the present Congo turbiditesystem from the canyon to the distal lobe area.

854 N. BABONNEAU ET AL. J S R

mooring located in the present channel axis (Kripounoff et al. 2003).Good correlations in time are observed between major river floods andturbidity currents. The activity of the Congo turbidite system is thusdirectly linked to the Congo River behavior (Heezen et al. 1964).

The thalweg of the whole Congo canyon and channel displays asmooth, concave-up, longitudinal depth profile (Fig. 2A) and a remark-able deep incision of the channel floor. The origin of the strongentrenchment of the canyon and the thalweg incision across the wholeshelf are not fully understood. The backward erosion of the canyon headprobably generated the canyon’s vertical incision across the shelf, duringperiods of sea-level rise. The continuity of the turbidity-current activityprobably maintains the connection between the river estuary and thecanyon head (Babonneau et al. 2002).

Flow properties (flow velocity, turbulence, and erosive power) ofturbidity currents depends directly on the characteristics of the sedimentsupply, such as sand/mud ratio, grain-size distribution, concentration,and volume, and also depends on the slope gradient along the canyon(e.g., Pratson et al. 2000). The necessary conditions to erode are attainedin the Congo system, despite the relatively low sand/mud content ofsediment supply (estimated by the measurements of the sedimentdischarge in the Congo River estuary) (Moguedet 1988). We interpretthat the strong entrenchment of the upper part of the Congo channelsuppresses the overflow of turbidity currents and maintains the flow ofenergy. The deposition of the sediment load of the turbidity currentsoccurs mainly along the lower fan, lengthening the Congo Channel andinducing the progradation of the distal lobes (Babonneau et al. 2002).

Despite the strong entrenchment, the present Congo Channel is highlysinuous. The length of the Congo turbidite system, along the meanderingpathway, from the canyon head to the lobes, is about 1200 km for astraight-line distance from the estuary of 760 km. The total sinuosity ofthe channel is 1.58, and it can be called a ‘‘meandering’’ channel (Leopoldand Wolman 1957). The sinuosity of the Congo channel is maximumalong the upper-fan valley and along the upper part of the channel–leveesystem, where cut-off meanders are numerous (Babonneau et al. 2004).The sinuosity irregularly decreases down to the lower part of channel–levee system.

DATA

The results presented in this paper are based on the analysis of dataacquired during GUINESS (1992, 1993), and ZAIANGO (1998, 2000)cruises. The general bathymetric map and sonar backscatter images werecompiled from Simrad EM12 dual multibeam data. Complementarybathymetric data were collected with a Simrad EM300 dual multibeamsounder on narrow areas of the proximal part of the turbidite system.EM300 data have higher vertical and lateral resolution in water depthsless than 3500 m. At 3000 m water depth, the processing of EM12 dataprovides a final terrain numerical grid with 100 m horizontal grid spacingand a vertical resolution around 6 10 m, whereas EM300 data provides afinal grid with 50 m horizontal spacing and an average vertical resolutionaround 6 5 m.

Seismic-reflection surveys collected during the ZAIANGO cruises weremade using several configurations. Profile Z2-21 (presented in this paper)is a high-resolution seismic profile collected at a speed of 5 knots, using aseismic source made of six small mini GI airguns shooting at a rate of 10 s(mean frequency 120 Hz) and a 2500-m-long 96-channel streamer. Thevertical resolution of the profile is about 10 meters.

Side-scan sonar images and 3.5 kHz echo-sounder profiles, illustratingthis paper, were obtained with the SAR system (a deep-towed side-scansonar developed by IFREMER in 1982). The vehicle is towed at about100 m above the sea floor at 2 knots, and it is equipped with a very high-resolution side-scan sonar (170–190 kHz) and a 3.5 kHz echosounderprofiler. The SAR backscatter images provide information on sea-floor

texture and microtopography. Side-scan sonar images are firstlyinterpreted picking all the visible features. The interpretation ofechosounder profiles corresponding to the sonar images give comple-mentary information about acoustic facies (sediment type and texture),detailed morphology, and surface structure.

Kullenberg cores were collected in various part of the channel–leveesystem. They were analyzed using a Multi-Sensor Core Logger (GeotekLtd.), which provides gamma density, magnetic susceptibility, and P-wave velocity measurements. Visual descriptions distinguished the mainlithologies (clay, silty clay, silt, and fine sand), and vertical successions ofsedimentary facies. A series of 1-cm-thick sediment slabs were collectedfor each split core section and X-radiographed using a SCOPIX digital X-ray imaging system. Images of the sedimentary structures are enhancedafter X-ray image processing.

MORPHOLOGY, SEISMIC ARCHITECTURE, AND SEDIMENTARY FACIES

Sinuosity and Meander Morphology at Regional Scale

The morphological analysis of the Congo channel was based on thegeneral EM12 bathymetric map (Fig. 1). Quantitative parameters such aschannel relief, channel width, slope gradient along the thalweg, andsinuosity were analyzed by Babonneau et al. (2002). Figure 2 illustratesthe longitudinal depth profile of the Congo channel (Fig. 2A), thesinuosity calculated along channel segments 10 km long (Fig. 2B), andthe mapping of the Congo channel (Fig. 2C).

The sinuosity of the Congo channel changes along its path (Fig. 2B). Inthe canyon, the sinuosity of the thalweg is relatively low because of itsstrong entrenchment in a deep V-shaped valley. Meanders are small andrestricted by the steep canyon walls. Thalweg sinuosity increases along theupper-fan valley, due to the decrease of the confinement. Babonneau et al.(2002) interpreted that the evolution of the thalweg sinuosity along theupper-fan valley directly depends on the valley relief. There is a relationbetween lateral erosion capacity and thalweg entrenchment.

The Congo channel–levee system is divided into two parts. The upperpart of the channel–levee system is characterized by the highest sinuosityvalues (black sections in Fig. 2C), which also exhibits numerousabandoned meanders, easily identifiable in the sea-floor morphology.The lower part of the Congo channel–levee system is characterized bydecrease of sinuosity until the channel becomes a fairly straight path (lightgray sections). In this area, meanders have simple shape and noabandoned meanders are visible.

The longitudinal slope gradient of the present channel is smooth, with agradual downslope decrease (Fig. 2A) from 8 m/km in the head of thecanyon to 2m/km in lower channel–levee system. The downslope decreaseof sinuosity could be linked to the decrease in slope and flow energy.However, ancient channels of the Congo Fan, which extended farther intothe distal plain than the present system, were highly sinuous with thesame order of very low slope gradient. Their paths are highly meanderingat the depth of the present distal lobes. Thus, the relation between slopegradient and sinuosity is probably not as simple as in fluvial meanderingvalleys and remains unclear in turbidite environments (Babonneau et al.2002).

Details of the Channel–Levee Morphology from the EM12Backscattering Image

The EM12 backscattering image provides information concerning thesurface sedimentary cover. The backscattering image (Fig. 3A) is focusedon the channel–levee system. Contrasts in backscatter facies result mainlyfrom lithological and grain-size contrasts, variations in relief, andcompaction variations of the surface sediments.

Figure 3B is the interpreted map of the backscattering image combinedwith the analysis of 3.5 kHz echosounder profiles. Levee boundaries of


the present channel are underlined by black lines visible in both Figure 3Band 3C. Levees extend from 4 to 20 km on each side of the channel(average 10–12 km). Large sediment waves cover the whole levees, asdescribed by Migeon et al. (2004). The distribution of sediment waves isnot homogeneous. They are arranged in multiple fields. In the upper partof the channel–levee system, the sediment waves are well developed in theexternal area of sharp meanders. In the lower part of the channel–leveesystem, they are more continuous along quasi-rectilinear channelsegments (Migeon et al. 2004).

The sinuosity and the meander shape change along the channel. Sevenareas are distinguished by the morphological characteristics of meanders(Fig. 3):

N Areas 1 and 2 belong to the upper-fan valley. Area 1 is characterizedby a very irregular sinuosity, which increases downslope (Fig. 2B).The morphology of the meanders is complex (asymmetrical and withseveral orders of wavelength and amplitude). In numerous areas, themeanders are abandoned. Area 2 is fairly similar but with higheraverage values of sinuosity.

N Areas 3, 4, and 5 belong to the upper part of the channel–levee system.In Area 3, abandoned meanders are less frequent and less visible thanalong the upper-fan valley. Meander morphology in Area 3 is complexwith asymmetrical geometries. The sinuosity in Area 3 is very high andirregular. Area 4 is characterized by numerous abandoned meanders.The sinuosity in Area 4 is high (. 1.5). The meander morphology inArea 4 becomes simple and mainly symmetrical. In Area 5, thesinuosity has a lower value. The meanders in Area 5 are simple andsymmetrical, with only five loop cut-offs.

N Areas 6 and 7 belong to the lower part of the channel–levee system.Along this part of the channel, there is no abandoned meander. In

Area 6, the sinuosity is medium and relatively irregular. The meandermorphology in Area 6 is simple and asymmetrical. In Area 7, thesinuosity is low and decreases downslope. The meanders in Area 7 aresymmetrical with simple shape.

Meander morphology appears more complicated upslope (abandonedmeander, asymmetrical, and composite meander shapes) than downslope,in spite of a very low gradient change (less than 2m/km). The difference inmorphology probably is because the upstream reaches are older and haveundergone a longer evolutionary history.

Contribution of EM300 Data High-Resolution Bathymetry

The first bathymetric map of the Congo Fan was obtained using amultibeam EM12 Simrad sounder (in 1998). The quantitative morpho-logical analysis of the present Congo channel (Babonneau et al. 2002) wasbased on EM12 dataset. It expressed values of the main morphologicalcharacteristics and their general evolution along the channel pathway. Atthe scale of one or several meanders, the bathymetric contours weresmoothed and the channel section appears symmetrical (Fig. 4A).

The data collected with a multibeam EM300 Simrad sounder provideda higher-resolution image of the seafloor morphology (Fig. 4B).However, such high-resolution data were collected only along a 100-km-long segment of the upper Congo channel–levee system, from 3500 to4000 m water depth (location in Fig. 2). At this water depth, the EM300swath is limited to 3 km.

The EM300 contour map (Fig. 4B) shows small crescent-shaped planarareas in the channel banks, forming terraces. They are located mainly on theinner bank of the channel, but their location varies depending on the meandermorphology. Bathymetric sections across the channel indicate that thechannel section is almost asymmetrical. The channel section presents a steep

FIG. 3.—A) EM12 Backscatter sonar image along the channel–levee system. B) Interpretation of the sonar image. C) Simplified map of the present channel–leveesystem, grayscale of the channel floor corresponds to sinuosity values.


FIG. 4.—A) EM12 and B) EM300 bathymetry (contours every 20 m), showing a segment of the Congo channel (Location in Fig. 2). The EM300 bathymetric mapshows more morphological details and particularly small terraces created by channel migration.


bank in the outer part of the meander and a smoother inner bank. Innerbanks are frequently composed of one or several steps of small terraces. Thismorphology looks like a fluvial pattern across a meander, depicted by aneroded abrupt outer bank and a smooth opposite bank with point-baraccretion (Leopold and Wolman 1960; Schumm and Winkley 1994).

The asymmetric sections of the Congo channel, clearly visible on theEM300 map (Fig. 4B), suggest a progressive migration of the thalweg.The terraces probably correspond to the imprints of the morphologicalchange of the channel. Arrows (Fig. 4B) illustrate the surface dip of theterraces, which probably indicates the local migration direction. Differenttypes of migration are visible: (1) lateral extension of meander or meandergrowth, (2) downslope translation of the channel path, and (3) loop cut-off. Similar geometries are observed along meandering rivers, suggestinga very similar planform evolution of channels (Brice 1974).

The morphological imprint of lateral migration, responsible formeander growth (Zoom 1 of Fig. 4), often shows several steps in theinner bank. The steps suggest that meander development is not acontinuous process but occurs in several stages with changes in migrationdirection (Zoom 1 of Fig. 4). In contrast, the morphological imprints ofdownslope translation (Zoom 2 of Fig. 4) are small terraces with a quiteregular and low slope gradient.

The morphological features of the terraces strongly indicate lateralmigration of meanders. Although the processes are still unclear, the probablemorphological evolution of the turbidite channel is similar to those observedin rivers, and results from the erosion of the outer bank and depositionalprocesses in the opposite bank. The presence of terraces (single or stepped

terraces) could indicate a vertical incision of the channel coeval with lateralmigration. However, these terraces can also correspond to sedimentarydeposits inside the channel. It is necessary to study seismic sections tointerpret the possible vertical evolution of the channel floor through time.

Morphology and Surface Sedimentary Structures Imaged by theSAR System

High-resolution side-scan sonar images and 3.5 kHz echosounderprofiles were taken on the channel–levee system using the deep-towedSAR system. Figure 5 shows two SAR images and their related 3.5 kHzprofiles.

Profile and image G2Z07 (Fig. 5) are located in the area covered byEM300 data (location in Figs. 1 and 4B). The channel is 200 m deep andmore than 500 m wide. The SAR image G2Z07 shows a small flat terracebordering the thalweg. The terrace is located upslope of the mainmeander and is interpreted to reflect a downslope thalweg translation.The terrace surface contains several converging lineaments oriented in theslope direction that correspond to erosional features (furrows). In the3.5 kHz echosounder profile G2Z07 (Fig. 5), the acoustic facies coveringthe terrace is quite similar to the levee facies (bedded acoustic facies), butthe penetration is poorer than in the levee crest. This slight contrast ofpenetration may indicate coarser or more consolidated deposits on theterrace than on the levee crest.

Profile and image G2Z-04 (Fig. 5) are located upstream of the G2Z07profile (see location in Fig. 1). The channel is 220 m deep and the channel

FIG. 5.—SAR images, map interpretations, and 3.5 kHz echosounder profiles across the present meandering channel: G2Z-04 and G2Z07 (Location in Fig. 2),showing morphological features (terraces) created by the thalweg migration.


floor is 250 m wide. The image covers a well-developed tight meanderbend. The profile follows the axis of the meander loop and shows theabrupt outer wall (northern bank) and stepped structures in the internalbank of the meander. The 3.5 kHz echosounder profile shows beddedacoustic facies on top of the outer flank, typical of levee deposits. Thesedimentary deposits in the stepped structures are also characterized by abedded acoustic facies with good signal penetration. Acousticalcharacteristics appear relatively similar to levee acoustic facies, but thereare numerous oblique discontinuities, which are interpreted as smallfaults or sliding surfaces. The three steps of terrace seem to correspond tothree stages of meander growth with changes of migration direction (quitesimilar to the stepped terraces described in Fig. 4).

Seismic Structure across Meanders of the Present Channel

Most seismic data available across the Congo channel has too poorhorizontal and vertical resolution to define sedimentary structures insidethe channel. Moreover, the number of profiles crossing the presentchannel in an optimal position to observe meander migration is small.For these reasons, we firstly interpreted the sedimentary units observedinside the present channel axis as large slides of the channel banks, by

analogy with what was observed in other submarine canyons (Kenyon etal. 1995; Klaucke and Hesse 1996; Liu et al. 1993). Later, we collectedhigher-resolution data combining SAR and EM300 data. As we showedpreviously, the high-resolution data highlighted depositional terraces inthe channel axis in relation with channel migration and meanderformation. This new interpretation led us to reinterpret the initial seismicdata.

The seismic profile Z2-21 (Fig. 6) is one of the best profiles of theZAIANGO seismic data crossing the channel–levee system. It shows well-developed levees on both sides of the channel, above a basal complex withhigh-amplitude reflections (HARPs) and chaotic layers (Fig. 6). Inside thechannel, a large sediment accumulation built a terrace in the northernchannel bank. At the base of the accumulation, we can distinguish obliqueto sigmoidal high-amplitude reflectors corresponding to the lowersedimentary unit (underlined by dashed line in Fig. 6). The reflectors areinclined toward the south, forming downlap geometry toward the presentchannel axis. The reflectors underlined by dashed lines are laterallycorrelated and probably correspond to paleo-bathymetric surfaces.Considering the high vertical exaggeration of the seismic profile, thereflectors show angles less than 3u. The lower sedimentary unit is coveredby lower-amplitude and subhorizontal reflectors forming the upper part of

FIG. 6.—High-resolution seismic profile Z2-21 across the present channel–levee system (location in Fig. 2), showing a small terrace (levees in light gray, HARP’s indark gray, and chaotic layers in medium gray).


the terrace (above the upper dashed line). The lower unit is interpreted aspoint-bar structures, synchronous with the lateral migration of the channel.In contrast, the upper unit has a draping shape and low-amplitudediscontinuous reflections. The seismic facies is quite similar to the leveeseismic facies. Based on its draping geometry, it was deposited after thechannel-migration stage. The base of the well-defined concave-up high-amplitude reflectors, visible under the terrace, is located at about the samelevel as the present channel floor. They can correspond to the ancientlocations of the channel axis through time. If this interpretation is correct,following an incisional phase, the elevation of the channel floor remainedstable during lateral migration with only a small aggradation component.

Cores Collected in the GUINESS Area: Channel Floor,Channel Banks, and Levees

The GUINESS area corresponds to a short channel segment of theupper part of Congo channel–levee system (Fig. 1). In the GUINESSarea, the channel floor is about 200 m below the levee crests. The channelpath is highly meandering (Fig. 2) and shows one well-developedmeander and one well-shaped meander cutoff (Fig. 7). A set of sevenpiston cores were collected along a transect crossing the inner bank of thewell-developed meander (core locations in Fig. 7). Two cores are locatedon the outer levees (KG2Z-02 and KG2Z-03), two cores sampled the

FIG. 7.—Core descriptions across the channel in Guiness Area, see the location of cores in the bathymetric map and in the bathymetric profile: KG2Z-02 and 03 arelocated in the external levee, KZR19 and 21 sample the channel floor, and KZR17, 18, and 22 were collected in the inner levee.


channel floor (KZR19 and KZR21), and three others were taken atvarious levels in the terrace located the inner bank of the meander(KZR17, KZR18, and KZR 22). Unfortunately, no high-resolutionseismic profile is available along the core transect, and this limits ourinterpretation of the detailed sedimentary architecture in the GUINESSarea.

Cores Collected in the Levees: KG2Z-03 and KG2Z-02

KG2Z02 is located on the southern levee of the Congo channel at awater depth of 3490 m and 200 m above the channel floor (Fig. 7). Thecore consists of a succession of fine-grained turbidite deposits, dominatedby silt, silty clay, and clay. There is no sandy interval. The silty intervalsare less than 5% of the core (95% clay and silty clay) and are concentratedat the base of the core in the lowest 5 m. The bases of turbidite depositsare sharp, or gradual where there is intense bioturbation. Sand beds arenormally graded. Silt and silty-clay intervals are characterized by a darkcolor, induced by the high content of organic matter (small vegetal debris

and oxides). At the core location, the levee was fed by the overflows fromthe abandoned meander when it was active. Present sedimentation isprobably due to turbidity currents overflowing from the upstreammeander.

KG2Z03 is located on the northern levee at 3530 m water depth(Fig. 7). This area is fed by sediments provided by turbidity currentsoverflowing the outer bank of the well-developed meander. The coreconsists of a succession of fine-grained turbidite deposits, dominated bysilt, silty clay, and clay. There is no sand interval. The silty intervals areless than 2% of the core (98% of clay and silty clay). The verticalsuccession shows a fining-upward evolution with more frequent siltysequences at the base of the core. Photographs and X-radiograph imagesof KG2Z03 (Fig. 8) show examples of the fine-grained turbidite deposits,with a vague basal contact and a lower silty interval (several centimetersthick). These deposits are structureless and highly bioturbated. Silt andsilty-clay intervals are also characterized by a dark color, induced by thehigh content of organic matter (vegetal debris and oxides).

FIG. 8.—Cores photographs and X-ray images, showing the main sedimentary facies: massive sand for the channel floor and a fine-grained sequence typical oflevee facies.


Sedimentary Facies of the Terrace Deposits: KZR17, KZR18, and KZR22

Three cores were collected in the terrace located in the inner bank of themeander. They are located respectively at 3526 m, 3577 m, and 3660 mwater depth. In other terms, KZR22 is located at 40 m above the presentchannel floor, KZR18 is 120 m above the channel floor, and KZR17 is170 m above the channel floor (50 m below the levee crest).

The sedimentary facies are quite similar to levee facies, with successionsof fine-grained turbidite deposits (silt and silty clay, normally gradeddeposits). However, content of silt and fine sand is higher than in corestaken in the levees. These three cores show fining-upward successions. Siltand silty-clay intervals are characterized by a dark color, induced by thehigh content of organic matter (vegetal debris). The clayey intervals arebrown and bioturbated. Figure 8 shows photographs and X-ray imagesof selected core sections. These sections illustrate the most representativefacies of the cores with the best-preserved structures. They correspond tosections 70–80 cm long, located in the uppermost 5 meters at the top ofthe cores.

KZR17 is composed of a succession of fine-grained turbidite deposits,dominated by silt, silty clay, and clay. There are no sand layers in thiscore. The silt content is about 3–5% (95% of clay and silty clay). Turbiditedeposits present sharp bases and normal grading (Fig. 8).

KZR18 is also characterized by a succession of fine-grained turbiditedeposits, dominated by silt, silty clay, and clay. Sandy layers are presentbut sparse and thinner than 5 mm (sand content lower than 1%). Thesilt content is about 10–15% (85% of clay and silty clay). Turbiditedeposits present sharp bases and normal grading. The thickest siltyintervals present horizontal lamination. The vertical stacking ofturbidite deposits shows fining-upward sequences of three or fourturbidite deposits (Fig. 8). Several intervals of KZR18 are locallyunstructured and liquefied because of the presence of gas (methane) inthe sediment. The sediment collected in the Congo fan is very rich inorganic matter, and particularly vegetal debris directly coming from theCongo River. Locally, the sediment is partly saturated in methane. Thepressure changes during the coring induce sediment liquefaction anddeformation.

KZR22 consists of a succession of fine-grained turbidite deposits,dominated by silt and silty clay. Sandy layers are visible but are sparseand characterized by thin intervals about 1 to 2 cm thick. The sandcontent is about 3% and the silt content is about 40–45% (55% of clay andsilty clay). Turbidite deposits present sharp bases, variable thicknesses,normal grading, and frequent horizontal lamination in the thickest siltyintervals (visible in the X-radiograph image in Fig. 8).

A gradual evolution of sedimentary facies is visible from KZR17 toKZR22, with more abundant and thicker silt beds seen in the cores closerto the channel axis.

Cores in the Channel Floor: KZR19 and KZR21

KZR19 and KZR21 were collected in the channel axis. Both consist ofmassive, structureless medium to fine sand, and locally contain vegetaldebris and mud clasts. The sand content is about 95 to 100%.

KZR21 is characterized by higher mud content than KZR19 (poorlysorted sand), perhaps due to its proximity to the outer channel bank. Thispart of the channel floor is probably occasionally laterally fed bysediments coming from outer-bank erosion and instabilities (dominantlymuddy blocks and debris). The absence of structure in visual descriptionand X-radiograph data is difficult to interpret. It could be partly theresult of coring disturbance (sand liquefaction). In some core sectionshowever, the lack of obvious coring disturbance (flow in) suggests thepresence of amalgamated massive sand layers. It indicates significantreworking of the sediment deposited in the channel floor (erosion andtraction processes) at the bases of the turbidity currents.

DISCUSSION

The high-resolution bathymetric, seismic, and core data presented inthis paper illustrates the meandering morphology and sedimentaryarchitecture of the Congo turbidite channel in unprecedented detail.Integrating the information provided by this dataset, we propose a modelfor the channel evolution and for the development of the meanders andassociated terraces.

Longitudinal Evolution of the Sinuosity along the Present Congo Channel

The longitudinal evolution of the sinuosity along the Congo channelshows a high and nonlinear variability (Fig. 2). The meanderingmorphology of the Congo canyon and submarine valley is probably notcontrolled by one simple rule, such as the link between slope gradient andsinuosity value.

The morphology of the upper part of the channel–levee system is morecomplicated (abandoned meander, asymmetrical and composite meandershapes) than the lower part (Figs. 2, 3). Channel gradients are very low,evolving from 4 m/km to 2 m/km. The complete study of the Congo deep-sea fan and its seismic structure reveals that the active channel was builtprogressively basinward. The distal part of the active channel is younger(approximately 10 ka) than the upper part of channel, which is estimatedto be older than 100 ka (Babonneau et al. 2002). Therefore theunderstanding of the meandering pattern of the Congo submarine valleyis not simple, because the morphology of the different channel segmentscorresponds to different scales of evolution with probable changes in flowprocesses in time and in space. The distal part of the channel being young,it is relatively immature. It is less sinuous than the upper channelsegments (Fig. 2) and probably displays how the Congo channel lookedin its initial form.

The observation that sinuosity in many submarine channels increasesthrough time has been documented by several previous studies (Deptucket al. 2003; Deptuck et al. 2007; Kolla 2007; Kolla et al. 2001; Peakall etal. 2000). In our study, the increase of irregularity and the complexity ofmeandering shape (abandoned meander, asymmetrical and compositemeander shapes, visible in Fig. 3) toward the upper part of the channel–levee system is probably directly induced by the maturity of the channel,reflecting a longer evolutionary history with the superposition of severalmeander generations.

Evidence of Lateral Migration of the Thalweg

Data acquired on the Congo channel clearly show a gradual lateralmigration of the thalweg and meander development. The EM300bathymetric map illustrates the morphological imprints of the migration(Fig. 3B), underlined by the presence of terraces inside the channel. InFigure 3, a great diversity of migration types are visible: lateral extensionof meander, downstream translation, and loop cutoffs.

We define two types of terrace: (1) single terraces (Fig. 4, Zoom 2), and(2) stepped terraces (Fig. 4, Zoom 1). The two types of terrace are alsoimaged with the SAR system and 3.5 kHz echosounder profiles inFigure 5 (single terrace in SAR image G2Z-07 and stepped terraces inG2Z04). The single terraces are characterized by a nearly flat surfacegently inclined downstream toward the channel axis (Fig. 5, G2Z-07). Weinterpret that the slope direction of the terrace indicates the maindirection of channel migration. These terraces are the imprints of theprogressive downstream translation of the channel.

The stepped terraces are characterized by several levels of successivesmall terraces (Fig. 5, G2Z-04), with various directions and values ofgradient. We also interpret that the slope directions of the steppedterraces indicate the directions of channel migration. These terraces arelocated in the inner banks of well-developed meanders and correspond tothe different stages of meander development. In Figure 5 (image and


section G2Z-04), the first stages of meander development are illustratedby the upper terraces and show the lateral migration of the meander.Then, the channel migration changes in direction and turns about 90u inthe lower terrace, toward the downstream direction of the general slope.The changes in migration do not occur progressively but by successivestages forming the stepped terraces. The change of migration directioncould be linked to upstream morphological changes or changes in flowdynamics.

In the high-resolution morphology, we describe two types of terracecorresponding to two types of channel migration: lateral extension ofmeander, and downstream translation. The other type of channelmigration visible in the morphology is loop cutoffs (Figs. 3, 4). Thesemigration directions and forms are very similar to patterns and processesobserved in fluvial environments (Brice 1974).

What are the Terraces? Incisional Imprints or Depositional Units?

The interpretation of seismic sections shows that the present channel isover-incised (Babonneau et al. 2002). The present channel floor is locatedbelow the base of the levee deposits. Profile Z2-21 shows a well-developedterrace, which is composed of a thick accumulation of sediment inside thechannel (Fig. 6). This seismic structure suggests an infill during and afterthe meander migration. It also suggests that the channel was alreadydeeply incised before the lateral migration and the formation of theterraces (Fig. 6).

In the other seismic sections available across the present Congochannel, all terraces are characterized by thick accumulations filling theincised channel. In none of our seismic profiles have we observed terracesinduced by the progressive incision of the channel, simultaneously with itslateral migration. It suggests that the vertical incision of the channeloccurred very early and probably without significant lateral migration, asFigure 9 illustrates (stages 1 and 2).

The base of the terrace accumulation (Fig. 6) is located at a level closeto the present channel floor (or below). It means the channel migration

occurred with a channel floor level constant or with a little aggradation.These data suggest that lateral migration and meander development occurmainly after incision for a relatively stable channel floor, maybe when theturbidity currents are sufficiently channelized and confined to erode thechannel bank (Fig. 9, stage 3).

Profile Z2-21 (Fig. 6) also shows the internal structure of the sedimentaccumulation corresponding to the terrace. The accumulation is dividedinto two units: the lower unit is characterized by sigmoidal high-amplitude reflectors, and the upper unit is a nearly constant-thicknessunit draping pre-existing reflections (Fig. 6). The obliquity of thesigmoidal reflectors is low and amplified by the vertical exaggeration ofseismic sections (apparent dip less than 3u). However, this obliquity isprobably more important in a section perpendicular to the channel (Z2-21is oblique to the channel and probably not ideally oriented to observe thestructure). The complete geometry of the lower unit, underlined bydashed lines, suggests a sigmoidal shape, typical of lateral-accretiondeposits.

In rivers, point bars correspond to sandy accumulations, deposited onthe inner banks of fluvial meanders. These sediment bodies arecharacterized by large oblique structures, which migrate toward the riveraxis during the gradual lateral migration of the thalweg. There is erosionof the outer bank and accretion of the inner bank (Bridge 1992). In ourdata, the lower unit of the terrace identified in Z2-21 can be interpreted aspoint-bar deposits, with a shape very similar to those in fluvial systems.

Until now, true point-bar structures in large meandering turbiditechannels were not easy to identify on seismic data (Deptuck et al. 2003;Deptuck et al. 2007; Kolla 2007; Kolla et al. 2001; Kolla et al. 2007).Abreu et al. (2003) showed significant oblique structures, which theyinterpreted as lateral-accretion packages, named LAPs. With our dataalong the present Congo channel, we probably image the same type oflateral-accretion deposits. The presence of point-bar deposits demon-strates that the meanders of the Congo submarine channel are migratingin ways analogous to those in rivers, in agreement with previous work by

FIG. 9.—Model illustrating the initiation and evolution of meanders of the Congo submarine channel.


Kolla et al. (2007) and Abreu et al. (2003). It implies that the basal flow ofthe turbidity currents occurring in the Congo channel behaves in afashion similar to that of river flow (Fig. 9, Stage 3) (Imran et al. 2007).

The upper unit of the terrace of the seismic section Z2-21 (Fig. 6)drapes the point-bar deposits. It is characterized by a seismic facies withlow-amplitude reflectors, very similar to levee seismic facies. It probablycorresponds to the same type of deposit observed at the tops of theterraces imaged by the SAR system (Fig. 5). In 3.5 kHz sections (Fig. 5),the terrace deposits have also the same characteristics as levee deposits,with good penetration and bedded acoustic facies. Cores collected at thetops of terraces show the sedimentary facies covering the terraces(KZR17, KZR18, and KZR22 in Figs. 7 and 8). The sedimentary faciesare successions of fine-grained turbidite deposits, typical of overflowdeposits (levee facies). In core data, there are more silty sequences in thecores collected in the terrace than on the true external levees (dominatedby muddy turbidite deposits). The draping upper unit is probably built bythe overflow processes of the turbidity currents on the terrace and can beconsidered as inner-levee deposits (Fig. 9, stage 4). In the cores collectedin the terrace, frequency and thickness of silty and sandy layers increasetoward the channel axis. It correlates with decrease of the terrace heightand distance from the active channel floor. It is probably due to verticalgrain-size distribution in turbidity currents (Pirmez and Imran 2003).

Inner-levee deposits are also observed in buried channels imaged by 3Dsubsurface seismic data (Deptuck et al. 2007; Kolla 2007). They are alsoobserved in outcrops (Elliot 2000; Labourdette et al. 2008). It has beendescribed for different channel evolution: vertical aggradation with alateral migration component, and lateral migration. For the Congochannel, the development of inner levees is identified for lateral migrationwith a stable channel level. The inner-levees presented in this paper aggradeabove point-bar deposits after the lateral migration of the channel.

In the high-resolution data (morphology and SAR images), weobserved that the largest terraces show several steps (Figs. 4 and 5,G2Z-04), with changes in the probable migration direction. Thisobservation implies that the terrace formation often occurred bysuccessive stages (Fig. 9, stage 4): erosional stages inducing rapid lateralmigration of the channel alternating with more depositional stagesallowing inner-levee deposits with a relatively stable channel floor. Thisbehavior could be explained by cyclic changes of flow characteristics, withperiods of large and erosional currents alternating with periods of smalland low-energy currents. This variations could be correlated with long-term fluctuations of the river discharge and climatic changes, but there isno accurate age control on the evolution of the terraces.

Flow Behavior

In this paper, we illustrate with high-resolution data that turbiditycurrents generate channel morphology and sedimentary architecture verysimilar to river flows. Turbidity currents are offset in the meander outerbank and erode it. The highest and most dilute part of the turbiditycurrents partly overflows the outer bank onto the levee crest. Most of thefine suspended sediment remains in the channel and feeds the terraces(Fig. 9, stage 4).

Physical experiments and numerical models provide new informationabout flow dynamics and physical factors controlling the evolution ofchannel morphology with the three main components: incision, aggrada-tion, and lateral migration (Kassem and Imran 2004; Keevil et al. 2006;Peakall et al. 2007; Straub et al. 2008). Some authors have recently shownthat variations in basal flow and basal shear stress influence theformation, development, and facies of inner-bend accumulations insubmarine channels, and ultimately may affect the evolution of thechannel (Corney et al. 2006; Imran et al. 2007). In the Congo channel, thelow aggradation of the channel floor is probably maintained by highbasal shear stress, induced by a thin high-concentration basal flow layer.

The basal flow reworks sediments of the channel floor (the sandy depositscollected in KZR19 and KZR21 in Fig. 7) and erodes the outer banks ofmeanders. Above the basal layer, a thick low-concentration plume feedsfine-grained sediment to the inner banks of the meanders. The energy andvelocity of the plume are probably relatively low. The low-density part ofcurrents overflows and deposits fine-grained sediments on the outer levee(outer bank of meanders) and inside the channel feeding the inner levees.

The work presented in this paper provides new information about themeander development of an incised channel. It provides useful constraintsfor fundamental flume tank and numerical modeling work done tounderstand flow dynamics through sinuous channels (Kassem and Imran2004; Keevil et al. 2006; Peakall et al. 2007; Straub et al. 2008). Thepresence of structures very similar to point-bar deposits in submarinesinuous channels supports that helical flow fields in submarine settingshave the same direction as in rivers, in agreement with the numericalmodel of Das et al. (2004) and Kassem and Imran (2004).

CONCLUSIONS

The present Congo turbidite channel is a long incised submarinechannel, which is presently active. It is highly meandering, and itssinuosity is highly variable all along its pathway. The Congo channel hasbeen built gradually by progradation. Its distal extremity is relativelyyoung and reflects the immature stage of the channel (low sinuosity). Incontrast, the upper part of the channel has undergone a longerevolutionary history. Its pathway is complex and highly sinuous, withnumerous abandoned meanders visible in the morphology.

The data available on the Congo active submarine channel showevidence of progressive channel migration and meander development. Adetailed characterization of channel morphology indicates that theevolution of the channel path is very similar to that of fluvial meanderingsystems. We identify three main types of channel migration: (1) lateralmeander extension leading to the formation of stepped terraces in theinner bank of the meander, (2) downstream translation of the thalwegforming single terraces, and (3) meander cutoff.

Seismic and 3.5 kHz echosounder profiles show that terraces are not theimprints of incisional processes but are true depositional units infilling thechannel. Terraces are built during and after the lateral migration of thechannel and are composed of (1) point-bar deposits and (2) inner-leveedeposits aggrading above point-bar units. Point bars are characterized bylow-angle oblique reflectors forming sigmoidal-shaped deposits, verysimilar to those observed in fluvial systems. The channel evolution(behavior and sedimentary architecture) shows many similarities with riversand indicates that the basal parts of the turbidity currents flowing in theCongo submarine channel can be considered to be very similar to river flow.

With these interpretations, the morphological evolution of the Congochannel can be better constrained. We suggest that the initial channelprobably incises vertically without significant lateral migration. After theincision stage, its level becomes stable and its pathway begins to migratelaterally and to develop meanders. The formation of point-bar depositsoccurs during the migration. Later, terraces aggrade, fed by the overflowsof turbidity currents, and form inner-levee deposits.

The high-resolution dataset collected in the present Congo turbiditechannel provides a new description of the channel morphology and evolution,at a scale intermediate between outcrop observations and 2D and 3D seismicdata. It provides detailed interpretation of intra channel sedimentationassociated with lateral channel migration and also provides new data forinterpretation of flow dynamics in submarine meandering channels.

ACKNOWLEDGMENTS

This work was initiated in the ZAIANGO project, which was financiallysupported by TOTAL and IFREMER. N. Babonneau acknowledgesIFREMER for Ph.D. financial support and the opportunity to work on the


ZAIANGO and GUINESS data set. We thank the Genavir technician teams,crew members of the IFREMER Ships for the ZAIANGO and GUINESScruises, and in particular we thank people in charge of data processing: B.Loubrieu, A. Normand, E. LeDrezen, H. Nouze, and R. Apprioual fromIFREMER. Special thanks to all the people collaborating in the ZAIANGOproject and to D.J.W. Piper for discussing with us some of our results.

The authors also thank the two reviewers D. Pyles and K. Straub, and theassociate editor C. Pirmez for their comments and suggestions thatsignificantly improved the paper.

This paper is dedicated to the memory of Bruno Savoye, our colleague andfriend. Bruno’s death is an enormous loss for marine geosciences, for hisfamily, friends, colleagues, and students. Our sympathy goes to Babette, hiswife, and to Charlotte and Marion, his daughters.

REFERENCES

ABREU, V., SULLIVAN, M., PIRMEZ, C., AND MOHRIG, D., 2003, Lateral accretionpackages (LAPs): an important reservoir element in deep water sinuous channels:Marine and Petroleum Geology, v. 20, p. 631.

ARNOTT, R.W.C., 2007, Stratal architecture and origin of lateral accretion deposits(LADs) and conterminous inner-bank levee deposits in a base-of-slope sinuouschannel, lower Isaac Formation (Neoproterozoic), East-Central British Columbia,Canada: Marine and Petroleum Geology, v. 24, p. 515.

BABONNEAU, N., SAVOYE, B., CREMER, M., AND KLEIN, B., 2002, Morphology andArchitecture of the present canyon and channel system of the Zaire deep-sea fan:Marine and Petroleum Geology, v. 19, p. 445–467.

BABONNEAU, N., SAVOYE, B., CREMER, M., AND BEZ, M., 2004, Multiple terraces withinthe deep incised Zaire canyon: are they confined levees? in Lomas, S.A., and Joseph,P., eds., Turbidite Sedimentation in Confined Systems; Geological Society of London,Special Publication, v. 222, p. 91–114.

BRICE, J.C., 1974, Evolution of meander loops: Geological Society of America, Bulletin,v. 85, p. 581–86.

BRIDGE, J.S., 1992, A revised model for water flow, sediment transport, bed topographyand grain size sorting in natural river bends: Water Resources Research, v. 28, p.999–1013.

CORNEY, R.K.T., PEAKALL, J., PARSONS, D.R., ELLIOTT, L., AMOS, K.J., BEST, J.L.,KEEVIL, G.M., AND INGHAM, D.B., 2006, The orientation of helical flow in curvedchannels: Sedimentology, v. 53, p. 249–257.

DAMUTH, J.E., KOLLA, V., FLOOD, R.D., KOWSMANN, R.O., MONTEIRO, M.C., GORINI,M.A., PALMA, J.J.C., AND BELDERSON, R.H., 1983, Distributary channel meanderingand bifurcation patterns on the Amazon deep-sea fan as revealed by long-range side-scan sonar (GLORIA): Geology, v. 11, p. 94–98.

DAS, H.S., IMRAN, J., PIRMEZ, C., AND MOHRIG, D., 2004, Numerical modeling of flowand bed evolution in meandering submarine channels: Journal of GeophysicalResearch, v. 109, p. 1–17.

DEPTUCK, M.E., STEFFENS, G.S., BARTON, M., AND PIRMEZ, C., 2003, Architecture andevolution of upper fan channel-belts on the Niger Delta slope and in the Arabian Sea:Marine and Petroleum Geology, v. 20, p. 649.

DEPTUCK, M.E., SYLVESTER, Z., PIRMEZ, C., AND O’BYRNE, C., 2007, Migration–aggradation history and 3-D seismic geomorphology of submarine channels in thePleistocene Benin–Major Canyon, western Niger Delta slope: Marine and PetroleumGeology, v. 24, p. 406.

DROZ, L., AND BELLAICHE, G., 1985, Rhone Deep-Sea Fan: Morphostructure andgrowth pattern: American Association of Petroleum Geologists, Bulletin, v. 69, p.460–479.

DROZ, L., MARSSET, T., ONDREAS, H., LOPEZ, M., SAVOYE, B., AND SPY-ANDERSON, F.L.,2003, Architecture of an active mud-rich turbidite system: The Zaire Fan (Congo–Angola margin, southeast Atlantic). Results from ZAIANGO 1 and 2 cruises:American Association of Petroleum Geologists, Bulletin, v. 87, p. 1145–1168.

ELLIOTT, T., 2000, Depositional architecture of a sand-rich, channelized turbiditesystem: the Upper Carboniferous Ross sandstone Formation, Western Ireland: GulfCoast Section, SEPM Foundation, 20th Annual Bob F. Perkins Research Conference,Deep-Water Reservoirs of the World, p. 342–373.

FERRY, J.-N., BABONNEAU, N., MULDER, T., PARIZE, O., AND RAILLARD, S., 2004,Morphogenesis of Congo submarine canyon and valley: implications about thetheories of the canyon’s formation: Geodynamica Acta, v. 17, p. 241–251.

FLOOD, R.D., AND DAMUTH, J.E., 1987, Quantitative characteristics of sinuousdistributary channels on the Amazon Deep-Sea Fan: Geological Society of America,Bulletin, v. 98, p. 728–738.

HEEZEN, B.C., MENZIES, R.J., SCHNEIDER, E.D., EWING, W.M., AND GRANELLI, N.C.L.,1964, Congo Submarine Canyon: American Association of Petroleum Geologists,Bulletin, v. 48, p. 1126–1149.

HUBSCHER, C., SPIESS, V., BREITZKE, M., AND WEBER, M.E., 1997, The youngest channel–levee system of the Bengal Fan: results from digital sediment echosounder data:Marine Geology, v. 141, p. 125–145.

IMRAN, J., ISLAM, M.A., HUANG, H., KASSEM, A., DICKERSON, J., PIRMEZ, C., AND

PARKER, G., 2007, Helical flow couplets in submarine gravity underflows: Geology,v. 35, p. 659–662.

JANSEN, J.H.F., VAN WEERING, T.C.E., GIELES, R., AND VAN IPRESEN, J., 1985, Middleand late Quaternary oceanography and climatology of the Zaire–Congo fan and theadjacent eastern Angola Basin: Netherlands Journal of Sea Research, v. 17, p.201–249.

KASSEM, A., AND IMRAN, J., 2004, Three-dimensional modeling of density current: II.Flow in sinuous confined and unconfined channel: Journal of Hydraulic Research,v. 42, p. 591–602.

KASTENS, K.A., AND SHOR, A.N., 1986, Evolution of a channel meander on theMississippi deep-sea fan: Marine Geology, v. 71, p. 165–175.

KEEVIL, G.M., PEAKALL, J., BEST, J.L., AND AMOS, K.J., 2006, Flow structure in sinuoussubmarine channels: Velocity and turbulence structure of an experimental submarinechannel: Marine Geology, v. 229, 241 p.

KENYON, N.H., AMIR, A., AND CRAMP, A., 1995, Geometry of the youngersediment bodies of the Indus Fan, in Pickering, K.T., Hiscott, R.N., Kenyon,N.H., Ricci Lucchi, F., and Smith, R.D.A., eds., Atlas of Deep Water Environ-ments: Architectural style in Turbidite Systems: London, Chapman and Hall,p. 89–93.

KLAUCKE, I., AND HESSE, R., 1996, Fluvial features in the deep-sea: new insights from theglacigenic submarine drainage system of the Northwest Atlantic Mid-Ocean Channelin the Labrador Sea: Sedimentary Geology, v. 106, p. 223–234.

KOLLA, V., 2007, A review of sinuous channel avulsion patterns in some major deep-seafans and factors controlling them: Marine and Petroleum Geology, v. 24, p. 450.

KOLLA, V., AND COUMES, F., 1987, Morphology, internal structure, seismic stratigraphyand sedimentation of the Indus Fan: American Association of Petroleum Geologists,Bulletin, v. 71, p. 650–677.

KOLLA, V., BOURGES, P., URRUTY, J.M., AND SAFA, P., 2001, Evolution of deep-waterTertiary sinuous channels offshore Angola (West Africa) and implications forreservoir architecture: American Association of Petroleum Geologists, Bulletin, v. 85,p. 1373–1405.

KOLLA, V., POSAMENTIER, H.W., AND WOOD, L.J., 2007, Deep-water and fluvial sinuouschannels—characteristics, similarities and dissimilarities, and modes of formation:Marine and Petroleum Geology, v. 24, p. 388.

KRIPOUNOFF, A., VANGRIESHEIM, A., BABONNEAU, N., CRASSOUS, P., DENNIELOU, B., AND

SAVOYE, B., 2003, Direct observation of intense turbidity activity in the Zairesubmarine Valley by 4000 m depth: Marine Geology, v. 194, p. 151–158.

LABOURDETTE, R., CRUMEYROLLE, P., AND REMACHA, E., 2008, Characterisation ofdynamic flow patterns in turbidite reservoirs using 3D outcrop analogues: Example ofthe Eocene Morillo turbidite system (south-central Pyrenees, Spain): Marine andPetroleum Geology, v. 25, p. 255.

LEOPOLD, L.B., AND WOLMAN, M.G., 1957, River channel pattern: braided, meandering,and straight: U.S. Geological Survey, Professional Paper 282-B, p. 39–84.

LEOPOLD, L.B., AND WOLMAN, M.G., 1960, River meanders: Geological Society ofAmerica, Bulletin, v. 71, p. 769–794.

LIU, C.S., LUNDBERG, N., REED, D.L., AND HUANG, Y.L., 1993, Morphological andseismic characteristics of the Kaoping Submarine Canyon: Marine Geology, v. 111, p.93–108.

LONCKE, L., GAULLIER, V., BELLAICHE, G., AND MASCLE, J., 2002, Recent depositionalpatterns of the Nile deep-sea fan from echo-character mapping: American Associationof Petroleum Geologists, Bulletin, v. 86, p. 1165–1186.

MARTON, L.G., TARI, G.C., AND LEHMANN, C.T., 2000, Evolution of the Angola passivemargin, West Africa, with emphasis on post-salt structural styles, in Moriak, W., andTalwani, M., eds., Atlantic Rifts and Continental Margins: Washington, D.C.,American Geophysical Union, p. 129–149.

MIGEON, S., SAVOYE, B., BABONNEAU, N., AND SPY-ANDERSON, F.L., 2004, Processes ofsediment-wave construction along the present Zaire deep-sea meandering channel:Role of meanders and flow stripping: Journal of Sedimentary Research, v. 74, p.580–598.

MOGUEDET, G., 1988, Les relations entre le fleuve Congo et le sedimentation recente surla marge continentale entre l’embouchure et le Sud Gabon: etude hydrologique,sedimentologique et geochimique [unpublished These d’etat thesis]: University ofAngers, Angers, 187 p.

PEAKALL, J., MCCAFFREY, B., AND KNELLER, B., 2000, A process model for the evolution,morphology, and architecture of sinuous submarine channels: Journal of SedimentaryResearch, v. 70, p. 434–448.

PEAKALL, J., AMOS, K.J., KEEVIL, G.M., BRADBURY, W.P., AND GUPTA, S., 2007, Flowprocesses and sedimentation in submarine channel bends: Marine and PetroleumGeology, v. 24, p. 470.

PICKERING, K.T., COLEMAN, J., CREMER, M., DROZ, L., KOHL, B., NORMARK, W.R.,O’CONNELL, S., STOW, D.A.V., AND MEYER-WRIGHT, A., 1986, A high-sinuosity,laterally migrating submarine fan channel–levee–overbank: results from DSDP Leg96 on the Mississippi Fan, Gulf of Mexico: Marine and Petroleum Geology, v. 3, p.3–18.

PIRMEZ, C., AND FLOOD, R.D., 1995, Morphology and structure of Amazon Channel, inFlood, R.D., Piper, D.J.W., and Klaus, A., eds., Proceedings of Ocean DrillingProgram, Initial Reports, v. 155, p. 23–45.

PIRMEZ, C., AND IMRAN, J., 2003, Reconstruction of turbidity currents in AmazonChannel: Marine and Petroleum Geology, v. 20, p. 823–849.

PRATSON, L.F., IMRAN, J., PARKER, G., AND SYVITSKI, J.P.M., 2000, Debris flowvs. turbidity currents: a modeling comparison of their dynamics and deposits,in Bouma, A.H., Coleman, J.M., and Meyer, A.W., et al., ed., Fine-Grained Tur-bidite Systems: American Association of Petroleum Geologists, Memoir, p. 57–72.


SAVOYE, B., COCHONAT, P., APPRIOUAL, R., BAIN, O., BALTZER, A., BELLEC, V., BEUZART,P., BOURILLET, J.F., CAGNA, R., CREMER, M., CRUSSON, A., DENNIELOU, B., DIEBLER,D., DROZ, L., ENNES, J.C., FLOCH, G., FOUCHER, J.P., GUIOMAR, M., HARMEGNIES, F.,KERBRAT, R., KLEIN, B., KHUN, H., LANDURE, J.Y., LASNIER, C., LE DREZEN, E.,LE FORMAL, J.P., LOPEZ, M., LOUBRIEU, B., MARSSET, T., MIGEON, S., NORMAND, A.,NOUZE, H., ONDREAS, H., PELLEAU, P., SAGET, P., SERANNE, M., SIBUET, J.C., TOFANI,R., AND VOISSET, M., 2000, Structure et evolution recente de l’eventail turbiditique duZaıre: premiers resultats scientifiques des missions d’exploration ZaiAngo 1 & 2(Marge Congo–Angola): Academie des Sciences de la Terre, Comptes Rendus, v. 331,p. 211–220.

SAVOYE, B., BABONNEAU, N., DENNIELOU, B., AND BEZ, M., 2009, Geological overview ofthe Angola–Congo Margin, the Congo deep-sea fan and its submarine valleys: Deep-Sea Research II, v. 56, p. 2169–2182.

SCHUMM, S.A., AND WINKLEY, B.R., 1994. The Variability of Large Alluvial Rivers: NewYork, American Society of Civil Engineers Press, 467 p.

SCHWENK, T., SPIESS, V., HUBSCHER, C., AND BREITZKE, M., 2003, Frequent channelavulsions within the active channel–levee system of the middle Bengal Fan—anexceptional channel–levee development derived from Parasound and Hydrosweepdata: Deep-Sea Research II, v. 50, p. 1023–1045.

STRAUB, K.M., MOHRIG, D., MCELROY, B., BUTTLES, J., AND PIRMEZ, C., 2008,Interactions between turbidity currents and topography in aggrading sinuoussubmarine channels: A laboratory study: Geological Society of America, Bulletin,v. 120, p. 368–385.

TORRES, J., DROZ, L., SAVOYE, B., TERENTIEVA, E., COCHONAT, P., KENYON, N.H., AND

CANALS, M., 1997, Deep-sea avulsion and morphosedimentary evolution of the RhoneFan Valley and Neofan during the Late Quaternary (north-western MediterraneanSea): Sedimentology, v. 44, p. 457–477.

VAN WEERING, T.C.E., AND VAN IPEREN, J., 1984, Fine-grained sediments of the Zairedeep-sea fan, southern Atlantic Ocean, in Stow, D.A.V., and Piper, D.J.W., eds., Fine-grained Sediments: Deep-Water Processes and Facies; Geological Society of London,Special Publication, v. 15, p. 95–113.

Received 18 June 2008; accepted 26 May 2010.


Documents

SEDIMENTARY ARCHITECTURE IN MEANDERS OF A …basin.earth.ncu.edu.tw/download/courses/seminar_MSc/2010/0106-1... · SEDIMENTARY ARCHITECTURE IN MEANDERS OF A ... Sinuous deep-water