Rip currents as a geological tool

This paper considers the nature and sedimentary significance of rip currents. Rip currents are the main factor responsible for the transport of coarse sediments from the littoral zone to greater depths. Such sediments, deposited outside the zone of wave deformation, may be identified as storm rip current increments within sediments deposited during fair weather. Composite beds deposited during a given wave cycle by storm rip currents are closely similar to ‘turbidites’ and many so called ‘fluxo-turbidites’ described from flysch deposits. Using data gathered in studies conducted on the modern Baltic coast, supplemented by experimental work and theoretical considerations, a sedimentary model is proposed. The model may be used to interpret possible rip current deposits among shallow water ‘turbidites’, and both modern storm sediments and ancient ‘tempestites’.     

The exaggerated radiocarbon age of deposit‐feeding molluscs in calcareous environments

Throughout northern Canada, live‐collected, pre‐bomb, deposit‐feeding marine molluscs from calcareous sediments yield greater apparent radiocarbon ages than do suspension feeders. We explore the size of this effect in a set of 57 paired datings of deposit feeders, mainly Portlandia arctica, and suspension feeders, mainly Hiatella arctica and Mya truncata, collected from both calcareous and non‐calcareous Holocene sediments. Deposit feeders from calcareous sediments are older than their suspension‐feeding counterparts by as much as 2240±130 ¹⁴C years. This is attributed to the uptake of ‘old’ bicarbonate derived from calcareous bedrock. The age discrepancy between suspension and deposit feeders in calcareous terrain is non‐systematic in space and time, thereby invalidating the application of a correction. In contrast, the age comparisons are concordant at sites located on the Precambrian Shield. In terrestrial environments underlain by carbonate, previous acceptance of dates on deposit feeders led to erroneous interpretations of deglaciation and relative sea‐level history, in both the North American and the Eurasian Arctic. This has prompted several researchers to exclude deposit feeders from their late Quaternary reconstructions. The same chronological problem of deposit‐feeding molluscs now needs to be more widely acknowledged by the marine community.     

Rollers and ripples in sand, streams and sky: rhythmic alteration of transverse and longitudinal vortices in three orders

Sediment ripples are caused by systematically-spaced transverse roller vortex systems in a moving fluid undergoing shear. With greater shear, these transverse rollers change over into longitudinal (helicoidal) vortices. This is the basic cause for the change from so-called ‘lower flow regime’ conditions to ‘upper flow regime’ conditions. All characteristics of these two regimes (sediment transport rate, bed form, sedimentary structures) are logically explained by attributing them to change in type of vortex system. For currents depositing sediments, there are three orders of magnitude of vortices, each order beginning with transverse rollers, passing through festoon to longitudinal rollers. A chaos zone (antidunes) ensues, followed by resumption of transverse rollers that are five to ten times as large as those in the previous order. Features of river sediments, marine sands, turbidites, desert sand dunes, sky, and stars are satisfactorily explained by this model.     

Grain-size characteristics of fine-grained unflocculated sediments I:'one-round' distributions

A simple physical model of gravitational settling from an unsorted, unflocculated source suspension is presented and an equation derived to describe the grain-size spectra of the resulting bottom sediment. Results of grain-size analyses of sediments from a variety of environments and geographical locations are shown to conform with the postulated model. The characteristic size spectrum, termed ‘one-round’ sediment, identifies a deposit which has settled from suspension with no subsequent reworking resulting in modification of the grain-size distribution. The distribution of settling rates of grains in the suspension may be inferred from an analytical form fit to the bottom sediment grain-size spectrum, along with knowledge of certain physical characteristics of the fluid (e.g. mean velocity profile).     

Carbonate sediment transport pathways based on foraminifera: case study from Frank Sound, Grand Cayman, British West Indies

Foraminifera can be used to determine the source(s) of carbonate sediment and the directions of sediment transport in shallow, coastal lagoons such as Frank Sound on the south-central coast of Grand Cayman. These determinations, based on the distribution of foraminiferal assemblages and ‘tracer species’ (numerically abundant species that live in known physiographic units and/or ecological conditions), show that the lagoonal sediments are a mixture of grains that originated in the lagoon and forereef. The variable proportions of these foraminifera throughout the lagoon reflects the dynamic processes that control lagoonal sedimentation. Amphistegina gibbosa, Discorbis rosea, and Asterigerina carinata lived in the forereef environment. The fact that these ‘tracer species’ are found throughout Frank Sound and in the beach sands, shows that they were transported across the reef crest and the lagoon. Abrasion-resistant Archaias angulatus, a‘tracer species’ indicative of a lagoonal setting, forms up to 50% of foraminiferal assemblages found in the lagoonal sediments. Preferential winnowing of small tests from these populations indicates sorting under high energy conditions. The vertical distribution of the forereef and lagoonal foraminifera in the sediment blanket that covers the floor of Frank Sound indicates that these processes are temporally persistent. Transportation of forereef foraminifera into and around the lagoon and sorting of the lagoonal foraminifera cannot take place under ‘normal’ conditions when the tranquil lagoon is characterized by weak currents. Storms and/or hurricanes, however, generate short-lived high-energy events that can move and sort the sediment and foraminifera. At the height of a storm, water and sediment are moved over the reef and then piled and held onshore by the onshore winds and the constant flow of water over the reef and across the lagoon. These currents can mix some of the lagoonal and forereef sediments. As a storm wanes, however, the ‘piled water’ flows offshore via strong rip currents that pass into the ocean through the channels which transect the reef. These currents winnow and/or strip sediment from the lagoon and may transport lagoonal sediments into the forereef area. As a result, residual lagoonal sediment is commonly characterized by larger and abrasion-resistant foraminifera.     

Spatial distribution and temporal evolution of sediment transport pathway of deltaic deposits in a rift basin: an example from Liaodong Bay sub-basin,Bohai Bay,China

The Liaodong Bay sub-basin is a classic non-marine rift sub-basin in the Bohai Bay, northeastern China. The study area is located on the east side of Liaoxi uplift in the west slope of Liaodong Bay sub-basin. It sits on a draped anticline zone above the paleo-uplift and contains the second biggest offshore hydrocarbon field found in China to date. The sub-basin is bound to the west by the TanLu fault zone, the most active and largest fault active zone in eastern of China, and has been active from the Mesozoic to present. The spatial distribution and temporal evolution of the depositional systems in the lacustrine rift basin were significantly controlled by topography of paleo-uplift and the distribution of sediment transport pathways. Using 3D seismic and densely spaced well data, we systematically analysed the spatial distribution and temporal evolution of sediment transport pathway of the deltaic deposits in the SZ36-1 oilfield in the eastern slope of the Liaozhong sag. Two types of sediment transport pathway, including the fault relay ramps and erosional valleys, were recognised: (i) the fault relay ramps between two sub-parallel faults; and (ii) a series erosional valleys on the uplift, with ‘V-’, ‘U-’ or ‘W-'shaped cross-sectional patterns. Seismic stratal slices reveal that the erosional valleys branch and converge from upstream to downstream. The paleomorphology of the Liaoxi uplift (e.g., erosional valleys and sub-uplifts) and their evolution control the depositional systems and the pattern of sediment dispersal in the rift lacustrine basin area. The research indicates that paleogeomorphology controlled the direction of sediment transport, the capacity and position of sediment accommodation, influenced the type of sedimentary micro-facies and the spatial distribution pattern of the sediments. Seismic stratal slices and paleogeography maps reveal the erosional valleys shrunk progressively with sedimentary fills, resulting in decreasing gradients of the depositional slope, and provide a gentle geomorphology for a large-scale fluvial-delta depositional system to develop.     

洋山港海域现代沉积物运移趋势

地形、物源和水动力条件是洋山港海域现代沉积环境的主要控制因素，粉粒和粘粒是现代沉积物的主要组成部分，潮流是起动和搬运现代沉积物的主要动力，悬移是现代沉积物最主要的搬运方式，跃移是次要搬运方式，现代沉积物的起动和沉积过程交替频繁。研究区现代沉积物长期净运移趋势表明：研究区东部的黄泽洋向洋山港海域的输沙趋势显著；洋山港岛链峡道深槽的形成，系由峡道东部和西部的现代沉积物各向峡道两端分异运移所致；小洋山—唐脑山—线将是今后港区泥沙淤积的主要部位之一。     

Textural characteristics of the Krishna River sediments,India

To evaluate the magnitude of variation in grain size distribution in the Krishna river, bed sediments and suspended sediments collected along the length of the river have been studied. There are both temporal and seasonal variation in the grain size distribution of suspended sediments. The statistical parameters show the change along the river in a non-linear fashion which may be due to human interference and due to different types of sediments contributed by tributaries to the Krishna river. The suspended sediments are mostly fine silt (4 to 16m), poorly sorted, showing coarse to fine skewed and are platyto leptokurtic. The bed sediments are mostly medium sand (350m) showing moderate to well sorted, coarse to fine skewed and are platy- to leptokurtic. The CM diagram of Krishna river bed sediments suggests that deposition takes place by (1) rolling (2) rolling and suspension and (3) graded suspension. The suspended sediments represent deposits of uniform suspension.     

A laboratory flume investigation of the formation of fossil stem fills

The mechanism of formation of fossil plant ‘pith casts’ has been investigated experimentally using a small laboratory flume tank. The extent to which filling occurs is dependent on stem length and diameter, and on the current velocity. Stem fills produced by sediment carried in suspension show a distinct structure of two wedges of sediment deposited at either end of the stem cavity. These wedges are deposited from small flow vortices which form when the flow around the stem separates, as the flow through the stem becomes restricted. Stems filled by bedload currents do not show such a uniform structure. The experimentally produced stem fills are comparable with fossil pith casts, e.g. Catamites. The formation of cortical casts of Stigmaria is not simply explained in terms of the processes operating in open-ended stem segments. This investigation of the mechanics of filling of hollow plant organs, and the structure of the fill produced, offers a basis for interpreting the sedimentary environment in which the fossil fills have formed, and leads to a fuller understanding of the mode of formation of such fossils.     

The terminology of structures in thrust belts

A review of structures and geometric relationships recognized in thrust belts is presented. A thrust is defined as any contractional fault, a corollary being that thrusts must cut up-section in their transport direction. ‘Flats’ are those portions of a thrust surface which were parallel to an arbitrary datum surface at the time of displacement and ‘ramps’ are those portions of thrusts which cut across datum surfaces. Ramps are classified on the basis of their orientation relative to the thrust transport direction and whether they are cut offs in the hangingwall or footwall of the thrust. Lateral variations in the form of staircase trajectories are joined by oblique or lateral ramps which have a component of strike-slip movement.An array of thrusts which diverge in their transport direction may form by either of two propagation models. These are termed ‘piggy-back’ propagation, which is foreland-directed, and ‘overstep’ propagation which is opposed to the thrust transport direction. An array of thrust surfaces is termed an ‘imbricate stack’ and should these surfaces anastamose upwards a ‘duplex’ will result; the fault-bounded blocks are termed ‘horses’. A duplex is bounded by a higher, ‘roof’ thrust and a lower, ‘floor’ thrust. The intersection of any two thrust planes is termed a ‘branch line’.Thrusts can be classified on the basis of their relationship to asymmetric fold limbs which they cut. A further classification arises from whether a particular thrust lies in the hangingwall or footwall of another one.The movement of thrust sheets over corrugated surfaces, or the local development of thrust structures beneath, will fold higher thrust sheets. These folds are termed ‘culminations’ and their limbs are termed ‘culmination walls’. Accommodation of this folding may require movement on surfaces within the hangingwall of the active thrust. These accommodation surfaces are termed ‘hangingwall detachments’ and they need not root down into the active thrust. This category of detachment includes dip-slip ‘hangingwall drop faults’ which are developed by differential uplift of duplex roofs, and ‘out-of-the-syncline’ thrusts which develop from overtightened fold hinges. Back thrusts, as well as forming as hangingwall detachments, may also form due to layer-parallel shortening above a sticking thrust or by rotation of the hangingwall above a ramp.     

An application of scanning electron microscopy of quartz sand surface textures to the environmental diagnosis of Neogene carbonate sediments, Finestrat Basin, south-east Spain

The Finestrat Basin, 25 km NE of Alicante, contains a poorly understood sequence of Neogene sediments that lack an indigenous fauna. Study of the surface textures of euhedral quartz grains, probably derived by first-cycle weathering from an underlying Triassic mudstone diapir, provides new data on the depositional environment of the basin. The possible influence of alluvial fans is recorded by mechanical breakage features in these non-glaciated sediments. The distribution and abundance of chemical etch pits suggest their diagenetic origin following non-marine deposition, while the density and abundance of mechanically induced ‘V’-shaped pits seems typical of fluvial and low-energy littoral environments. The presence of silica precipitated during transport, such as silica globules, suggests a silica-rich transport medium and a depositional environment that was not highly alkaline. In combination with independent sedimentological evidence, these data suggest that the sediments collected in a lake after derivation by rivers and alluvial fans from the diapir. It is concluded that the abundance and superposition of features must be noted for each grain in order to yield useful results, and that the mechanisms acting on each grain be deduced before attempting to identify specific environments.     

Depositional models for fine-grained sediment in the western Hellenic Trench, Eastern Mediterranean

Sediment in tectonically active, topographically restricted settings of the western Hellenic Arc, eastern Mediterranean, consists primarily of clayey silt and silty clay. Failure of metastable sediment temporarily stored on relatively steep slopes is triggered by earthquake tremors and eustatic oscillations. Redeposition of these materials by gravitative transport has resulted in markedly different lithofacies from site to site. Most piston cores include three Late Quaternary stratigraphic units that can be correlated with sections in other parts of the eastern Mediterranean; numerous radiocarbon-age determinations enhance the correlation. Seven fine-grained sediment types are identified in cores from eight distinct depositional environments. Some muds are closely related to specific environments (slump and debris flow deposits on slope and high-relief environments), or to time (well laminated mud during the latest Pleistocene-mid-Holocene), or to both (uniform and faintly laminated muds restricted to trench basins). Turbiditic and hemipelagic muds are common throughout the study area. Mud distribution patterns correlate closely with calculated sedimentation rates. We propose two depositional models for these sediments. The first emphasizes downslope transformations resulting in progressively reduced flow concentration during transport: from slump and debris flow–>turbidity current–>low density turbidity current or turbid layer mechanisms. The distal end-member deposits settling from low concentration flows are thick, rapidly emplaced, fine-grained uniform muds closely associated with faintly laminated muds. These were ponded in flat trench basin-plains. Planktonic and terrigenous fractions in the turbiditic, finely laminated and uniform muds record mixing of materials of gravitative and suspension origin during redeposition. This sequence prevails under conditions of minimal stratification of water masses, as characterized by the present Mediterranean. In the second model developed for conditions of well-developed water mass stratification, well laminated rather than uniform mud prevails as the end product of low concentration flows. These very finely laminated and graded muds record particle-by-particle settling from detached turbid layers concentrated along density interfaces; they include material from turbid layers complemented by the normal ‘rain’ of pelagic material. Stratification barriers resulted in region-wide distribution of such deposits, in both slope and trench environments.     

Geological evolution of the Hampshire Basin (southern England) during a global climate transition from ‘hothouse’ to ‘coolhouse’ in the Palaeogene

Palaeogene sediments of the Hampshire Basin were a sensitive recorder of fluctuations in climate and eustatic sea level as Earth's climate transitioned from the global early Eocene ‘hothouse’ to the early Oligocene ‘coolhouse’, accompanied by the first permanent continent-scale glaciation of Antarctica at the Eocene/Oligocene boundary. A study of the Palaeogene sediments of the Hampshire Basin is not only interesting from a palaeoclimate perspective, but the marine middle Eocene formations are renowned for containing some of the most abundant and diverse Palaeogene fossil assemblages in the world. In this article, I take you on a tour of the geological evolution of the Hampshire Basin from the end of the Paleocene through to the Pleistocene. I highlight some of the best exposures where sediments can be studied at outcrop and representative fossil assemblages can be collected, along with the economic and archaeological significance of these Palaeogene sediments.     

Particulate calcium carbonate in New England shelf waters: result of shell degradation and resuspension*

Aragonite and calcite needles, 30–500 μm long, were found to be prominent visual components throughout the water column in some areas of the New England (northeastern U.S.A.) continental shelf during winter months. Further investigation showed these‘needles’to be laths derived from the degradation of mollusc shells which were resuspended from bottom sediments during winter storms. Such degradation and subsequent transport/dissolution of carbonate particles may help explain why the terrigenous 'starved’shelf sediments off New England contain such small amounts of calcium carbonate.     

Cadmium transport in sediments by tubificid bioturbation: An assessment of model complexity

Biogeochemistry of metals in aquatic sediments is strongly influenced by bioturbation. To determine the effects of biological transport on cadmium distribution in freshwater sediments, a bioturbation model is explored that describes the conveyor-belt feeding of tubificid oligochaetes. A stepwise modelling strategy was adopted to constrain the many parameters of the model: (i) the tubificid transport model was first calibrated on four sets of microspheres (inert solid tracer) profiles to constrain tubificid transport; (ii) the resulting transport coefficients were subsequently applied to simulate the distribution of both particulate and dissolved cadmium. Firstly, these simulations provide quantitative insight into the mechanism of tubificid bioturbation. Values of transport coefficients compare very well with the literature, and based on this, a generic model of tubificid bioturbation is proposed. Secondly, the application of the model to cadmium dataset sheds a light on the behaviour of cadmium under tubificid bioturbation. Cadmium enters the sediment in two ways. In one pathway, cadmium enters the sediment in the dissolved phase, is rapidly absorbed onto solid particles, which are then rapidly transported to depth by the tubificids. In the other pathway, cadmium is adsorbed to particles in suspension in the overlying water, which then settle on the sediment surface, and are transported downwards by bioturbation. In a final step, we assessed the optimal model complexity for the present dataset. To this end, the two-phase conveyor-belt model was compared to two simplified versions. A solid phase-only conveyor-belt model also provides good results: the dissolved phase should not be explicitly incorporated because cadmium adsorption is fast and bioirrigation is weak. Yet, a solid phase-only biodiffusive model does not perform adequately, as it does not mechanistically capture the conveyor-belt transport at short time-scales.     

The Canary Debris Flow: source area morphology and failure mechanisms

The morphology of the source area of the Canary Debris Flow has been mapped using both GLORIA reconnaissance and TOBI high-resolution sidescan sonar systems. West of ≈19°W, the seafloor is characterized by a strongly lineated downslope-trending fabric. This fabric can be interpreted as being caused by streams of debris separated by longitudinal shears. Multiple flow pulses are indicated by a series of asymmetrical lateral ridges which mark the northern boundary of the flow. East of ≈19°W, GLORIA data show only a weak fabric of irregular patches and alongslope lineaments. The TOBI data show the patches to be coherent sediment blocks up to 10 km across, surrounded by debris flow material. These are interpreted as in situ areas of seafloor sediment which have survived the slope failure and debris flow event rather than transported fragments of a failed sediment slab. TOBI data from the best developed area of alongslope lineaments show a series of small faults downstepping to the west. This area of seafloor is interpreted as one of partial sediment failure, where the failure process became ‘frozen’ before total mobilization of the seafloor sediments could occur. The overall morphology of the failure area indicates removal of a slab-like body of sediment, although we cannot distinguish between retrogressive and slab-slide failure mechanisms. If the latter mechanism is applicable, fragmentation of the failing ‘slab’ must have commenced concurrently with the onset of downslope transport. Immediately upslope from the debris flow source area, a seafloor of characteristic rough blocky texture is interpreted as the surface of a debris avalanche derived from the slopes of the island of El Hierro. The debris flow and avalanche appear to be simultaneous events, with failure of the slope sediments occurring while the avalanche deposits were still mobile enough to fill and disguise the topographic expression of the debris flow headwall. Loading of the slope sediments by the debris avalanche most probably triggered the Canary Debris Flow.     

Late Quaternary mass movement events in Lake El′gygytgyn,North‐eastern Siberia

Lake El′gygytgyn is situated in a 3·6 Myr old impact crater in North‐eastern Siberia. Its sedimentary record probably represents the most complete archive of Pliocene and Quaternary climate change in the terrestrial Arctic. In order to investigate the influence of gravitational sediment transport on the pelagic sediment record in the lake centre, two sediment cores were recovered from the lower western lake slope. The cores penetrate a sub‐recent mass movement deposit that was identified by 3·5 kHz echo sounding. In the proximal part of this deposit, deformed sediments reflect an initial debris flow characterized by limited sediment mixture. Above and in front of the debrite, a wide massive densite indicates a second stage with a liquefied dense flow. The mass movement event led to basal erosion of ca 1 m thick unconsolidated sediments along parts of its flow path. The event produced a suspension cloud, whose deposition led to the formation of a turbidite. The occurrence of the turbidite throughout the lake and the limited erosion at its base mainly suggest deposition by ‘pelagic rain’ following Stokes’ Law. Very similar radiocarbon dates obtained in the sediments directly beneath and above the turbidite in the central lake confirm this interpretation. When applying the depositional model for the Late Quaternary sediment record of Lake El′gygytgyn, the recovered turbidites allow reconstruction of the frequency and temporal distribution of large mass movement events at the lake slopes. In total, 28 turbidites and related deposits were identified in two, 12·9 and 16·6 m long, sediment cores from the central lake area covering approximately 300 kyr.     

Composite methodology for interpreting sediment transport pathways from spatial trends in grain size: A case study of the Lithuanian coast

To achieve a more robust interpretation of sediment conditions and transport based on grain size, this study combines the perspectives of two different interpretation techniques that have been developed separately and applied in the literature: (i) ‘grain size trend analysis’ interprets changes between interrelated sampling sites assuming that sediment in the direction of transport should become either coarser, better sorted and more positively skewed, or finer, better sorted and more negatively skewed; and (ii) the ‘population anomalies’ method, which estimates the balance between erosion and accumulation processes at each sampling site based on individual site deviations compared with average values for grain‐size parameters in the area of interest. The simultaneous use of both approaches enables sedimentological conditions and transport alternatives to be distinguished. This improvement is necessary because site‐specific changes and general textural trends provide information that complements the strengths and weaknesses of each perspective. The methods are implemented here using two‐dimensional Geographic Information System tools and illustrated for the Lithuanian coast. Grain‐size characteristics of 712 surface samples of sandy sediments are used to interpret sediment transport pathways and predominant sedimentological conditions in the Lithuanian coastal zone. In general, coast‐parallel sediment transport dominates the entire investigation area, although wave‐induced movement perpendicular to the coastline is inferred in the shallow nearshore zone. The deepest areas offshore are characterized by sea floor erosion in the north, whereas an accumulation zone occurs in the relatively deep central part of the study area.     

Sedimentation and palaeoenvironments of Late Cretaceous crater-lake deposits in Bushmanland, South Africa

A large diameter borehole core from an epiclastic kimberlite remnant on the farm Stompoor in the Prieska district, Cape Province, contains a continuous 76 m section of fossiliferous sediments interpreted as having accumulated within a crater-lake during the Late Cretaceous. Three distinct facies associations reflect depositional processes that prevailed in offshore areas of the original lake. Facies Association A: matrix-supported pebble conglomerates comprising a chaotic assemblage of pyroclastic, basement and country rocks set in a fine-grained matrix. Flat, non-erosional basal surfaces with ‘frozen’ rip-up clasts, the protrusion of matrix-supported clasts above the upper surfaces and a direct relationship between maximum clast size and bed thickness suggest deposition from debris flows that originated subaerially on pyroclastic talus cones surrounding the crater. Facies Association B: alternating thin beds of matrix-supported granule conglomerate, structureless fine-grained sandstone and parallel laminated mudrock. Small fining-upward sequences within these beds are comparable to turbidite Bouma Tade, Tde. Numerous partings display petrified fish and frog skeletons, as well as bivalve, gastropod and ostracode shells, leaf impressions, insect wings and a possible bird bone. These beds were deposited by thin debris-flows and turbidity underflows interspersed with periods of ‘pelagic’ sedimentation. Facies Association C: microlaminated mudstone beds containing scattered ‘dropstone lapilli’. The lamination is imparted by alternating Ca-rich/Ca-poor layers which may reflect climatic seasonality. They are interpreted as the result of seasonally influenced suspension settling through a thermally stratified water column. Short-term periodicities in conglomerate bed thicknesses are interpreted as the result of successive block caving of a slump scar giving rise to several debris flows from the same source area. Seismic shock from nearby volcanism may have simultaneously triggered slumps on both subaerial and subaqueous slopes. Dropstone lapilli in Type C beds and the preponderance of load casting in Type B beds support this interpretation. An estimate of the time span involved in accumulating 76 m of crater lake sediments based on the possible seasonal imprint of Type C beds gives a figure of some 220,000 yr.     

Stream-driven, high-density gravelly traction carpets: possible deposits in the Trabeg Conglomerate Formation, SW Ireland and some theoretical considerations of their origin

Within high-density flood flows a prominent mechanism of gravel transport and deposition is by stream-driven, high-density traction carpet (with a rheology similar to grain flow). These gravel carpets are envisaged to form the basal portion of a bipartite high-density flood flow, decoupled from an overlying sand- and silt-laden turbulent flow. Several examples already documented in the literature are reviewed and an additional case from the Lower Old Red Sandstone of southwest Ireland is presented. Two mechanisms of traction carpet initiation are discussed: by rapid entrainment of gravel into suspension on rising stage, followed by settling into the gravel traction carpet at peak and falling stage; and by overconcentration of a ‘normal’, low-density bedload. Gravel entrainment, suspension and traction carpet development are significantly easier if the flood water already carries a high concentration of sand and silt in suspension. Theoretical consideration further shows that gravelly traction carpets can be maintained in channels of relatively low gradient by the shear stress exerted by the high-density, sand-bearing turbulent flood flow above. This tangential shear stress is converted to dispersive pressure, which aids buoyancy and quasi-static grain-to-grain contacts in the support of the clasts within the gravel carpet. The carpet is thought to have a quasi-plastic rheology but behave much like a viscous fluid at high shear rates. Stream-driven gravelly traction carpets are expected to produce sheet-like units of clast- to matrix-supported conglomerate, characterized by a parallel or an a(p)a(i) clast fabric. These units may be ungraded, normally or inversely graded, depending on the rate of shear, the viscosity of the flow and the celerity of deposition.