Assessing the effect of vegetation‐related bank strength on channel morphology and stability in gravel‐bed streams using numerical models

Bank strength due to vegetation dominates the geometry of small stream channels, but has virtually no effect on the geometry of larger ones. The dependence of bank strength on channel scale affects the form of downstream hydraulic geometry relations and the meandering‐braiding threshold. It is also associated with a lateral migration threshold discharge, below which channels do not migrate appreciably across their floodplains. A rational regime model is used to explore these scale effects: it parameterizes vegetation‐related bank strength using a dimensionless effective cohesion, C_r*. The scale effects are explored primarily using an alluvial state space defined by the dimensionless formative discharge, Q*, and channel slope, S, which is analogous to the Q–S diagrams originally used to explore meandering‐braiding thresholds. The analyses show that the effect of vegetation on both downstream hydraulic geometry and the meandering‐braiding threshold is strongest for the smallest streams in a watershed, but that the effect disappears for Q* > 10⁶. The analysis of the migration threshold suggests that the critical discharge ranges from about 5 m³/s to 50 m³/s, depending on the characteristic rooting depth for the vegetation. The analysis also suggests that, where fires frequently affect riparian forests, channels may alternate between laterally stable gravel plane‐bed channels and laterally active riffle‐pool channels. These channels likely do not exhibit the classic dynamic equilibrium associated with alluvial streams, but instead exhibit a cyclical morphologic evolution, oscillating between laterally stable and laterally unstable end‐members with a frequency determined by the forest fire recurrence interval. Copyright © 2008 John Wiley & Sons, Ltd.     

Life of a bifurcation in a gravel‐bed braided river

Channel bifurcation is a key element in braided rivers, determining the water and sediment distribution and hence controlling the morphological evolution. Recent theoretical and experimental findings, as well as field observations, showed that bifurcations in gravel‐bed braided rivers are often asymmetrical and highly unstable. In this paper field data are presented on a bifurcation in the Tagliamento River, northeast Italy. The planform configuration of the bifurcation and its temporal evolution was monitored by an automatic digital camera during a series of seven floods with different magnitudes. This remote sensing technique allowed a high temporal resolution (pictures were acquired every hour) that was proved to be essential in a highly dynamic system as the one considered here. Digitized maps of the channels provided information on the location of the bifurcation, the width of the anabranches, the angle between them, along with the occurrence and migration of sediment bars. Data were acquired at two different water levels, giving the possibility to compare low and high flow conditions. The monitored bifurcation is largely unstable and shows sudden changes in the water distribution, mainly driven by the bar migrating in the upstream channel and entering the distributaries. A relationship between width asymmetry and flood magnitude was observed, confirming previous analyses. Moreover, recent theoretical findings were applied, in order to test the possibility to estimate general trends in bifurcation evolution. The analysis pointed out the relevance of a correct assessment of the characteristic temporal scales, as the bifurcation evolves on a timescale similar to that of bar migration and flood duration. Understanding the interactions between these processes is therefore crucial in order to increase the ability to model and predict the morphological evolution of a braided network. Copyright © 2012 John Wiley & Sons, Ltd.     

Form and growth of bars in a wandering gravel‐bed river

The changing form of developing alluvial river bars has rarely been studied in the field, especially in the context of the fixed, compound, mainly alternate gravel bars that are the major morphological feature of the wandering style. Century scale patterns of three‐dimensional growth and development, and the consequent scaling relations of such bars, are examined along the gravel‐bed reach of lower Fraser River, British Columbia, Canada. A retrospective view based on maps and aerial photographs obtained through the twentieth century shows that individual bars have a life history of about 100 years, except in certain, protected positions. A newly formed gravel bar quickly assumes its ultimate thickness and relatively quickly approaches its equilibrium length. Growth continues mainly by lateral accretion of unit bars, consistent with the lateral style of instability of the river. Bar growth is therefore allometric. Mature bars approach equilibrium dimensions and volume that scale with the overall size of the channel. Accordingly, the bars conform with several published criteria for the ultimate dimensions of alternate barforms. Sand bars, observed farther downstream, have notably different morphology. Fraser River presents a typical wandering channel planform, exhibiting elements of both meandered and low‐order braided channels. Hydraulic criteria to which the Fraser bars conform illustrate why this planform develops and persists. The modest rate of bed material transfer along the channel – typical of the wandering type – determines a century‐length time scale for bar development. This time scale is consistent with estimates that have been made for change of the macroform elements that determine the overall geometry of alluvial channels. Copyright © 2009 John Wiley & Sons, Ltd.     

Riparian vegetation and island formation along the gravel‐bed Fiume Tagliamento,Italy

After more than 300 years of river management, scientific knowledge of European river systems has evolved with limited empirical knowledge of truly natural systems. In particular, little is known of the mechanisms supporting the evolution and maintenance of islands and secondary channels. The dynamic, gravel‐bed Fiume Tagliamento, Italy, provides an opportunity to acquire baseline data from a river where the level of direct engineering intervention along the main stem is remarkably small. Against a background of a strong alpine to mediterranean climatic and hydrological gradient, this paper explores relationships between topography, sediment and vegetation at eight sites along the active zone of the Tagliamento. A conceptual model of island development is proposed which integrates the interactions between large woody debris and vegetation, geomorphic features, sediment calibre and hydrological regime. Islands may develop on bare gravel sites or be dissected from the floodplain by channel avulsion. Depositional and erosional processes result in different island types and developmental stages. Differences in the apparent trajectories of island development are identified for each of the eight study sites along the river. The management implications of the model and associated observations of the role of riparian vegetation in island development are considered. In particular, the potential impacts of woody debris removal, riparian tree management, regulation of river flow and sediment regimes, and changes in riparian tree species' distribution are discussed. Copyright © 2001 John Wiley & Sons, Ltd.     

Particle path length distributions in meandering gravel‐bed streams: results from physical models

In gravel‐bed rivers with well‐de?ned pool–bar morphology, the path length of transported bed particles must be, at least during ‘channel‐forming’ ?ows, equal to the length scale of the morphology. This is the basis for some methods for estimating bed material transport rates. However, previous data, especially from ?eld tests, are often strongly positively skewed with mean much shorter than the pool–bar spacing. One possible explanation is that positively skewed distributions occur only in channels lacking distinct pool–bar topography or only at lower discharges in pool–bar channels. A series of ?ume experiments using ?uorescent tracers was used to measure path length distributions in low‐sinuosity meandering channels to assess the relation with channel morphology and ?ow conditions. At channel‐forming ?ows, 55 to 75 per cent of the tracer grains were deposited on the ?rst point bar downstream of the point of tracer input, with 15 per cent passing beyond the ?rst bar. Path length distributions are symmetrical with mean equal to the pool–bar spacing and can be described with a Cauchy distribution. In some cases there was a secondary mode close to the point of tracer introduction; this bimodal distribution ?ts a combined gamma–Cauchy distribution. Only when discharge was reduced below the channel‐forming ?ow were frequency distributions unimodal and positively skewed with no relation to the pool–bar spacing. Thus, path length distributions become more symmetrical, and mean path length increases to coincide with pool–bar spacing, as ?ow approaches channel‐forming conditions. This is a substantial modi?cation of existing models of particle transfer in gravel‐bed rivers. Copyright © 2003 John Wiley & Sons, Ltd.     

Effect of changes in fine‐grained matrix on bedload sediment transport in a gravel‐bed river

While clay and silt matrices of gravel‐bed rivers have received attention from ecologists concerned variously with the deteriorating environments of benthic and hyporheic organisms, their impact on sediment entrainment and transport has been explored less. A recent increase of such a matrix in the bed of Nahal Eshtemoa, an ephemeral river of the northern Negev, has more than doubled the boundary shear stress needed to initiate bedload, from 7 N m^‐2 (τ^* = 0.027) during the flash floods of 1991–2001 to 15 N m^‐2 (τ^* = 0.059) during those of 2008–2009. The relation between bedload flux and boundary shear stress continues to be well‐defined, but it is displaced. The matrix now contains a significant amount of silt and clay size material. The reasons for the increased entrainment threshold of bedload are explored. Large‐scale laser scanning of the dry bed reveals a reduction in grain‐scale morphological roughness, while artificial in situ tests of matrix integrity indicate considerable cohesion. The implications for adopting bed material sampling strategies that account for matrix development are assessed. Copyright © 2012 John Wiley & Sons, Ltd.     

Modeling the effects of pulsed versus chronic sand inputs on salmonid spawning habitat in a low‐gradient gravel‐bed river

It is widely recognized that high supplies of fine sediment, largely sand, can negatively impact the aquatic habitat quality of gravel‐bed rivers, but effects of the style of input (chronic vs. pulsed) have not been examined quantitatively. We hypothesize that a continuous (i.e. chronic) supply of sand will be more detrimental to the quality of aquatic habitat than an instantaneous sand pulse equal to the integrated volume of the chronic supply. We investigate this issue by applying a two‐dimensional numerical model to a 1 km long reach of prime salmonid spawning habitat in central Idaho. Results show that in both supply scenarios, sand moves through the study reach as bed load, and that both the movement and depth of sand on the streambed mirrors the hydrograph of this snowmelt‐dominated river. Predictions indicate greater and more persistent mortality of salmonid embryos under chronic supplies than pulse inputs, supporting our hypothesis. However, predicted mortality varies both with salmonid species and location of spawning. We found that the greatest impacts occur closer to the location of the sand input under both supply scenarios. Results also suggest that reach‐scale morphology may modulate the impact of sand loads, and that under conditions of high sand loading climate‐related increases in flow magnitude could increase embryo mortality through sand deposition, rather than streambed scour. Copyright © 2013 John Wiley & Sons, Ltd.     

Bedload transport measurements with impact plate geophones: comparison of sensor calibration in different gravel‐bed streams

Indirect bedload transport measurements have been made with the Swiss plate geophone system in five gravel‐bed mountain streams. These geophone sensors record the motion of bedload particles transported over a steel plate mounted flush with the channel bed. To calibrate the geophone system, direct bedload transport measurements were undertaken simultaneously. At the Erlenbach in Switzerland, a moving‐basket sampler was used. At the Fischbach and Ruetz streams in Austria, a Helley–Smith type bedload sampler provided the calibration measurements. A Bunte‐type bedload trap was used at the Rofenache stream in Austria. At the Nahal Eshtemoa in Israel, Reid‐type slot bedload samplers were used. To characterize the response of the geophone signal to bedload particles impacting on the plate, geophone summary values were calculated from the raw signal and stored at one second intervals. The number of impulses, i.e. the number of peaks above a pre‐defined threshold value of the geophone output signal, correlated well with field measured gravel transport loads and was found to be a robust parameter. The relations of impulses to gravel transport loads were generally near‐linear, but the steepness of the calibration relations differed from site to site. By comparing the calibration measurements from the different field sites and utilizing insights gained during preliminary flume experiments, it has been possible to identify the main factors that are responsible for site specific differences in the calibration coefficient. The analysis of these calibration measurements indicates that the geophone signal also contains some information about the grain size distribution of bedload. Copyright © 2013 John Wiley & Sons, Ltd.     

Flow resistance in gravel‐bed channels with large‐scale roughness

A previously published mixing length (ML) model for evaluating the Darcy–Weisbach friction factor for a large‐scale roughness condition (depth to sediment height ratio ranging from 1 to 4) is brie?y reviewed and modi?ed (MML). Then the MML model and a modi?ed drag (MD) model are experimentally tested using laboratory measurements carried out for gravel‐bed channels and large‐scale roughness condition. This analysis showed that the MML gives accurate estimates of the Darcy–Weisbach coef?cient and for Froude number values greater than 0·5 the MML model coincides with the ML one. Testing of the MD model shows limited accuracy in estimating ?ow resistance. Finally, the MML and MD models are compared with the performance of a quasi‐theoretical (QT) model deduced applying the P‐theorem of the dimensional analysis and the incomplete self‐similarity condition for the depth/sediment ratio and the Froude number. Using the experimental gravel‐bed data to calibrate the QT model, a constant value of the exponent of the Froude number is determined while two relationships are proposed for estimating the scale factor and the exponent of the depth/sediment ratio. This indirect estimate procedure of the coef?cients (b₀, b₁ and b₂) of the QT model can produce a negligible overestimation or underestimation of the friction factor. Copyright © 2003 John Wiley & Sons, Ltd.     

A two‐dimensional morphodynamic model of gravel‐bed river with floodplain vegetation

Artificially straight river channels tend to be unstable, and ultimately develop into river meanders through bank erosion and point‐bar deposition. In this paper account is taken of the effects of riparian and floodplain vegetation on bank strength, floodplain flow resistance, shear stress partitioning, and bedload transport. This is incorporated into an existing 2D hydrodynamic‐morphological model. By applying the new model to an initially straight and single‐threaded channel, the way that its planform and cross‐sectional geometry evolve for different hydraulic and floodplain vegetation conditions is demonstrated. The results show the formation and upstream migration of gravel bars, confluence scouring and the development of meandering and braiding channel patterns. In cases where the channel becomes unstable, the instability grows out of bar formation. The resulting braiding patterns are similar to analytical results. The formation of a transition configuration requires a strong influence from vegetation. Copyright © 2010 John Wiley & Sons, Ltd.     

Comparison of hydraulic geometry between sand‐ and gravel‐bed rivers in relation to channel pattern discrimination

A comparison has been made between the hydraulic geometry of sand‐ and gravel‐bed rivers, based on data from alluvial rivers around the world. The results indicate a signi?cant difference in hydraulic geometry among sand‐ and gravel‐bed rivers with different channel patterns. On this basis, some diagrams for discrimination of meandering and braided channel patterns have been established. The relationships between channel width and water discharge, between channel depth and water discharge, between width–depth ratio and water discharge and between channel slope and water discharge can all be used for channel pattern discrimination. The relationship between channel width and channel depth can also be used for channel pattern discrimination. However, the accuracy of these relationships for channel pattern discrimination varies, and the depth–discharge relationship is a better discriminator of pattern type than the classic slope–discharge function. The cause for this difference has been explained qualitatively. To predict the development of channel patterns under different natural conditions, the pattern discriminator should be searched on the basis of independent or at least semi‐independent variables. The relationship between stream power and bed material grain size can be used to discriminate channel patterns, which shows a better result than the discriminator using the slope–discharge relationship. Copyright © 2004 John Wiley & Sons, Ltd.     

Effects of river discharge on hyporheic exchange flows in salmon spawning areas of a large gravel‐bed river

The flow magnitude and timing from hydroelectric dams in the Snake River Basin of the Pacific north‐western US is managed in part for the benefit of salmon. The objective of this research was to evaluate the effects of Hells Canyon Dam discharge operations on hydrologic exchange flows between the river and riverbed in Snake River fall Chinook salmon spawning areas. Interactions between river water and pore water within the upper 1 m of the riverbed were quantified through the use of self‐contained temperature and water level data loggers suspended inside of piezometers. The data were recorded at 20 min intervals over a period of 200 days when the mean daily discharge was 218–605 m³ s^?1, with hourly stage changes as large as 1·9 m. Differences in head pressure between the river and riverbed were small, often within ± 2 cm. Measured temperature gradients in the riverbed indicated significant interactions between the surface and subsurface water. At the majority of sites, neither hydraulic nor temperature gradients were significantly affected by either short‐ or long‐term changes in discharge operations from Hells Canyon Dam. Only 2 of 14 study sites exhibited acute flux reversals between the river and riverbed resulting from short‐term, large magnitude changes in discharge. The findings suggest that local scale measurements may not be wholly explanatory of the hydrological exchange between the river and riverbed. The processes controlling surface water exchange at the study sites are likely to be bedform‐induced advective pumping, turbulence at the riverbed surface, and large‐scale hydraulic gradients along the longitudinal profile of the riverbed. By incorporating the knowledge of hydrological exchange processes into water management planning, regional agencies will be better prepared to manage the limited water resources among competing priorities that include salmon recovery, flood control, irrigation supply, hydropower production, and recreation. Copyright © 2007 John Wiley & Sons, Ltd.     

Particle clusters in gravel‐bed rivers: an experimental morphological approach to bed material transport and stability concepts

Structured gravel river beds clearly exert a major influence on bed stability. Indexing structural stability by field measurements of bed strength neglects the processes operating to entrain and transport bed material in different parts of each structure. This study takes a morphological approach to interpreting the critical processes, using particle tracing to determine the movement of individual cluster particles over a range of flood event magnitudes and durations. The experiment was carried out on the River South Tyne, UK; it uses flow hydrographs measured nearby and also benefits from previous studies of historical development, channel morphology and sediment transport at the same site. More than 30 clusters were monitored over a seven‐month period during which clusters occupied 7–16 per cent of the bed. Threshold flows delimiting three apparently contrasting bed sediment process regimes for cluster particles are tentatively set at 100 m³ s^?1 and 183 m³ s^?1; durations of flow at these levels are critical for cluster development, rather than flow peak values. Wake particles are transported most easily. Flow straightening in the wandering channel planform reduces the stability of clusters, since mechanical strength is markedly reduced by this change of direction. The overall area covered by clusters between significant transport events varies little, implying a dynamic equilibrium condition. Copyright © 2005 John Wiley & Sons, Ltd.     

Effects of sand content on initial gravel motion in gravel‐bed rivers

When fine sediments are present in gravel streambeds (gravel‐framework beds), the gravel can be more easily removed from its original position, compared with gravel in a streambed without fine sediment but otherwise under the same hydraulic conditions. In this study, the effect of the presence of sand on the initiation of gravel motion in gravel riverbeds was investigated using flume experiments. The relationship between the critical Shields stress for gravel motion initiation and the fraction of sand in the bed was determined experimentally. The results can be summarized as follows. (1) When the fraction of sand in the bed is smaller than about 0.4, the critical Shields stress for the initiation of gravel motion decreases with increasing fraction of sand. The critical Shields stress increases, however, with increasing fraction of sand when it is larger than about 0.4. (2) The difference between the value of the critical Shields stress predicted by the Egiazaroff equation and the value obtained from the experimental data becomes maximum at about 0.4 of the fraction of sand. Here an empirical relation between the critical Shields stress and the fraction of sand is proposed so as to consider the effects of the ratio of the characteristic gravel size to the mean size of the bed material on the critical Shields stress. (3) Gravel in armored beds can be more easily mobilized by supplying sand as part of a sediment augmentation scheme. The sand fraction in the subsurface layer of the bed appears to reduce the friction angle of exposed particles. Sediment augmentation using sand has been recently demonstrated to be a viable alternative for mobilizing gravel for the restoration of gravel‐bed rivers downstream of dams. The quantitative evaluation obtained through the experiments reported here may be useful for the design of augmentation schemes. Copyright © 2017 John Wiley & Sons, Ltd.     

Computational fluid dynamics modelling of boundary roughness in gravel‐bed rivers: an investigation of the effects of random variability in bed elevation

Results from a series of numerical simulations of two‐dimensional open‐channel flow, conducted using the computational fluid dynamics (CFD) code FLUENT, are compared with data quantifying the mean and turbulent characteristics of open‐channel flow over two contrasting gravel beds. Boundary roughness effects are represented using both the conventional wall function approach and a random elevation model that simulates the effects of supra‐grid‐scale roughness elements (e.g. particle clusters and small bedforms). Results obtained using the random elevation model are characterized by a peak in turbulent kinetic energy located well above the bed (typically at y/h = 0·1–0·3). This is consistent with the field data and in contrast to the results obtained using the wall function approach for which maximum turbulent kinetic energy levels occur at the bed. Use of the random elevation model to represent supra‐grid‐scale roughness also allows a reduction in the height of the near‐bed mesh cell and therefore offers some potential to overcome problems experienced by the wall function approach in flows characterized by high relative roughness. Despite these benefits, the results of simulations conducted using the random elevation model are sensitive to the horizontal and vertical mesh resolution. Increasing the horizontal mesh resolution results in an increase in the near‐bed velocity gradient and turbulent kinetic energy, effectively roughening the bed. Varying the vertical resolution of the mesh has little effect on simulated mean velocity profiles, but results in substantial changes to the shape of the turbulent kinetic energy profile. These findings have significant implications for the application of CFD within natural gravel‐bed channels, particularly with regard to issues of topographic data collection, roughness parameterization and the derivation of mesh‐independent solutions. Copyright © 2001 John Wiley & Sons, Ltd.     

Three‐dimensional flow structure around small‐scale bedforms in a simulated gravel‐bed environment

Pebble clusters are common small‐scale morphological features in gravel‐bed rivers, occupying as much as 10 per cent of the bed surface. Important links exist between the presence of pebble clusters and the development of flow structures. These links are poorly understood at the three‐dimensional level. Particularly neglected has been the effect of clusters on the lateral flow characteristics. A laboratory study was conducted using a hydraulic flume, within which simulated pebble clusters were superimposed onto a plane bed of gravel material. High‐resolution three‐dimensional flow data were collected above the bed at two different flow depths using an acoustic Doppler velocimeter. The results present evidence of the importance of lateral flow in the development of turbulent flow structure. Narrow regions of high lateral and downstream turbulence intensity exist to both sides of clusters and in a three‐dimensional separation zone in their lee. This may indicate the presence of horseshoe‐type vortical structures analogous to those identified in less hydraulically rough environments. However, it is likely that these structures are more complicated given the mutual interference of the surrounding medium. The lateral flow was also identified as a key component in the upwelling identified by other authors in the lee of pebble clusters. The results of the vertical flow analysis confirm the hypothesis that six regions with distinct vertical flow characteristics exist above clusters: flow acceleration up the stoss‐side of the cluster; recirculation behind the cluster in the wake region; vortex shedding from the pebble crest and shear layer; flow reattachment downstream of the cluster; upwelling of flow downstream of the point of reattachment; and recovery of flow. Copyright © 2001 John Wiley & Sons, Ltd.     

The role of channel morphology on the mobility and dispersion of bed sediment in a small gravel‐bed stream

This study uses a unique 10‐year tracer dataset from a small gravel‐bed stream to examine bed mobility and sediment dispersion over long timescales and at a range of spatial scales. Seasonal tracer data that captured multiple mobilizing events was examined, while the effects of morphology on bed mobility and sediment dispersion were captured at three spatial scales: within morphological units (unit scale), between morphological units (reach scale) and between reaches with different channel morphologies (channel scale). This was achieved by analyzing both reach‐average mobility and travel distance data, as well as the development of ‘mobility maps’ that capture the spatial variability in tracer mobility within the channel. The tracer data suggest that sediment transport in East Creek remains near critical the majority of the time, with only rare large events resulting in high mobility rates and grain travel distances large enough to move sediment past dominant bedforms. While a variable capturing both the magnitude and frequency of flow events within a season yielded a better predictor to sediment mobility and dispersion than peak discharge alone, the distribution of events of different magnitude within the season played a large role in determining tracer mobility rates and travel distances. The effects of morphology differed depending on the analysis scale, demonstrating the importance of scale, and therefore study design, when examining the effect of morphology on sediment transport. Copyright © 2016 John Wiley & Sons, Ltd.     

Persistence of the surface texture of a gravel‐bed river during a large flood

We present herein clear field evidence for the persistence of a coarse surface layer in a gravel‐bed river during flows capable of transporting all grain sizes present on the channel bed. Detailed field measurements of channel topography and bed surface grain size were made in a gravel‐bed reach of the Colorado River prior to a flood in 2003. Runoff produced during the 2003 snowmelt was far above average, resulting in a sustained period of high flow with a peak discharge of 27 m³/s (170% of normal peak flow); all available grain sizes within the study reach were mobilized in this period of time. During the 2003 peak flow, the river avulsed immediately upstream of the study reach, thereby abandoning approximately one half kilometer of the former channel. The abandonment was rapid (probably within a few hours), leaving the bed texture essentially frozen in place at the peak of the flood. All locations sampled prior to the flood were resampled following the stream abandonment. In response to the high flow, the surface median grain size (D_50s) coarsened slightly in the outer part of the bend while remaining nearly constant along the inner part of the bend, resulting in an overall increase from 18 to 21 mm for the study reach. Thus, the coarse bed surface texture persisted despite shear stresses throughout the bend that were well above the critical entrainment value. This may be explained because the response of the bed texture to increases in flow strength depends primarily upon the continued availability of the various grain size percentiles in the supply, which in this case was essentially unlimited for all sizes present in the channel. Copyright © 2007 John Wiley & Sons, Ltd.     

Flow modelling in gravel‐bed rivers: rethinking the bottom boundary condition

A key problem in computational fluid dynamics (CFD) modelling of gravel‐bed rivers is the representation of multi‐scale roughness, which spans the range from grain size, through bedforms, to channel topography. These different elements of roughness do not clearly map onto a model mesh and use of simple grain‐scale roughness parameters may create numerical problems. This paper presents CFD simulations for three cases: a plane bed of fine gravel, a plane bed of fine gravel including large, widely‐spaced pebble clusters, and a plane gravel bed with smaller, more frequent, protruding elements. The plane bed of fine gravel is modelled using the conventional wall function approach. The plane bed of fine gravel including large, widely‐spaced pebble clusters is modelled using the wall function coupled with an explicit high‐resolution topographic representation of the pebble clusters. In these cases, the three‐dimensional Reynolds‐averaged continuity and Navier–Stokes equations are solved using the standard k ? ε turbulence model, and model performance is assessed by comparing predicted results with experimental data. For gravel‐bed rivers in the field, it is generally impractical to map the bed topography in sufficient detail to enable the use of an explicit high‐resolution topography. Accordingly, an alternative model based on double‐averaging is developed. Here, the flow calculations are performed by solving the three‐dimensional double‐averaged continuity and Navier‐Stokes equations with the spatially‐averaged 〈k ? ε〉 turbulence model. For the plane bed of fine gravel including large, widely‐spaced pebble clusters, the model performance is assessed by comparing the spatially‐averaged velocity with the experimental data. The case of a plane gravel bed with smaller, more frequent, protruding elements is represented by a series of idealized hypothetical cases. Here, the spatially‐averaged velocity and eddy viscosity are used to investigate the applicability of the model, compared with using the explicit high‐resolution topography. The results show the ability of the model to capture the spatially‐averaged flow field and, thus, illustrate its potential for representing flow processes in natural gravel‐bed rivers. Finally, practical data requirements for implementing such a model for a field example are given. Copyright © 2011 John Wiley & Sons, Ltd.