Grain‐size distribution patterns of a point bar system in the Usri River,India

Grain‐size distribution patterns in a point bar system of the Usri River, India, were critically analysed in the light of log‐normal, log‐hyperbolic and log‐skew‐Laplace distribution models. Sand samples were collected from the cross‐bedding foreset of different sizes of bedform; the objectives were to (i) study whether bedform heights have any role in grain‐size distribution patterns, (ii) offer a best‐fit statistical model, (iii) study the downstream variation of size‐sorting in a point bar system, and (iv) study the mechanism of grain sorting. The results indicate that the bedform heights have no role in grain‐size distribution patterns. Quantitatively when the errors in three distribution models were analysed, it was observed that the log‐normal distribution is the best‐fit statistical model and the next one is the log‐skew‐Laplace. However, in the upper reaches of the river, log‐normal distribution is the best‐fit model in the case of large bedforms, whereas in the lower reaches the log‐normal model is the best‐fit one in the case of small bed forms. It is also observed that within a point bar, for large and small bedforms, there is a tendency for mean grain size to decrease downstream. Between point bars for large bedforms there is no consistency in decreasing grain size downstream, whereas for small bed forms the decrease of grain size downstream is observed except near the confluence at Palkia. With distance of transport, the coarser and finer fractions of sediments are gradually chopped off. The coarser fractions are buried below the advancing bedforms on the lee sides and the finer ones are transported further downstream. Thus the finer admixture giving rise to the fining‐upward sequence overlies a carpet of coarser materials. This mechanism provides a clue to the process of grain sorting in the fluvial environment. An interpretation has been offered for the log‐normality of the grain‐size distribution pattern. During prolonged transportation in a fluvial environment, the larger grain‐size fractions are gradually chopped off and buried below the advancing bedforms on their lee sides. On the other hand, the finer fractions are transported further downstream in suspension. Thus the narrow, intermediate size fraction takes active part in the distribution patterns leading to the generation of unimodality and a symmetric distribution pattern downstream, which are the main criteria for log‐normality. Similarly, increase of bedform size is the effect of increase of stream power and Froude number leading to the selective segregation of bed materials. Thus the intermediate size fractions take a more active part than the coarser and the finer size fractions in developing log‐normality. Besides the hydrodynamic parameters of the Usri, coarsening of grain size downstream has been attributed to (i) the aggrading nature of the Usri downstream, and (ii) the contribution of coarser materials to the Usri by its tributaries and bank erosion. Copyright © 2006 John Wiley & Sons, Ltd.     

Suspended sediment transport and deposition over a dune: Río Paraná, Argentina

Field data from the Rio Paraná, Argentina, are used to examine patterns of suspended sediment transport over a sand dune. Measurements of three‐dimensional velocity are made with an acoustic Doppler current profiler whilst suspended sediment concentration and particle size have been quantified using a laser in situ sediment scattering transmissometer. Suspended sediment concentration and streamwise and vertical sediment flux are highest close to the bed, with an upward vertical flux over the stoss side of the dune and downward flux over the lee side. Suspended sediment concentrations are higher over the crest compared with the trough and suspended sediment is coarsest near the bed. About 17% of the suspended‐load transported over the crest is deposited in the lee side before it reaches the trough. Most of this deposited sand is coarser sediment that originates close to the bed over the crest, a result consistent with simulations based on the model of Mohrig and Smith (Water Resources Research 1996; 32: 3207–3217) for the excursion lengths of sediment dispersed in the lee side of a dune. Copyright © 2009 John Wiley & Sons, Ltd.     

Dune deformation in a multi‐directional wind regime: White Sands Dune Field,New Mexico

As with most dune fields, the White Sands Dune Field in New Mexico forms in a wind regime that is not unimodal. In this study, crescentic dune shape change (deformation) with migration at White Sands was explored in a time series of five LiDAR‐derived digital elevation models (DEMs) and compared to a record of wind direction and speed during the same period. For the study period of June 2007 to June 2010, 244 sand‐transporting wind events occurred and define a dominant wind mode from the SW and lesser modes from the NNW and SSE. Based upon difference maps and tracing of dune brinklines, overall dune behavior consists of crest‐normal migration to the NE, but also along‐crest migration of dune sinuosity and stoss superimposed dunes to the SE. The SW winds are transverse to dune orientations and cause most forward migration. The NNW winds cause along‐crest migration of dune sinuosity and stoss bedforms, as well as SE migration of NE‐trending dune terminations. The SSE winds cause ephemeral dune deformation, especially crestal slipface reversals. The dunes deform with migration because of differences in dune‐segment size, and differences in the lee‐face deposition rate as a function of the incidence angle between the wind direction and the local brinkline orientation. Each wind event deforms dune shape, this new shape then serves as the boundary condition for the next wind event. Shared incidence‐angle control on dune deformation and lee‐face stratification types allows for an idealized model for White Sands dunes. Copyright © 2015 John Wiley & Sons, Ltd.     

Turbulence fields in the lee of two‐dimensional transverse dunes simulated in a wind tunnel

The turbulence field of airflow in the lee of a dune has significant impacts on dune dynamics and related processes. We used particle image velocimetry in a wind tunnel simulation to obtain detailed velocity measurements in the lee of two‐dimensional transverse dune models, then used the results to analyse their turbulence fields. The dune models used in this study had a single lee angle of 30°, and a total of six stoss angles: 3°, 5°, 10°, 15°, 20° and 25°. We used vorticity, turbulence intensity, Reynolds stress and turbulent kinetic energy to characterize the turbulence fields. These parameters were functions of stoss angle, wind velocity, distance from the dune crest and height above the ground surface. The stoss angles could generally be divided into two groups based on the profiles of mean velocity, turbulence and Reynolds stress. Stoss angles of 3° and 5° usually had similar profiles, and angles of 15°, 20° and 25° formed a second group with similar profiles. The profiles for the stoss angle of 10° were usually transitional and were intermediate between the two groups. Vorticity, Reynolds stress and turbulent kinetic energy increased monotonically with increasing free‐stream wind velocity, but their variations with respect to the stoss angle were complex. The stoss angles of 15° and 20° had the maximum values of these three parameters, thus these angles may have special significance in dune development given the characteristics of the mean velocity fields and turbulence fields they produce within the lee airflow. It is the streamwise velocity component and its turbulence that determine the surface shear stress. Copyright © 2008 John Wiley and Sons, Ltd.     

Secondary flow deflection in the lee of transverse dunes with implications for dune morphodynamics and migration

Measurements of lee‐side airflow response from an extensive array of meteorological instruments combined with smoke and flow streamer visualization is used to examine the development and morphodynamic significance of the lee‐side separation vortex over closely spaced transverse dune ridges. A differential deflection mechanism is presented that explains the three‐dimensional pattern of lee‐side airflow structure for a variety of incident flow angles. These flow patterns produce reversed, along‐dune, and deflected surface flow vectors in the lee that are inferred to result in net ‘lateral diversion’ of sand transport over one dune wavelength for incident angles as small as 10° from crest‐transverse (i.e. 80° from the crest line). This lateral displacement increases markedly with incident flow angle when expressed as the absolute value of the total deflection in degrees. Reversed and multi‐directional flow occurs for incident angles between 90° and 50°. These results document the three‐dimensional nature of flow and sand transport over transverse dunes and provide empirical evidence for an oblique migration model. Copyright © 2013 John Wiley & Sons, Ltd.     

Simulation and measurement of surface shear stress over isolated and closely spaced transverse dunes in a wind tunnel

Topographic interactions generate multidirectional and unsteady air?ow that limits the application of velocity pro?le approaches for estimating sediment transport over dunes. Results are presented from a series of wind tunnel simulations using Irwin‐type surface‐mounted pressure sensors to measure shear stress variability directly at the surface over both isolated and closely spaced sharp‐crested model dunes. Findings complement existing theories on secondary air?ow effects on stoss transport dynamics and provide new information on the in?uence of lee‐side air?ow patterns on dune morphodynamics. For all speeds investigated, turbulent unsteadiness at the dune toe indicates a greater, more variable surface shear, despite a signi?cant drop in time‐averaged measurements of streamwise shear stress at this location. This effect is believed suf?cient to inhibit sediment deposition at the toe and may be responsible for documented intermittency in sand transport in the toe region. On the stoss slope, streamline compression and ?ow acceleration cause an increase in ?ow steadiness and shear stress to a maximum at the crest that is double that at the toe of the isolated dune and 60–70 per cent greater than at ?ow reattachment on the lower stoss of closely spaced dunes. Streamwise ?ow accelerations, rather than turbulence, have greater in?uence on stress generation on the stoss and this effect increases with stoss slope distance and with incident wind speed. Reversed ?ow within the separation cell generates signi?cant surface shear (30–40 per cent of maximum values) for both spacings. This supports ?eld studies that suggest reversed ?ow is competent enough to return sediment to the dune directly or in a de?ected direction. High variability in shear at reattachment indicates impact of a turbulent shear layer that, despite low values of time‐averaged streamwise stress in this region, would inhibit sediment accumulation. Downwind of reattachment, shear stress and ?ow steadiness increase within 6 h (h = dune height) of reattachment and approach upwind values by 25 h. A distance of at least 30 h is suggested for full boundary layer recovery, which is comparable to ?uvial estimates. The Irwin sensor used in this study provides a reliable means to measure skin friction force responsible for sand transport and its robust, simple, and cost‐effective design shows promise for validating these ?ndings in natural dune settings. Copyright © 2003 John Wiley & Sons, Ltd.     

Offshore aeolian sediment transport across a human‐modified foredune

Concepts derived from previous studies of offshore winds on natural dunes are evaluated on a dune maintained for shore protection during three offshore wind events. The potential for offshore winds to form a lee‐side eddy on the backshore or transfer sediment from the dune and berm crest to the water are evaluated, as are differences in wind speed and sediment transport on the dune crest, berm crest and a pedestrian access gap. The dune is 18–20 m wide near the base and has a crest 4.5 m above backshore elevation. Two sand‐trapping fences facilitate accretion. Data were obtained from wind vanes on the crest and lee of the dune and anemometers and sand traps placed across the dune, on the beach berm crest and in the access gap. Mean wind direction above the dune crest varied from 11 to 3 deg from shore normal. No persistent recirculation eddy occurred on the 12 deg seaward slope. Wind speed on the berm crest was 85–89% of speed at the dune crest, but rates of sediment transport were 2.27 times greater during the strongest winds, indicating that a wide beach overcomes the transport limitation of a dune barrier. Limited transport on the seaward dune ramp indicates that losses to the water are mostly from the backshore, not the dune. The seaward slope gains sand from the landward slope and dune crest. Sand fences causing accretion on the dune ramp during onshore winds lower the seaward slope and reduce the likelihood of detached flows during offshore winds. Transport rates are higher in access gaps than on the dune crest despite lower wind speeds because of flatter slopes and absence of vegetation. Transport rates across dunes and through gaps can be reduced using vegetation and raised walkover structures. Copyright © 2017 John Wiley & Sons, Ltd.     

Understanding river dune splitting through flume experiments and analysis of a dune evolution model

Forecasts of water level during river floods require accurate predictions of the evolution of river dune dimensions, because the hydraulic roughness of the main channel is largely determined by the bed morphology. River dune dimensions are controlled by processes like merging and splitting of dunes. Particularly the process of dune splitting is still poorly understood and – as a result – not yet included in operational dune evolution models. In the current paper, the process of dune splitting is investigated by carrying out laboratory experiments and by means of a sensitivity analysis using a numerical dune evolution model. In the numerical model, we introduced superimposed TRIAS ripples (i.e. triangular asymmetric stoss side‐ripples) on the stoss sides of underlying dunes as soon as these stoss sides exceed a certain critical length. Simulations with the model including dune splitting showed that predictions of equilibrium dune characteristics were significantly improved compared to the model without dune splitting. As dune splitting is implemented in a parameterized way, the computational cost remains low which means that dune evolution can be calculated on the timescale of a flood wave. Subsequently, we used this model to study the mechanism of dune splitting. Literature showed that the initiation of a strong flow separation zone behind a superimposed bedform is one of the main mechanisms behind dune splitting. The flume experiments indicated that besides its height also the lee side slope of the superimposed bedform is an important factor to determine the strength of the flow separation zone and therefore is an important aspect in dune splitting. The sensitivity analysis of the dune evolution model showed that a minimum stoss side length was required to develop a strong flow separation zone. Copyright © 2014 John Wiley & Sons, Ltd.     

A field study of mean and turbulent flow characteristics upwind,over and downwind of barchan dunes

Field‐measured patterns of mean velocity and turbulent airflow are reported for isolated barchan dunes. Turbulence was sampled using a high frequency sonic anemometer, deriving near‐surface Reynolds shear and normal stresses. Measurements upwind of and over a crest‐brink separated barchan indicated that shear stress was sustained despite a velocity reduction at the dune toe. The mapped streamline angles and enhanced turbulent intensities suggest the effects of positive streamline curvature are responsible for this maintenance of shear stress. This field evidence supports an existing model for dune morphodynamics based on wind tunnel turbulence measurements. Downwind, the effect of different dune profiles on flow re‐attachment and recovery was apparent. With transverse incident flow, a re‐attachment length between 2·3 and 5·0h (h is dune brink height) existed for a crest‐brink separated dune and 6·5 to 8·6h for a crest‐brink coincident dune. The lee side shear layer produced elevated turbulent stresses immediately downwind of both dunes, and a decrease in turbulence with distance characterized flow recovery. Recovery of mean velocity for the crest‐brink separated dune occurred over a distance 6·5h shorter than the crest‐brink coincident form. As the application of sonic anemometers in aeolian geomorphology is relatively new, there is debate concerning the suitability of processing their data in relation to dune surface and streamline angle. This paper demonstrates the effect on Reynolds stresses of mathematically correcting data to the local streamline over varying dune slope. Where the streamline angle was closely related to the surface (windward slope), time‐averaged shear stress agreed best with previous wind tunnel findings when data were rotated along streamlines. In the close lee, however, the angle of downwardly projected (separated) flow was not aligned with the flat ground surface. Here, shear stress appeared to be underestimated by streamline correction, and corrected shear stress values were less than half of those uncorrected. Copyright © 2011 John Wiley & Sons, Ltd.     

Turbulent flow over a dune: Green River,Colorado

Detailed echo‐sounder and acoustic Doppler velocimeter measurements are used to assess the temporal and spatial structure of turbulent flow over a mobile dune in a wide, low‐gradient, alluvial reach of the Green River. Based on the geometric position of the sensor over the bedforms, measurements were taken in the wake, in transitional flow at the bedform crest, and in the internal boundary layer. Spatial distributions of Reynolds shear stress, turbulent kinetic energy, turbulence intensity, and correlation coefficient are qualitatively consistent with those over fixed, two‐dimensional bedforms in laboratory flows. Spectral and cospectral analysis demonstrates that energy levels in the lee of the crest (i.e. wake) are two to four times greater than over the crest itself, with minima over the stoss slope (within the developing internal boundary layer). The frequency structure in the wake is sharply defined with single, dominant peaks. Peak and total spectral and cross‐spectral energies vary over the bedform in a manner consistent with wave‐like perturbations that ‘break’ or ‘roll up’ into vortices that amalgamate, grow in size, and eventually diffuse as they are advected downstream. Fluid oscillations in the lee of the dune demonstrate Strouhal similarity between laboratory and field environments, and correspondence between the peak frequencies of these oscillations and the periodicity of surface boils was observed in the field. Copyright © 2005 John Wiley & Sons, Ltd.     

Early‐stage aeolian protodunes: Bedform development and sand transport dynamics

Early‐stage aeolian bedforms, or protodunes, are elemental in the continuum of dune development and act as essential precursors to mature dunes. Despite this, we know very little about the processes and feedback mechanisms that shape these nascent bedforms. Whilst theory and conceptual models have offered some explanation for protodune existence and development, until now, we have lacked the technical capability to measure such small bedforms in aeolian settings. Here, we employ terrestrial laser scanning to measure morphological change at the high frequency and spatial resolution required to gain new insights into protodune behaviour. On a 0.06 m high protodune, we observe vertical growth of the crest by 0.005 m in two hours. Our direct measurements of sand transport on the protodune account for such growth, with a reduction in time‐averaged sediment flux of 18% observed over the crestal region. Detailed measurements of form also establish key points of morphological change on the protodune. The position on the stoss slope where erosion switches to deposition is found at a point 0.07 m upwind of the crest. This finding supports recent models that explain vertical dune growth through an upwind shift of this switching point. Observations also show characteristic changes in the asymmetric cross‐section of the protodune. Flow‐form feedbacks result in a steepening of the lee slope and a decline in lower stoss slope steepness (by 3°), constituting a reshaping of protodune form towards more mature dune morphology. The approaches and findings applied here, (a) demonstrate an ability to quantify processes at requisite spatial and temporal scales for monitoring early‐stage dune evolution, (b) highlight the crucial role of form‐flow feedbacks in enabling early‐stage bedform growth, alluding to a fluctuation in feedbacks that require better representation in dune models, and (c) provide a new stimulus for advancing understanding of aeolian bedforms. Copyright © 2017 John Wiley & Sons, Ltd.     

Aeolian dynamics over a coastal foredune,Prince Edward Island,Canada

Near‐surface airflow over a morphologically simple, vegetated, 8 m high foredune with a small wave‐cut scarp was measured for onshore to oblique‐onshore conditions during a low‐moderate (5–6 m s^‐1 ) wind event and a high velocity (11–18 m s^‐1) sand‐transporting gale event. Flow across the foredune was characterized by significant flow compression and acceleration up and across the foredune during both events. During the gale, a pronounced jet (speed bulge) developed at the foredune crest, which increased in magnitude with increasing wind speed. The vertical (W) velocity component of the 3D flow field was positive (upwards) across the stoss slope under low wind conditions but negative (downwards) during gale wind conditions, with upslope acceleration. During the low velocity event, there was speed‐down within the vegetation canopy, as would be expected for a porous roughness cover. During the strong wind event there was speed‐up in the lower portion of the vegetation canopy, and this was found up the entire stoss slope. Sediment transport during the gale force event was substantial across the beach and foredune despite the moderate vegetation cover and minimum fetch. Aeolian suspension was evident in the lee of the dune crest. The observations presented herein show that strong storm winds are an effective mechanism for translating sediment landwards across a high vegetated foredune, contributing sediment to the stoss slope, crest and leeward slopes of the foredune and backing dunes. Copyright © 2013 John Wiley & Sons, Ltd.     

Lateral migration,thermoluminescence chronology and colour variation of longitudinal dunes near Birdsville in the Simpson Desert,central Australia

The longitudinal dunes of the Simpson Desert, in the vicinity of Birdsville, have been reworked largely during the Holocene from dunes deposited up to 80000 years ago or earlier. The widespread asymmetry of these roughly northward-trending dunes, with steeper eastern faces and more gentle western faces, supports wind-rose data showing sand-transporting winds from the southwest obliquely intersecting the dunes. While this suggests a change in the wind pattern since the dune field was oriented, it does not indicate that the dunes are necessarily shifting leeward (eastward) as a consequence. It is hypothesized that the direction of migration is controlled by the extent to which the dunes are vegetated. Relatively well-vegetated dunes can accrete sand on their gentle stoss slopes and erode on their lee slopes causing them to shift westward and hence obliquely into the wind, a condition that probably prevails in wetter regions and during episodes of relatively humid climate. In contrast, in very dry areas or during arid phases, sand can move unimpeded up a sparsely vegetated stoss face and over the crest to form an avalanche or slip face on the lee side, thereby causing the dunes to shift eastward. Despite evidence that longitudinal dune crests can shift laterally to some extent, the dunes in the western part of the Simpson Desert have not migrated, either westward or eastward, more than 100m or so from their Pleistocene cores. Aeolian transport and partial or complete removal of iron cutans from around quartz grains results in dunes of widely varying colour yet of similar age.     

Athabasca Glacier,Canada – a field example of subglacial ice and till erosion?

The Athabasca Glacier, resting on a rigid bed, provides an excellent example of subglacial ice and till erosion. The presence of a thin mobile till layer is shown by the presence of flutes, saturated till layer, push moraines and ploughed boulders. Cross‐cutting striations, v‐shaped striations and reversed stoss‐and‐lee clasts are indicative of clasts rotating within this layer. As the till moves it erodes the bedrock and clasts within it. A combination of erosion by ice and till produces stoss‐and‐lee‐clasts and generates striations on flutes and embedded clasts, as well as eroding the bedrock into a continuum of smoothed, rounded and streamlined forms. Copyright © 2005 John Wiley & Sons, Ltd.     

Modeling river dune development and dune transition to upper stage plane bed

Large asymmetric bedforms known as dunes commonly dominate the bed of sand rivers. Due to the turbulence generation over their stoss and lee sides, dunes are of central importance in predicting hydraulic roughness and water levels. During floods in steep alluvial rivers, dunes are observed to grow rapidly as flow strength increases, undergoing an unstable transition regime, after which they are washed out in what is called upper stage plane bed. This transition of dunes to upper stage plane bed is associated with high transport of bed sediment in suspension and large decrease in bedform roughness. In the present study, we aim to improve the prediction of dune development and dune transition to upper stage plane bed by introducing the transport of suspended sediment in an existing dune evolution model. In addition, flume experiments are carried out to investigate dune development under bed load and suspended load dominated transport regimes, and to get insight in the time scales related to the transition of dunes to upper stage plane bed. Simulations with the extended model including the transport of suspended sediment show significant improvement in the prediction of equilibrium dune parameters (e.g. dune height, dune length, dune steepness, dune migration rate, dune lee side slope) both under bed load dominant and suspended load dominant transport regimes. The chosen modeling approach also allows us to model the transition of dunes to upper stage plane bed which was not possible with the original dune evolution model. The extended model predicts change in the dune shapes as was observed in the flume experiments with decreasing dune heights and dune lee slopes. Furthermore, the time scale of dune transition to upper stage plane bed was quite well predicted by the extended model. Copyright © 2015 John Wiley & Sons, Ltd.     

Spatio-temporal variation of transition probability of bursting events over the ripples at the bed of open channel

In this study, the turbulent flow structure over the ripples at the bed of open channel was investigated experimentally. An analysis of bursting process was used to recognize the susceptible regions for sediment entrainment and deposition over the ripples. Two artificial ripples were built according to the ripples, formed by nature. Three different configurations of ripples were used by changing the wavelength. According to analysis of bursting event it was found that at the stoss side of ripples, quadrants (II) and (IV) were dominant to the quadrants (I) and (III) and at the lee side of the ripple it was vice versa. Also the transition probabilities of the bursting events were determined. The results showed that stable organizations of each class of the events had highest transition probabilities whereas cross organizations had lowest transition probabilities. Additionally, an effort was made to find the average inclination angle of the bursting events in quadrants (II) and (IV). The results showed that the mean angle of events in quadrants (II) and (IV) increases at the downstream of stoss side to the crest in each experimental test. Also, at the lee side where the sediment particles were deposited, the inclination angles had the highest values.     

Monitoring surface change on a Namib linear dune

In tackling the apparently intractable problem of linear dune initiation and maintenance there has been a move away from large-scale deductive models to smaller-scale field studies of individual dunes. This paper reports a study of surface change on a large, complex linear dune in the Namib Desert, southern Africa. The dune surface responds to a markedly seasonal wind regime. In summer westerly winds erode sand from the west flank of the dune and deposit it on the easterly lee side of the dune crest. In winter this pattern is reversed. Easterly winds erode sand from the east slope and deposit it on the west slope. The crest therefore moves back and forth some 15 m each year returning at the end of a year's cycle to its position at the beginning. The position of the base of the dune appears to remain fixed, even though sand is moving throughout the dune system. The dune does extend northward along some resultant of the westerly and easterly winds. Despite relatively high levels of activity, especially at the dune crest, there is no evidence of the breakdown of the linear dune form. The conclusion must therefore be that linear dunes can be maintained in bimodal wind regimes and are not necessarily related to unidirectional parallel regimes as others have suggested.     

Secondary airflow and sediment transport in the lee of a reversing dune

Lee-side windspeed and sediment transport were measured over a small (1·2 m) transverse ridge in the Silver Peak dunefield, west-central Nevada, USA, using an intensive array of 25 cup anemometers and seven total flux traps. During crest-transverse and transporting flow conditions (u_0·3crest ≈ 8·4 m s⁻¹), windspeed near the surface of the lee slope averaged half (48 per cent) that of crest speeds. Dimensionless speeds in the separation zone ranged from 0·2 to 0·8 that of the outer flow (u₁₂). Along the boundary of the separation cell, windspeed increased by 10 per cent of the crest speed before separation. Equilibrium of upper and lower wake regions was not observed by the documented eight dune heights, suggesting that wake recovery may not occur over closely spaced dunes. Sediment transport measured directly on both the lee slope and interdune surfaces averaged approximately 15 per cent of crest inputs. This suggests that a significant amount (c. 70–95 per cent) of sediment transported over the crest moved as fallout. For this data set, flux was approximately proportional to the cube of the near-surface windspeed (u_0·3) and in general there was an order of magnitude difference between flux measured at the crest and that measured within the separation zone. Transport direction in the separation zone was acutely oblique to the incident direction owing to secondary flow deflection. Beyond the interdune, transport direction progressed from oblique to crest-transverse. This indicates that an appreciable amount of sediment may move laterally along the lee slope and interdune corridor under crest-transverse flows. Regarding the grain size and sorting properties of transported sediment, there was no significant difference in mean grain size over the dune, although in general particles were finer and more poorly sorted in the lee. Copyright © 1999 John Wiley & Sons, Ltd.