More than 4000 hourly wind profiles measured on three topographically different foredunes are analysed and discussed. Wind flow over the foredunes is studied by means of the relative wind speed: the ratio between wind speed at a certain location and the reference wind speed at the same height. Relative wind speeds appear to be independent of general wind speed but dependent on wind direction. For perpendicular onshore winds the flow over the foredune is accelerated due to topographic changes and decelerated due to changes in surface roughness. Accelerations dominate over decelerations on the seaward slope. The pattern of acceleration and deceleration in relation to wind direction is more or less comparable for different foredunes, but the magnitudes differ. An increase in foredune height from 6 to 10m leads to an increase in speed-up near the top of the seaward slope from 110 to 150 per cent during onshore wind, but further increase of foredune height from 10 to 23 m appears to have little effect, due to increased roughness and deflection of flow.Topography also influences the direction of the flow. Between beach and top, the flow deflects in the direction of the normal during onshore winds. During offshore winds the flow is deflected to the parallel. Near the dunefoot, deflection is always in the direction of the parallel, and increases with steeper topography. The maximum deflection near the dunefoot was 90", over a 23 m high dune, observed during offshore winds.Patterns of erosion and sedimentation resulting from winds from different directions can be explained by the observed accelerations and decelerations. Owing to speed-up on the seaward front of the foredune, sand transport capacity of the wind increases, which results in erosion if vegetation is absent. During strong onshore wind, sand is lifted near the dunefoot and moves over the foredune in suspension. During weaker winds, vertical wind velocities do not exceed fall velocities of the sand grains, and most of the sand is deposited near the dunefoot.

Soil loss caused by wind erosion is a widespread phenomenon in the Sahelian zone of WestAfrica. According to Sahelian farmers, scattered vegetation standing in amongst the crop has the potential for a wind erosion control strategy. This study was conducted to study the effect of single vegetation elements on the pattern of average wind speed and sediment transport. This was done by two experiments that were carried out during the rainy seasons of 2002 and 2003 in north Burkina Faso, West Africa. Wind speeds were measured using three sonic anemometers, at a sampling frequency of 16 Hz. Sediment transport was determined by calculating the mass fluxes from 17 MWAC catchers. In this study, a shrub was defined as a vegetation element with branches until ground and a tree as a vegetation element with a distinctive trunk below a canopy.Behind shrubs wind speed near the soil surface was reduced up to approximately seven times the height of the shrub. The observed reduction in wind speed in the area where wind speed was reduced was 15 per cent on average. At the sides of the shrub, wind speed was increased, by on average 6 per cent. As the area of increase in wind speed is one-third of the area of decrease in wind speed, the net effect of a shrub is a reduction in wind speed. A similar pattern was visible for the pattern of sediment transport around a shrub. Downwind of a shrub, sediment transport was diminished up to seven times the height of the shrub. Probably most of this material was trapped by the shrub. Trees showed a local increase of wind around the trunk, which is expected to relate to an increase in sediment transport around the trunk. Mass flux measurements of sediment transport were not made, but visual observations in the field substantiate this. Behind the canopy of a tree, a tree acts similarly to a shrub regarding its effects on average wind speed, but as a tree is generally a larger obstacle than a shrub the extent of this effect is larger than for shrubs. Thus, whereas shrubs are more effective than trees regarding their direct effect on soil loss by trapping sand particles near the soil surface, trees are more effective in affecting soil loss indirectly by reducing the wind speed downwind more effectively than shrubs. Therefore, to reduce soil loss in an area, the presence of both trees and shrubs is crucial.

Fast response wind and turbulence instruments, including sonic anemometers are used more and more in the research of aeolian sediment transport. These instruments provide data on mean wind, but also on friction velocity, wind speed fluctuations and turbulence statistics, such as the U-W and W-T covariance (C UW and C WT ), which are a measure for the momentum flux and the sensible heat flux. This short paper will examine two problems that may arise when using sonic anemometers, namely the low/high frequency losses and the interpretation of the sonic anemometer measurements over sloping or non-homogeneous terrain.

The effect of turbulent flow structures on saltation sand transport was studied during two convective storms in Niger, West Africa. Continuous, synchronous measurements of saltation fluxes and turbulent velocity fluctuations were made with a sampling frequency of 1Hz. The shear stress production was determined from the vertical and streamwise velocity fluctuations. The greatest stress-bearing events were classified as turbulent structures, with sweep, ejection, inward interaction, and outward interaction described according to the quadrant technique. The classified turbulent structures accounted for 63·5 per cent of the average shear stress during the first storm, and 56·0 per cent during the second storm. The percentage of active time was only 20·6 per cent and 15·8 per cent, respectively. High saltation fluxes were associated with sweeps and outward interactions. These two structures contribute positively (sweeps) and negatively (outward interactions) to the shear stress, but have in common that the streamwise velocity component is higher than average. Therefore, the horizontal drag force seems primarily responsible for saltation sand transport, and not the shear stress. This was also reflected by the low correlation coefficients (r) between shear stress and saltation flux (0·12 and 0·14, respectively), while the correlation coefficients between the streamwise velocity component and saltation flux were much higher (0·65 and 0·57, respectively).

Wind erosion is an important soil erosion and hence a soil degradation problem in the Sahelian zone of West Africa. Potentially, the characteristic dryland vegetation with scattered trees and shrubs can provide for soil erosion protection from wind erosion, but so far adequate quantification of vegetation impacts is lacking. The aim of this study was to develop a model of windblown soil erosion and sediment transport around a single shrub-type vegetation element. Starting with the selection of a suitable transport equation from four possible sediment transport equations, the effects of a single vegetation element on wind speed were parameterized. The modified wind speed was then applied to a sediment transport equation to model the change in sediment mass flux around a shrub. The model was tested with field data on wind speed and sediment transport measured around isolated shrubs in a farmer's field in the north of Burkina Faso. The simple empirical equation of Radok (Journal of Glaciology 19: 123-129, 1977) performed best in modelling soil erosion and sediment transport, both for the entire event duration and for each minute within an event. Universal values for the empirical constants in the sediment transport equation could not be obtained because of the large variability in soil and roughness characteristics. The pattern of wind speed, soil erosion and sediment transport behind a shrub and on either side of it was modelled. The wind speed changed in the lee of the vegetation element depending on its porosity, height and downwind position. Wind speed was recovered to the upstream speed at a downwind distance of 7·5 times the height of the shrub. The variability in wind direction created a 'rotating' area of influence around the shrub. Compared to field measurements the model predicted an 8% larger reduction in sediment transport in the lee of the vegetation element, and a 22% larger increase beside the vegetation element.