Search

International audience; We review selected experimental saltation studies performed in laboratory wind tunnels and collision experiments performed in (splash-) laboratory facilities that allow detailed observations between impinging particles on a stationary bed.We also discuss progress in understanding aeolian transport in nonterrestrial environments. Saltation studies in terrestrial wind tunnels can be divided into two groups. The first group comprises studies using a short test bed, typically 1–4m long, and focuses on the transitional behavior near the upwind roughness discontinuity where saltation starts. The other group focuses on studies using long test beds — typically 6 m or more — where the saturated saltation takes place under equilibrium conditions between wind flow and the underlying rough bed. Splash studies using upscaled model experiments allow collision simulations with large spherical particles to be recorded with a high speed video camera. The findings indicate that the number of ejected particles per impact scales linearlywith the impact velocity of the saltating particles. Studies of saturated saltation in several facilities using predominantly Particle Tracking Velocimetry or Laser Doppler Velocimetry indicate that the velocity of the (few) particles having high trajectories increases with increasing friction velocity. However, the speed of the majority of particles that do not reachmuch higher than Bagnold's focal point is virtually independent of Shields parameter—at least for lowor intermediate u⁎-values. In this case mass flux depends on friction velocity squared and not cubed as originally suggested by Bagnold. Over short beds particle velocity shows stronger dependence on friction velocity and profiles of particle velocity deviate from those obtained over long beds. Measurements using horizontally segmented traps give average saltation jump-lengths near 60–70 mm and appear to be only weakly dependent on friction velocity, which is in agreement with some, but not all, older or recent wind tunnel observations. Similarly some measurements performed with uniform sand samples having grain diameters of the order of 0.25–0.40mmindicate that ripple spacing depends on friction velocity in a similar way as particle jump length. The observations are thus in agreementwith a recent ripple model that link the typical jump length to ripple spacing. A possible explanation for contradictory observations in some experiments may be that long observation sequences are required in order to assure that equilibrium exists between ripple geometry and wind flow.Quantitative understanding of saltation characteristics onMars still lacks important elements. Based upon image analysis and numerical predictions, aeolian ripples have been thought to consist of relatively large grains (diameter N 0.6mm) and that saltation occurs at high wind speeds (N26 m/s) involving trajectories that are significantly longer than those on Earth (by a factor of 10–100). However, this is not supported by recent observations from the surface of Mars, which shows that active ripples in their geometry and composition have characteristics compatible with those of terrestrial ripples (Sullivan et al., 2008). Also the highest average wind speeds on Mars have been measured to be b20 m/s, with even turbulent gusts not exceeding 25 m/s. Electrification is seen as a dominant factor in the transport dynamics of dust onMars, affecting the structure, adhesive properties and detachment/entrainment mechanisms specifically through the formation of aggregates (Merrison et al., 2012). Conversely for terrestrial conditions electric fields typically observed are not intense enough to significantly affect sand transport rates while little is known in the case of extra-terrestrial environments.