Timescale dependence of aeolian sand flux observations under atmospheric turbulence

Abstract

[1] The transport of sand in saltation is driven by the persistently unsteady stresses exerted by turbulent winds. Based on coupled high-frequency observations of wind velocity and sand flux on a desert dune during intermittent saltation, we show here how observations of saltation by natural winds depend significantly on the timescale and method used for determining shear stress and sand flux. The correlation between sand flux and excess shear stress (stress above a threshold value) systematically improves for longer averaging timescale, T, and is better for stress determined by the law-of-the-wall versus the Reynolds stress method. Fitting parameters for the stress-flux relationship do not converge with increasing T, which may be explained by the nonstationary nature of wind velocity statistics. We show how it may be possible, based on the scale-dependent statistics of stress fluctuations, to rescale saltation flux predictions for wind observations made at different timescales. However, our observations indicate hysteresis and time lags in thresholds for initiation and cessation of saltation, which complicate threshold-based approaches to predicting sediment transport at different timescales.

Scaling of Dust Flux With Friction Velocity: Time Resolution Effects

“…The reduced correlation between F wd and u * at small time resolution corroborates the correlation reduction observed between the saltation flux and u * by Sterk et al (1998), Shao and Mikami (2005), or Martin et al (2013), although the methods used by these previous studies were less suited than the wavelet method to estimate u * at small time scale, as explained in section 2.1. Shao (2008) suggested that this reduced correlation between saltation flux and near-surface u * at small time scale could be either explained (1) by the difference at small scales between near-surface and surface values of u * or (2) by the absence of saltation equilibrium state at small time scales where the second Owen's (1964) hypothesis of the skin friction equal to the fluid threshold shear stress would not remain.…”

“…This is explained by variations of mesoscale background conditions, adding to the diurnal cycle forcing. This explains that an increasing number of studies focused on the smaller time scale behavior (<10 min) of the saltation flux in regard to the instantaneous shear stress and wind velocity, added to the fact that high-frequency sensors became more affordable these days (Bauer et al, 1998;Butterfield, 1991;Martin et al, 2013;Namikas et al, 2003). Even so, current parametrizations of the saltation flux still differ substantially from one to the other (Sherman and Bailiang, 2012), with 10.1029/2019JD031192 systematic differences between wind-tunnel and field measurements (Raupach, 1991).…”

Large Effects of Particle Size Heterogeneity on Dynamic Saltation Threshold

“…Experiments further revealed that, above the region of strongly suppressed near‐surface wind, there is a focal region at which the wind velocities are nearly constant with u * (the Bagnold focus ): (Bagnold, ), where z f and u f are constants. This focal point approximation is equivalent to an exponentially increasing surface roughness, , as found in simulations (Durán et al, , ) and measurements (Creyssels et al, ; Ho et al, ; Martin et al, ) of saturated saltation transport. In contrast, z o changes much more slowly with u * when saltation transport is strongly undersaturated or absent: z o ≃const.…”

“…In contrast to the static threshold above which saltation can be initiated (e.g., Bagnold, ; Burr et al, ; Chepil, ; de Vet et al, ; Gillette et al, ; Iversen & Rasmussen, ; Iversen et al, ; Merrison et al, ; Nickling, ; Raffaele et al, , and references therein), u t has only rarely been systematically studied in controlled laboratory settings, especially in recent history (there are numerous poorly controlled field studies though; Barchyn & Hugenholtz, ; Li et al, ; Martin & Kok, , ; Martin et al, , and references therein). Even today, we largely rely on the old data sets by Bagnold () and Chepil (), who measured u t by visual means.…”

Unsteady saltation on Mars

“…Most importantly, sand transport could still be observed within long-time runs even though the fitted values of bed shear stress using conventional laboratory equations for mass transport predicted that there should be no transport (Rasmussen and Sørensen, 1999). By analyzing the observed data, many believe that the quasi-coherent structure in the turbulent wind in atmospheric surface layer is crucial to the intermittent saltation (e.g., Bauer et al, 1998;Baas, 2007;Martin et al, 2013), and the variability should be incorporated into the classical steady transport models. In the wind tunnel, Butterfield (1998) reported an obvious increase of induced transport which was much higher than the measured value in steady wind when imposing sinusoidal wind velocity variations with periods of 6-20 s. The results demonstrated the significant effects of unsteady wind on sand transport rate.…”