Intermittent Aeolian Saltation: A Protocol For Quantification

Abstract

The quantification of saltation intermittency-the proportion of time when sand transport occurs-provides valuable information about the nature of aeolian systems, including insights concerning threshold conditions for sand transport. Intermittency has been measured in numerous studies, but a lack of measurement standardization often makes the comparison of results difficult. Four methodological factors influence estimates of intermittency: sample frequency, sample duration, sample area, and sensor elevation. For a given transport rate, the value of intermittency may decrease with sensor elevation, increased sample frequency, or decreased sensor area; it may increase or decrease as a consequence of sample duration. We therefore simulate the influences of these factors and suggest a common protocol for measuring and reporting intermittency based on measurements at 1 Hz with 100 mm 2 sensor area at 50 mm elevation above the surface, or equivalent values.

Distinct Thresholds for the Initiation and Cessation of Aeolian Saltation From Field Measurements

“…However, such estimates of threshold values and their effects on saltation occurrence may be complicated by uncertainties in measurements of wind speeds, turbulence properties, soil moisture, and sediment size distributions. In addition, translation of saltation activity and threshold values among different modes of investigation (i.e., observational, experimental, theoretical, and numerical) is fraught with issues of measurement scale and methodology (e.g., Barchyn, Martin, et al, ; Martin et al, ), leading to suggestions for standardized analysis protocols (Barchyn et al, ; Ellis et al, ; Sherman et al, ). Further work is needed to improve measurement techniques, standardize analysis protocols, and constrain the role of environmental factors in the determination of threshold values.…”

“…Due to counting uncertainty for the small sampling volume of Wenglor sensors counting the passage of individual particles over short time intervals, we apply a slight correction to f Q to account for the possibility of false negatives (see Appendix A). For each Δ t , we then calculate effective threshold wind speed u th by applying the TFEM (Davidson‐Arnott et al, ; Sherman et al, ; Stout, ; Stout & Zobeck, ; Wiggs, Atherton, & Baird, ): where Φ u (1 − f Q ) is the value in the cumulative distribution of wind speeds Φ u corresponding to the time fraction of inactive saltation, 1 − f Q (Figure f). When calculating Φ u for each Δ t , we use the δt‐averaged u values, in correspondence with our methods for calculating f Q .…”

“…Due to counting uncertainty for the small sampling volume of Wenglor sensors counting the passage of individual particles over short time intervals, we apply a slight correction to f Q to account for the possibility of false negatives (see Appendix A). For each Δt, we then calculate effective threshold wind speed u th by applying the TFEM (Davidson-Arnott et al, 2012;Sherman et al, 2017;Stout, 2004;Stout & Zobeck, 1997;:…”

Turbulent characteristics of saltation and uncertainty of saltation model parameters

“…In this study, we apply the Differential Evolution Adaptive Metropolis (DREAM) algorithm proposed by Vrugt et al (2011) for estimation of hydrologic model parameters. The algorithm integrates differential evolution (Storn and Price, 1997) and self-adaptive randomised subspace sampling to accelerate a Markov chain Monte Carlo simulation. A full description of the DREAM algorithm is beyond the scope of our study.…”

“…Turbulent saltation has attracted much attention in more recent years (e.g. McKenna-Neuman et al, 2000;Davidson-Arnott and Bauer, 2009;Sherman et al, 2017) and large-eddy simulation models have been under development to model the process (e.g. Dupont et al, 2013).…”