Evolution of Turbulent Boundary Conditions on the Surface of Large Barchan Dunes: Anomalies in Aerodynamic Roughness and Shear Velocity, Aeolian Threshold, and the Role of Dune Skewness

We recorded aerodynamic roughness and shear velocity along transects on and around mature crescent‐shaped barchan dunes of 4.5m $4.5\ \mathrm{m}$ and 27m $27\ \mathrm{m}$ height above the horizontal rock‐covered Qatar desert by fitting to the log‐law time‐averaged vertical velocity profiles acquired from triads of ultrasonic anemometers penetrating the inner turbulent boundary layer. Shear velocity first decreased, then recovered as air climbed on the dune, with a local maximum ahead of the crest as predicted by the Jackson and Hunt (1975, https://doi.org/10.1002/qj.49710143015) theory. Unlike flows over gentler bedforms without a slope discontinuity, an anomalous peak of shear velocity also arose on the dune centerline at the brink, which the theory attributed to skewness in the dune transect profile. The onset of aeolian transport produced a log‐law passing through the Bagnold (1941, https://doi.org/10.1007/978‐94‐009‐5682‐7) focal point. It was bracketed by noticeable hysteretic peaks in the correlation between wind speed and entrained sand flux. The dunes' rocky surroundings and topography produced an aerodynamic roughness at odds with the Nikuradse (1933, https://ntrs.nasa.gov/citations/19930093938) data for fully developed turbulent boundary layers. Large‐eddy numerical simulations illustrated the sensitivity of shear velocity to wide changes in aerodynamic roughness from desert floor to dune surface. Plain Language Summary Wind friction is the engine that erodes sand dunes, relentlessly pushing them over roads, houses, and infrastructure. Our records of wind speed on crescent‐shaped mobile dunes challenge the conventional understanding of this process. Comparing field measurements and models, we show that the highest friction occurs where the gentle upward dune surface abruptly gives way to a steeper avalanching downward slope. Our data also reveals that the “aerodynamic roughness,” a measure of wind friction on sand, is at odds with historical data meant for turbulent pipe flow. Because numerical simulations are used to predict flow over landforms that are inaccessible to detailed measurements, we validate them against data on a large dune. Our observations imply that, to achieve greater fidelity, simulations should subdivide the fluid neighborhood of the dune more finely, and revisit how they treat aerodynamic friction on its surface. Although our work involved large desert dunes, we expect these suggestions to apply more broadly to atmospheric, f