Aeolian saltation: Field measurements and numerical simulations

The nature of saltation trajectories, and their relation to the wind field, was investigated at Oceano Dunes, California. Several instruments were developed for the study. The horizontal distribution of mass flux was measured using a trough-type trap divided into 35 compartments, each directing sediment to a separate electronic weighing mechanism. The vertical distribution of mass flux was measured using 15 wedge-shaped trap heads covering discrete elevation increments up to 0.35 m. A mast of anemometers was used to monitor wind speed. Instruments were sampled at 1 Hz. Data runs encompassed a wide range of shear velocities (0.26 to 0.63 ms−1).; Pre- and post-experiment calibrations demonstrated that the instruments performed well. Shear velocity estimates derived from several anemometer sub-groups were similar, indicating that anemometer arrangement does not influence the estimates. Averaging intervals as brief as 15 seconds produced robust estimates. Corrections for displacement height generated demonstrably unreasonable results in most cases, and cannot be recommended. Grain-size distributions of samples from the vertical and horizontal traps were virtually identical, and total measured transport differed by only 4%, indicating that the traps had comparable efficiencies.; Numerical simulations replicated the measured distributions of mass flux extremely well. A tendency for the simulations to under-represent mass flux at the lowest/upwind compartments was attributed to creep-mode transport. An exponential launch velocity distribution was found to provide good results, while a gamma distribution worked well only with inclusion of an additional component to account for ejected/reptating grains. Best results were obtained with mean launch speed held constant, independent of shear velocity. It was inferred that additional kinetic energy gained by saltating grains at high shear velocities is largely transferred to the bed, and the mean launch speed is controlled by the elastic limit of the bed. This hypothesis was evaluated by simulating the distributions of individual grain-size fractions, with the launch speed of each fraction determined from equal kinetic energy. Measured vertical distributions were reproduced extremely well by this approach. Although results for the horizontal distributions were somewhat weaker, the hypothesis is considered viable.