Predicting aeolian sand transport rates: A reevaluation of models

Eight aeolian sand-transport models are evaluated using a field data set and a new approach to estimating shear velocity. The models are those of Bagnold, 1937, Kawamura, 1951, Zingg, 1953, Owen, 1964, Kadib, 1965, Hsu, 1971, Lettau and Lettau, 1978 and Sørensen, 2004. All of these models predict transport as a function of shear velocity to the third power. Shear velocities are estimated using wind profile data (log-linear slope) with the von Kármán constant and with the apparent von Kármán parameter and the results of the different approaches are evaluated based on comparison of regression statistics and RMS error. The models were not adjusted to account for sediment moisture content or local surface slope effects.All of the models have about the same statistical explanatory power, so evaluations were made by comparing slopes and intercepts of best fit (least-squares) lines and RMSE. From this basis, we conclude that predictions made with the Bagnold (1937) model best match our observations, with the models of Kadib, 1965 and Hsu, 1971 performing nearly as well. The Lettau and Lettau, 1978 and Kawamura, 1951 model predictions match observations least.

► Eight transport rate models are tested.
► No moisture content or slope corrections applied.
► Shear velocity estimated with κ = 0.4 and a variable von Kármán parameter, κa.
► Bagnold’s (1937) model best replicated observations.
► Predictions were improved using the κa parameter.