Sand Movement by Wind?Optimising Predictive Power

Chapman, David M.; ,

Sand Movement by Wind?Optimising Predictive Power

by David M. Chapman, Univ of Sydney, Sydney, Australia,

Document Type: Proceeding Paper

Part of: Coastal Sediments

Abstract:

The aim of the work reported here was to develop a model for the efficient utilisation of standard anemometer data for the prediction of potential sand movement. Accurate aeolian sand transport rate data are difficult to obtain, and many different formulae have been developed to fit experimental data and also to incorporate factors such as effects of wind turbulence, ratio of shear velocity to threshold velocity, vertical wind velocity gradients, and so on. Consequently, choice of a method for the prediction of aeolian sand transport is not straightforward. Consideration of the available formulae for prediction of aeolian sand transport led to the conclusion that, as different methods had been developed from a variety of experimental and theoretical approaches, there was scope for development of a model which utilised optimisation techniques to fine tune the predictive capacity of the best performing methods, employed in concert. The work involved assembly (from the literature) of an empirical database of shear velocity/transport rate determinations, the testing of a number of the best known formulae against the empirical data set, with rejection of those of lower predictive power, and the development of a statistical optimisation technique for the set of formulae (for both transport rate and shear velocity) retained. The resulting model offers predictive efficiency (measured by the statistic r²) of 10% to 20% better than any of the component formulae taken individually. The model was originally developed for application in the context of management problems in coastal dunefields, and has been applied to the characterisation of dunesand movement in coastal environments in eastern Australia.

Return to search