Improvement and validation of a snow saltation model using wind tunnel measurements

A Lagrangian snow saltation model has been extended for application to a wide variety of snow surfaces. Important factors of the saltation process, namely number of entrained particles, ejection angle and speed, have been parameterized from data in the literature. The model can now be run using simple descriptors of weather and snow conditions, such as wind, ambient pressure and temperature, snow particle sizes and surface density. Sensitivity of the total mass flux to the new parameterizations is small. However, the model refinements also allow concentration and mass flux profiles to be calculated, for comparison with measurements. Sensitivity of the profiles to the new parameterizations is considerable. Model results have then been compared with a complete set of drifting snow data from our cold wind tunnel. Simulation mass flux results agree with wind tunnel data to within the bounds of measurement uncertainty. Simulated particle sizes at 50 mm above the surface are generally larger than seen in the tunnel, probably as the model only describes particles in saltation, while additional smaller particles may be present in the wind tunnel at this height because of suspension. However, the smaller particles carry little mass, and so the impact on the mass flux is low. The use of simple input data, and parameterization of the saltation process, allows the model to be used predictively. This could include applications from avalanche warning to glacier mass balance. Copyright (C) 2008 John Wiley & Sons, Ltd.