Blowing snow redistribution

Considering the importance of snow hydrology for water supply in the Cana-dian Prairies, and the impact of blowing snow processes on the amount of snow available for spring melt, there is a need to develop models of blowing snow trans-port and sublimation in alpine terrain. A distributed model of blowing snow model (DBSM) was developed by Essery [2] based on MS3DJH/3R terrain wind flow model and the Prairie Blowing Snow Model (PBSM,[9]). This model applied to rolling topography (Trail Valley in the arctic tundra) showed good results. In our study, an attempt was made to run DBSM over steep mountain topography in Marmot Creek Basin, Canadian Rocky Mountains and to assess its performance. Breaking the assumption of rolling topography, the accuracy of the linear wind flow model is challenged. Using LiDar data for elevation and snow depth and observations from different weather station in the basin, the behaviour of the wind flow model for calculating blowing snow was tested. Although the model must show serious deficiencies in modelling wind on steep slopes, some observed features of redistributed snow covers might be reproduced. As an alternative, a commercial wind flow model (Windsim) using a 3D Reynolds averaged Navier-Stokes equations and the k − ǫ turbulence model was coupled to DBSM. This model showed promising results in the complex moun-tain topography. Nevertheless, the limited number of cells is too coarse to predict complex wind regimes, which are affected by fine scale topographic configurations. Both model do not reproduce the features of wind blowing over ridges. So, eventually, this particular case of blowing snow process is studied through a two dimensional approach with Ansys CFD program. Au vu de l’importance de l’hydrologie de la neige pour les ressources en eau des Prairies canadiennes et de l’impact du vent sur la redistribution de ce bien précieux, il existe un besoin fort de développement de modèles des processus éoliens de transport et de sublimation de la neige. Un modèle distribué simulant ce phénomène a été développé par Essery [2]. Ce modèle basé sur le “Prairi