EVALUATION OF THE SENSITIVITY OF PARAMETERIZATIONS OF SIMULATED LATENT HEAT FLUX OVER SNOW AT TWO COMPLEX MOUNTAIN SITES

The snow cover energy balance is typically dominated by net radiation, sensible and latent heat fluxes. Validation of the two latter components is rare and often difficult to undertake at complex mountain sites. Latent heat flux, the focus of this paper, is the primary coupling mechanism between the snow surface and the atmosphere. It accounts for the critical exchange of mass – sublimation or condensation, along with the associated substantial snowcover energy loss or gain. To evaluate how latent heat flux, and associated model parameters, varies across the landscape, measured latent fluxes using eddy covariance were compared at a wind-exposed and a wind-sheltered site. A well-tested and validated snow cover energy balance model, Snobal [Marks et al., 1999, 2002], was selected for this comparison because of previous successful application of the model at these sites, and because of the adjustability of the parameters specific to latent heat transfer within the model. Simulated latent heat flux and snow cover mass were not sensitive to different formulations of the stability profile functions. The adjustable parameters of snow surface roughness length and active layer thickness influenced how the model simulated latent heat flux as well as the development and ablation of the snowcover. At the exposed site, shorter roughness lengths and a thicker active layer were optimal, while at the sheltered site, longer roughness lengths and a thinner active layer were optimal. These findings are linked to the physical characteristics of the study sites and can be incorporated into the spatially distributed version of the model Isnobal by allowing critical parameters to vary over the modeling domain.