Test of two SVAT snow submodels during different winter conditions

Heat exchange at the snow surface is crucial for the heat and water fluxes in the soil during freezing and melting. Modeling the thermal dynamics at the boundaries between frozen soil, snow and atmosphere includes uncertainties due to simplified assumptions in the formulation of boundary conditions or unknown parameter settings. The present investigation aims to identify the major uncertainties for the simulation of thermal dynamics during winter-time. The second goal is to test the sensitivity of certain conditions where the snow cover, at times, is patchy and thin. Measurements of water and heat were conducted on soil plots of 4 m(2) in central Sweden for two winters. A one-dimensional SVAT-model using different assumptions for the thermal boundary conditions, e.g., an energy balance approach for the snow surface, was tested. For the first winter period, the model was shown to be sensitive during thin snow conditions, especially with regard to simulated snow depth and snow surface temperatures. A simplified assumption setting snow surface temperature equal to air temperature resulted in substantial discrepancies. Parameters describing the thermal properties of snow and soil proved to be of secondary importance. Simulations for the second winter corresponded well with measurements and did not confirm the sensitive influence for the upper boundary conditions, probably due to the fact that the snow cover was permanent and not considerably altered by intermediate melting events. (C) 1998 Elsevier Science B.V. All rights reserved.