Impact of snow cover on soil temperature and its simulation in a boreal aspen forest

Snow cover has low thermal conductivity and high albedo, which have large impacts on the soil thermal regime and microclimate. Accurate simulations of snow and soil processes are critical to ecosystem and climate modelling. In this study, we developed a detailed snow and soil scheme in an ecosystem model EALCO to (1) improve the simulations of snow and soil thermal regimes and (2) investigate the effects of different model configurations and parameterizations including snow and soil layering resolutions, fractional snow cover parameterization, water table and total soil column depths, and the inclusion/exclusion of a surface organic layer and unfrozen water in frozen soil. The model includes dynamic layering of snow, user-adjustable layering of soil, snow cover compaction and destructive metamorphisms, implicit solutions of the soil heat and water transfer equations using a finite difference method, distinct parameterizations of the surface organic layer and unfrozen water in frozen soil. The model was tested using nine years (1994-2002) of winter field measurements obtained in a boreal aspen forest. The tested variables included snow depth, snow temperature, soil heat flux, and soil temperatures at various depths. The results showed that the model well reproduced the snow and soil thermal regimes and their interactions observed at the site. The results also indicated that the simulation of snow cover duration was affected by the snow layering resolution and fractional snow cover parameterization. The simulation of soil temperature was affected by the soil layering resolution, the depths of water table and soil column, and the inclusion of a soil organic layer. Inclusion of unfrozen water in frozen soil was found to have only a small effect on soil temperature simulation at this site. (C) 2007 Elsevier B.V. All rights reserved.