A parameterization scheme of vertical energy/water exchange at the landsurface was developed for inclusion into atmospheric models and for local/regional estimates of heat/water budget components. The previous version of the model was tested for different natural conditions during several stages of PILPS (Project for Intercomparison of Land Surface Parameterization Schemes). In the last version of the scheme, main attention was paid to the inclusion of processes typical for cold seasons/regions: snow accumulation and melting, soil freezing and melting, infiltration and runoff of melt water.
The snow block of the scheme includes snow accumulation; evaporation; melting due to energy excess at the surface, presence of warm rain and fog, or warm soil surface; water holding capasity of the snow cover; refreezing of melt water within snow cover. The soil block includes freezing/melting of the soil due to energy balance at corresponding levels within it, soil water flow to the frozen zone from unfrozen one, melt water infiltration into soil and melt runoff formation. All these processes are influenced by and influence the vertical profiles of soil humidity, frozen water content and temperature. The description of the processes in the scheme was made in a simplified manner in order to save computer resources for experiments with atmospheric models.
The scheme was validated during PILPS2d experiments carried out for Valdai site located between Moscow and St.Petersburg. The terrain with moraine landscape abounds in lakes and hills, and the water table is rather high, providing additional water inflow into the soil layer. The 18-year weather data (1966-1983) were provided by PILPS working group. The scheme was run with these forcing data and showed quite good results as validated against the measured characteristics of water balance. The intensity of snow melting and the dates of snow disappearance were evaluated rather correctly. The length of the frozen soil period was calculated satisfactorily for the most years, including those with different thermal and snow cover conditions.
The scheme was also validated against the data obtained at Kolyma water balance station located in mountains of north-eastern Siberia where the permafrost is well-developed (the thickness of frozen ground varies from 50 to 250 meters). The meteorological measurements were carried out in a valley surrounded by mountains up to 700 meters of relative height. The cloudiness observations were used for calculation of radiation fluxes, while all other meteorologic forcing parameters were available with daily temporal resolution. The scheme was run for 3 years (1976-1978) with different weather conditions. The snow/soil surface temperature was modeled with good accuracy, as well as the dates of crossing of freezing point by deep soil temperatures. The seasonal amplitude of deep soil temperature was overestimated possibly due to rather porous gravel underlying the soil. Considering the vertical soil thermal and hydrological inhomogeneity is shown to be desirable for land-atmosphere energy/water exchange parameterization.