Numerical experiments were carried out using the thermodynamic model with the aim to optimize choice of parameterization of the density of fresh snow, its albedo, and thermal conductivity coefficient in order to reproduce the seasonal evolution of ice thickness in the North-Eastern part of the Sea of Azov. The simulation results were compared with each other as well as with the observations obtained at the costal station of the Southern Scientific Center of the Russian Academy of Sciences in

more »

... Taganrog Bay. It is shown that small differences in the schemes of parameterization of physical and thermal properties of snow and ice cover may result in significant scatter in the simulation results. To assess the quality of the forecasting of the seasonal course of the ice thickness, the standard deviation of the calculated ice thickness from the average value for the period of measuring ice thickness, the standard deviation, the correlation coefficient, and the verification of the forecast were determined. Based on the analysis of these parameters, the optimal configuration of the snow layer parameters is proposed, which allows adequate reproducing of the seasonal thermal dynamics of the sea ice thickness. For the conditions of winter 2010/2011 the most close values of calculated ice thickness to results of the measurements in the North-Eastern part of the Taganrog Bay were obtained by determining the dependence of the density of fresh snow on the temperature in the near-surface layer of the atmosphere by the algorithm CLASS, albedo of the snow surface - by the scheme EHAM5, and the coefficient of thermal conductivity of snow – by the formulas of N.I. Osokin or M. Janson.