Simulation of snow cover characteristics with atmospheric general circulation models

Abstract

A short review of snow cover parametrizations used in different atmospheric general circulation models is given. The results of comparison of 20-year mean integral characteristics of snow cover for North America and Eurasia obtained in three Russian models in the AMIP-2 experiments with observations and reanalysis data are analyzed. Results of the models of the Hydrometeorological Centre of Russia, the Institute of Numerical Mathematics of the Russian Academy of Sciences, and the Voeikov Main Geophysical Observatory are used. It is shown that all models better reproduce snow cover characteristics of Eurasia than those of North America, and snow area is reproduced with smaller errors than snow mass. It is also shown that the fall and winter snow cover formation is simulated more closely to the reference data than the spring snowmelt. It is also shown that 20-year mean snow cover changes in watersheds of the Siberian Ob, Yenisei, and Lena rivers are well reproduced in all models. In the fall and winter period of snow cover formation, the model results are close to one another. During a spring rapid snowmelt, the model of the Hydrometeorological Centre is closer to the reference data.