Vol. 4, No. 2, 179-190, 2005

Abstract
Large-scale snow cover extent and the atmospheric circulation have recently been recognized to interact mutually. Understanding the snow-atmosphere interaction is important for a better predictability of the cryospheric climate variability under the future climate change.

We investigated interactions between continental snow cover extent and the large-scale atmospheric circulation variability in the Northern Hemisphere, on a climatic scale (from seasonal to sub-decadal) for the recent three decades. Both observations and numerical simulations were used. Utilized snow cover data were derived from the visible sensor (AVHRR). The indices to present the atmospheric circulation variability, such as the Arctic Oscillation (AO), were computed from the reanalysis data. Statistical analyses showed that the Eurasian snow cover extent SCE_EUR and the AO have a significant coherency in sub-decadal periods, with the former leading the latter by several months. A more detailed inspection revealed a climatological change in the sub-decadal covariability between snow and the atmosphere in the mid 1980s. The change may be attributed to changes in the seasonal persistency of SCE_EUR, from autumn to winter, and in strength of winter-to-spring interaction between SCE_EUR and the AO. A series of numerical experiments was performed with an atmospheric general circulation model, to examine the importance of initial (January) circulation anomaly patterns for the change in the winter-to-spring interaction, and reproduced the observed change successfully.

Re-calculation of large-scale snow data using a physical snow model for the pre-satellite era has been planned to reconstruct a dataset of a longer period with consistent quality for an extension of the covariability research. Our preliminary attempt showed, among other atmospheric forcing inputs, the accuracy of the surface air temperatures is most important for a reliable re-construction.

ISSN 1729-3782