资源环境科技发展态势分析平台

The European Centre/Hamburg version 6 atmospheric model with a well-defined stratosphere and an internally generated Quasi-biennial oscillation (QBO) was used to study the relationship between the autumn Eurasian snow extent and the wintertime climate of the northern hemisphere. A positive snow cover anomaly was imposed over Eurasia in early autumn and held constant until spring. One hundred years of the snow anomaly run was compared with a 100 year control run. A dynamical response to the snow anomaly is seen in the northern polar stratosphere and troposphere during autumn and early winter, in line with previous modeling studies, and the monthly progression of the atmospheric anomalies follows the size of the surface forcing. However, this response is weaker, and occurs earlier in season, than that seen in observations. Considering the effect of QBO, we find a stratospheric vortex weakening during the easterly phase; however, the effect is weaker than that seen in observations. The strongest response of the polar vortex is found when both factors-the snow anomaly and the QBO phase-are considered together, with the response being close to an additive combination of the responses to the individual forcings. Although our study confirms the ability of snow cover variability to significantly influence the large-scale circulation, it also demonstrates that the inclusion of a well-resolved stratosphere is not a sufficient condition for reproducing the intensity and timing of the circulation response appearing in observations.

Plain Language Summary The ability to predict European wintertime temperatures weeks or months in advance would have many positive implications for society. However, standard weather prediction techniques are unable to make accurate forecasts much more than a week in advance. For long-term seasonal forecasts we rely on the slowly changing features of the climate system to give us information about the future weather. These slowly changing features include the autumn snow cover over Siberia and winds high in the tropical atmosphere. Both of these can influence the stratospheric winds over the northern hemisphere pole in winter, which then affects the weather at the surface. In this study we simulate these slowly changing features and compare the simulated response of the polar winds to the observed response. We find that the simulated response is weaker than the observed response; however, we also find that the Siberian snow and the high tropical winds work together to influence the polar winds more than each factor individually. This combined effect may provide us with additional information when making long-term seasonal forecasts.