Key points are not available for this paper at this time.
Seasonal prediction systems which contribute to the Copernicus Climate Change Service (C3S) multi-model suite are used operationally to forecast the state of the atmosphere a season ahead. During Northern Hemisphere winter, a source of potential seasonal forecast skill is the strength of the Arctic stratospheric polar vortex and its coupling to tropospheric modes of variability. The intraseasonal-to-seasonal variability in the vortex strength can induce shifts in the position and strength of the mid-latitude tropospheric jet, and an accurate representation of stratospheric variability in models has been shown to enhance tropospheric predictability at seasonal time scales. The strength of the Arctic vortex and the coupling between the stratosphere and the troposphere are commonly described by anomalies in the multi-level hemispheric atmospheric mode known as the North Annular Mode (NAM). In this work we investigate the relationship between NH stratospheric and tropospheric predictability during boreal winter from a systematic multi-model assessment of the vertical structure and skill of NAM. Overall, the ensemble-mean hindcasts from C3S models (1993/94 2016/17) show a poor performance in predicting the seasonal NAM across all atmospheric levels. The hindcasts display slightly higher prediction skill in the stratosphere than in the troposphere, consistent with its longer timescales, yet no clear pattern connecting stratospheric to tropospheric skill. A monthly-frequency analysis reveals a pronounced dip in the January tropospheric skill across all models and initialisation dates. Since this appears to be linked to a bias in the vertical troposphere-stratosphere NAM coupling during January, we investigate the origin of the models behaviour by looking into the vertical structures of large-scale mid-latitude planetary waves. The deterioration of the seasonal winter skill by the mid-winter model performance indicates a route towards improving seasonal forecasts of the NH winter troposphere.
Portal et al. (Fri,) studied this question.