Analogue-based identification of preconditioned polar vortex states preceding sudden stratospheric warmings

Sudden stratospheric warmings (SSWs) are known to be associated with the presence of a preconditioned polar vortex state that facilitates the upward propagation of planetary waves from the troposphere. Previous studies have suggested that this pre-warming state differs depending on the SSW type, with displacement events characterized by a weakened, funnel-shaped vortex, and split events associated with a narrower and more vertically aligned vortex displaced towards the pole. More recent evidence, however, indicates that the majority of SSWs, largely independent of their type, are preceded by enhanced tropical stratopause wave driving, which shifts the zero-wind line poleward, displaces the vortex towards higher latitudes and promotes the focusing of wave activity into the polar stratosphere and mesosphere.

Despite this progress, there remains an open debate as to whether SSWs require an anomalously strong pulse of tropospheric wave activity, or whether climatological tropospheric forcing is sufficient when combined with a favorable stratospheric state. In this study, we address this question by applying a Euclidean-distance-based analogue method to identify pre-warming polar vortex states using daily ERA5 reanalysis data. Analogous vortex configurations are objectively defined based on their similarity to a reference pattern and are subsequently classified according to whether they precede an SSW or not.

Our results show that approximately 75% of SSWs occurring between December and February during the period 1980–2021 (18 out of 23 events) are preceded by a recurrent preconditioned state characterized by a poleward-displaced vortex north of 60°N. This preconditioning phase persists over a variable number of consecutive days and terminates with a strong stratopause-level vortex deceleration, accompanied by the development of easterly winds that subsequently propagate downward through the stratosphere, marking the onset of vortex decline. A key distinction between cases that do and do not lead to an SSW lies in the strength of lower-tropospheric wave activity. Thus, while wave forcing is enhanced relative to climatology in both cases, it is stronger in SSW events, supporting the idea that both a favorable preconditioned vortex state and anomalously strong tropospheric wave forcing are necessary ingredients for SSW generation. Finally, split SSWs tend to be associated with stronger and more persistent tropospheric wave activity, both during the establishment of the preconditioning state and throughout the subsequent vortex breakdown.