Despite the recent changes in Arctic climate (Fig. 1), the influence of Arctic sea ice on mid-latitude weather remains uncertain. The evidence remains unclear in both observations and models (see Barnes et al. 2015 or Screen et al. 2018). This is likely a result of large internal variability, competing effects of upper-troposphere tropical warming vs. Arctic surface warming, nonlinear dynamics, and/or data uncertainties. In association with the next IPCC report (AR6) and Coupled Model Intercomparison Project Phase 6 (CMIP6), a coordinated series of modeling experiments called the Polar Amplification Model Intercomparison Project (PAMIP) will try to answer some of these questions (see Smith et al. 2019).
In our new modeling study, we evaluate how the atmospheric response to Arctic sea-ice loss may be sensitive to a climate phenomenon in the tropics – the Quasi-biennial Oscillation (QBO). The QBO is a variation of wind (easterly or westerly) in the tropical middle atmosphere (stratosphere) approximately every 28 months. This remarkable climate cycle is a result of momentum fluxes transferred by a spectrum of tropical waves that interact with the background mean flow (see reviews by Baldwin et al. 2001 and Anstey et al. 2014). These complex dynamics are still quite difficult to simulate in a global climate model. In fact, most climate models do not produce a realistic QBO cycle at all.
Earlier studies have shown that the phase of the QBO (easterly or westerly) can influence the propagation of planetary waves in the mid-latitudes and the strength of the Northern Hemisphere polar vortex. You can think of planetary waves as being associated with large-scale weather patterns, such as storm systems and the jet stream. The QBO is often used in seasonal weather prediction in the boreal winter. It is therefore reasonable to consider that the QBO may also influence how the atmosphere responds to future Arctic sea-ice decline.
To link the QBO and Arctic sea ice, we extend a series of climate model experiments from an earlier paper (Labe et al. 2018 – see summary) and separate westerly QBO winters from easterly QBO winters. We prescribe the model experiments with a realistic QBO cycle using radiosonde observations. We also set the model to levels of either historical (1976-2005) or future Arctic sea ice (thickness and concentration). The future levels of sea ice in the model use projections from an extreme emissions scenario (Representative Concentration Pathway 8.5) averaged over the 2051-2080 period. Since these are perturbation experiments, we use this large forcing in the model to better understand potential mechanisms.
Loss of Arctic sea ice contributes to a negative Northern Annular Mode (NAM/NAO)-like response during both westerly and easterly QBO winters. Hence, there are equatorward shifts of the polar jet stream. However, we see a significance difference between QBO phases in the stratosphere, especially in early winter. During easterly QBO years, we find that the stratospheric polar vortex weakens in response to sea-ice decline (Fig. 2-right). In contrast, we find that the polar vortex strengthens to Arctic sea-ice decline during westerly QBO years (Fig. 2-left). This result was also robust in an additional experiment using less sea-ice decline. To understand the cause of this difference in the strength of the polar vortex, we conduct analysis on how atmospheric waves vertically propagate into the stratosphere during each QBO phase (see background reading). In response to Arctic sea-ice loss during the easterly QBO, these atmospheric waves reach into the stratosphere and disturb (weaken) the polar vortex. This is not the case during westerly QBO years, and they are instead trapped within the troposphere. The strength of the Ural/Siberian High plays a key role in the linkages between the QBO and Arctic sea ice. During the easterly QBO years, we see a significant reinforcement of the Siberian High and subsequent increase in the frequency and intensity of cold temperature extremes in central Asia due to Arctic sea-ice loss. Again, this is not the case for westerly QBO years.
In summary, we show that the QBO can modulate the atmospheric response to Arctic sea-ice loss in our climate model. While the experiments are idealized, our results further suggest that the atmosphere is quite sensitive to changes in Arctic sea ice and the background climate state. This may affect the interpretation of Arctic-midlatitude linkages in climate models that do not resolve stratospheric processes, such as the QBO. All in all, how the polar vortex responds to future declines in Arctic sea ice is still very unclear in the scientific literature. We have planned additional targeted experiments that will look at these connections in longer simulations and in different climate models. Stay tuned!
Labe, Z.M., Y. Peings, and G. Magnusdottir (2019). The effect of QBO phase on the atmospheric response to projected Arctic sea ice loss in early winter, Geophysical Research Letters, DOI:10.1029/2019GL083095
[Plain Language Summary]
 Labe, Z.M., G. Magnusdottir, and Y. Peings. Linking the Quasi-Biennial Oscillation and Projected Arctic Sea-Ice Loss to Stratospheric Variability in Early Winter, 20th Conference on Middle Atmosphere, Phoenix, AZ (Jan 2019).
 Magnusdottir, G., Z.M. Labe, and Y. Peings. The role of the stratosphere, including the QBO, in Arctic to mid-latitude teleconnections associated with sea-ice forcing, 2018 American Geophysical Union Annual Meeting, Washington, DC (Dec 2018).