Simon’s Stratosphere Watch #1

Welcome to a new blog series! Yes, the title was indeed chosen to fit the SSW acronym. In these posts, I intend to give a brief summary of the state of the Arctic stratospheric polar vortex and the latest forecasts. I have yet to decide how regular these will be.

Tuesday 22 February 2022


The stratospheric polar vortex is currently unusually strong throughout the depth of the stratosphere, though slightly weaker than last week. Forecasts show a steady weakening in the mid-stratosphere consistent with the seasonal cycle, but most forecasts still do not show a signal for the final warming and maintain a stronger-than-normal vortex into April.


Past and current conditions

In general, the polar vortex has been strong for most of the season since November (Fig. 1, left). However, it is since the turn of the year that the vortex has pushed into unusual territory. Since on average the vortex becomes more disturbed from January onwards, the anomalous strength reached the largest amplitude during this time.

The evolution of the vortex strength is consistent with unusually weak stratospheric wave driving (Fig. 1, right). A component of this likely arises from a tropospheric height anomaly pattern which destructively interferes with the mean stationary waves, suppressing the amount of wave activity emanating from the troposphere. Such a pattern of destructive tropospheric wave activity was present during the exceptional winter of 2019/20, for example. However, the suppression of tropospheric wave activity is unlikely to explain the full magnitude of the anomalous vortex strength, since the stratosphere also contributes toward determining its own strength.

Figure 1: Left: 10 hPa 60°N zonal-mean zonal winds for 2021/22 (red & pink, with GEOS forecast in orange) versus MERRA-2 climate. Right: 45-day average of the 100 hPa eddy heat flux due to wavenumbers 1-3. Note how the very high heat flux in Dec-Jan 2020/21 preceded the major SSW, whilst the very low heat flux this year has led to the opposite vortex state. Figures from

There have also been several cases of negative eddy heat flux, which yields a downward-pointing Eliassen-Palm flux vector suggesting wave reflection — such reflection episodes can also increase vortex strength (as was the case during winter 2019/20). Although negative eddy heat flux events can also bring the time-average heat flux down, thus making it more difficult to determine if the low heat flux is due to fundamentally low driving or unusually frequent negative events, this winter appears to have been more dominated by only weakly positive heat flux rather than negative heat flux.

Last week, the vortex reached what appears to be the strongest it will be this season. The mid-stratospheric NAM exceeded 4 sigma and 10 hPa 60°N zonal-mean zonal winds exceeded 60 m/s. According to a comparison of analyses from NASA’s GEOS model and MERRA-2 reanalysis, this set daily records at some levels (e.g. 100 hPa), though the daily maxima are quite noisy and individual daily records are less important than the overall position of the zonal-mean winds within the overall climatology. This period coincided with strong cyclonic storms over the North Atlantic — including Storm Eunice over the UK, which led to the Met Office issuing Red weather warnings. It is quite plausible that the strength of the stratospheric vortex influenced the development of these storms: the strength of the lower-stratospheric jet has been shown to influence the evolution of baroclinic waves, whilst the current NAO+ regime is much more likely during a strong lower-stratospheric vortex. This is not to say that the vortex strength ’caused’ Eunice, but it did likely load the background state ‘dice’ in that direction.


In the very short-term, there is a particularly interesting few days from 24 February – 1 March. An upward-propagating stratospheric wave disturbance develops, but then the situation quickly turns strongly reflective: wave activity diminishes in the mid- and upper stratosphere, lower-stratospheric eddy heat flux turns very negative (Fig. 2), and the 10 hPa vortex pings to extreme strength once more and attains a wave-free ‘doughnut’ shape (Fig. 3, left). This is a case of dynamics driving a strong vortex — rather than an absence of dynamics (and why I will stand firm in my belief that strong Arctic vortices are interesting!).

Figure 2: Left: 00Z GFS forecast initialised 22 Feb 2022 for 100 hPa geopotential height (contours) and eddy heat flux (filled colours) valid 00Z 26 Feb 2022. Right: Same GFS run showing EP-flux vectors at 12Z 27 Feb 2022. Note the large region of downward-pointing vectors. The divergence of this vector is associated with an acceleration of the mid-stratospheric U-wind (red contours). Right-hand figure from

The intensification is very brief, and almost immediately thereafter, wave activity returns. However, there is quite a bit of spread in the forecast after this event (Fig. 3, right): the medium-range GEFS ranges from average to record-strong, which is quite a large amount of spread for that lead-time, and suggests the details of this reflection event are a point of longer-term uncertainty. Note that this does not mean the models are ‘struggling’ with this event: they might well be, but the development of spread/uncertainty does not in itself indicate that — it simply means this is a point of uncertainty leading to divergent forecast trajectories.

Figure 3: Left: ‘doughnut’ vortex with an absence of wave activity on 1 March (forecast initialised 00Z 22 Feb 2022). Right: GEFS and GFS forecast of 10 hPa 60°N zonal-mean zonal winds from 00Z 22 Feb 2022. Note the rapid growth in spread in intensity after the peak on 1 March.

Most forecast systems are subsequently showing a downward trend in the 60°N zonal winds across multiple layers in the stratosphere. However, such a weakening trend should be viewed in the context of its starting position: with the vortex at or near date-record strength, it will take quite a deceleration to take it down to even normal strength. The GFS has been recently trying to produce some very disrupted states by amplifying wavenumber-2 — with persistent Alaskan ridging present, transient forecast amplification of the Scandinavia-Greenland pattern appears to have been a contributor to this. These more aggressive GFS solutions have been inconsistent with each other and the GEFS. Be cautious of over-interpreting the saturated reds on the ECMWF 10 hPa temperature plots: these are not showing anywhere near a ‘sudden stratospheric warming’ and are more closely linked to the anomalous heat flux/wave activity around the vortex edge.

Into the longer-range, there remains substantial support from GEFS, ECMWF and CFS for the vortex to remain stronger-than-normal into April (Fig. 4). Reversals of the 10 hPa zonal winds are still hard to come by within these forecast systems. The climatological-mean date of the final warming is close to 15 April, and any reversals at this stage are likely to be the final warming. Some final warmings are driven by wave activity like major mid-winter SSWs, whilst others are driven primarily by the radiative effect of the sun returning to the Arctic and reversing the meridional temperature gradient. A later-than-normal final warming is more likely to be dynamically-driven, but they can be hybrid affairs (like in 2019). A later-than-normal final warming is also counter to the average of winters without a major SSW, as the final warming in those winters is often delayed by the strong radiative recovery following mid-winter SSWs. However, that probably also includes cases where the final warming ends up being, in effect, a very late SSW, and thus without such an event I see no reason to assume that a strong vortex must dissipate earlier.

Figure 4: Selection of extended-range forecasts of the zonal-mean zonal winds at 10 hPa 60°N. Left: 46-day ECMWF, initialised 21 February 2022 []. Right-top: 35-day GEFS, initialised 21 February 2022. Bottom-right: 44-day CFSv2, initialised between 06Z 21 February and 00Z 22 February 2021. Note that the ECMWF forecast is shown with respect to the model’s own 20-year reforecast climatology; GEFS is shown without bias-correction with respect to ERA5, whilst the CFSv2 forecast has been bias-corrected by subtracting the model’s 1999-2010 reforecast bias.

It’s also worth noting that whilst the middle and upper-stratospheric vortex will begin to seasonally decay, the lower-stratospheric vortex (which is currently very strong and contains some of the coldest air ever recorded in the Arctic lower stratosphere at this time of year) will likely persist for some time yet as the timescale of the lower stratosphere is much longer (especially in late winter-early spring). Since it is this which the troposphere ‘feels’, some of the changes going on at 10 hPa will be of less importance given what is present below. In fact, the GEFS ensemble mean shows very little change for the next 5 weeks (Fig. 5), demonstrating precisely why the lower stratosphere can be a useful subseasonal boundary condition to the troposphere.

Figure 5: 35-day forecast of 100 hPa 60°N zonal-mean zonal winds from the 00Z 21 February 2022 GEFS. Note that no bias correction has been applied, but GEFS biases in this metric are generally small.

It is also possible that there could be ozone feedbacks, especially as sunlight returns to the Arctic and allows for ozone depletion: polar cap ozone levels are currently very low, but this is more representative of the effect of a pole-centred vortex with a strong transport barrier and reduced wave driving, as the ozone minima are currently normal. However, should the cold vortex core be slightly displaced equatorward and exposed to sunlight, ozone levels may quickly drop. Using interactive ozone chemistry (i.e., where the ozone feels the dynamics and vice versa) in forecast models has been shown to influence the strength of the vortex and potentially the surface impacts, but it’s not yet fully understood.

And so in summary, if you made it this far, or if you just scrolled to the end: for now, I think persistence wins.

One thought on “Simon’s Stratosphere Watch #1

  1. Pingback: Simon’s Stratosphere Watch #2 | Simon Lee

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s