Good afternoon, everyone, and happy first day of meteorological winter!
The other day I wrote how the outlook appeared to be favoring above-normal temperatures for the month of December into early January, as the stratospheric polar vortex (SPV) is currently stronger than normal and model forecasts for that period had been aggressive with keeping the vortex strong. Since then, however, we have seen a shift in model guidance that bring the SPV under attack to a stronger degree, and a broader shift in the global weather regime may favor a round of sharply colder Canadian air in the December 20-30 timeframe.
Previous forecast model runs anticipated the development of a single, weak ridge of high pressure way up in the stratosphere that would appear to be a nuisance towards, but not disrupt, the stratospheric polar vortex. Consequentially, all stratospheric signals seemed to be aiming in a warmer direction for U.S. weather for the balance of December and into January.
Latest model guidance has changed its tune. The GFS ensembles are shown above, with the 10-millibar forecast of geopotential height anomalies shown in color and pressure contours juxtaposed on top. We see that the GFS ensembles are showing a vortex splitting event to occur around December 7-9, with two distinct ridges of high pressure (one in Eurasia and another in the Pacific) pinching the SPV across the North Pole so that one lobe of the vortex breaks off and settles over Greenland, while the second lobe drifts into Siberia.
This isn't the most impressive vortex splitting event you'll ever see, but it's enough to disrupt the flow of the stratospheric vortex and, in turn, open the door for a round of Arctic air about two to three weeks later. This puts us roughly in the December 20th - January 5th timeframe, and this will be the period to watch for a potential round of much colder than normal air in North America.
So we know that the stratospheric polar vortex will come under pressure over the next week or two, perhaps splitting in the process. Supposing a cold air event may become more likely two to three weeks later, how will the overall pattern look around then? We start with an analysis of tropical convection.
The chart above (from weathermodels.com) is complex at first blush, but can be easily broken down. This shows forecasted thunderstorm activity along the Equator (in a band between the 15 North and 15 South lines of latitude) from the long-range ECMWF model across all lines of longitude, from today through January 15th (y-axis scale). The variable being measured here is upper-level velocity potential anomalies, which is a fancy parameter for judging how conducive the atmosphere will be for thunderstorm activity.
Here, yellows and oranges represent above-normal velocity potential anomalies, which suppress thunderstorm activity. On the flip side, greens and blues indicate lower VP anomalies, which then enhance thunderstorm activity.
We can then ascertain what phase the Madden-Julian Oscillation (MJO) is in by seeing when those greens and blues pop up across different parts of the world. For example, over the next few days, we see blues and greens across the islands northwest of Australia and southeast of India, which correspond to MJO phases 3, 4 and 5. This feature has recently helped to enhance above-normal temperatures in the United States, as Phases 3-5 encourages warmer weather across the continent. How do we know this? The composite image on the right (from the Climate Prediction Center) shows temperature anomalies during the Nov-Dec-Jan period for each of the eight MJO phases.
By the same token, when enhanced Equatorial thunderstorm activity develops northeast of Australia in the open Equatorial Pacific, the MJO tends to be in Phases 7-8, which correlate colder for most of the country. The whole process starts over again with enhanced thunderstorms in the Equatorial Atlantic and over Africa, which is MJO Phase 1 territory.
Returning to the ECMWF forecast image above, after our MJO Phase 4 influence dies down by December 10th, we are expecting to see a robust resurgence in tropical convection in the Equatorial Atlantic, which I've highlighted in the graphic. While not explicitly shown, thunderstorm activity around those waters and into Africa correlates to a more-or-less MJO Phase 8 or 1 type of pattern, and will tend to lean cooler for most of the United States as a result. This enhanced thunderstorm activity in the Atlantic kicks off by December 10th and may stick around through early January, with ebbs and flows in the intensity of this thunderstorm activity.
Rather than rely on subjective interpretations of the MJO to decide where the oscillation is heading, we can also make use of an objective MJO phase-space analysis, shown above for the same long-range ECMWF model we were just subjectively analyzing. In this objective analysis above (from the CPC), the MJO is seen meandering (and strengthening) in Phase 7 from December 17th through the end of the forecast period, December 30th. This is a little earlier in the phase progression from the Phases 8 and 1 I had been expecting, but no matter - the general theme of a global pattern shift in favor of colder weather for the last ten days of December and into early January remains intact.
The stratosphere will be coming under some duress in the next several days, which may raise the risk of a cold air event in the U.S. for the end of December and into early January, while the MJO is shifting into pro-cold phases for the last ten days of December and perhaps onwards. Now that we see what kind of environment may be setting up, let's see if model guidance is latching on to any cold weather threats.
Beginning with the American CFS model, we take a look at 10-day temperature anomalies across North America from December 16th through the 26th. The pattern looks generally warm for the U.S., but the real thing I'm watching is that reservoir of colder-than-normal air building in western Canada, and similarly anomalous above-normal temperatures in the Arctic Circle. This suggests a shift in Arctic air to the south, displaced into western Canada and further east into Greenland. While not a necessity for cold air events, a reservoir of cold Canadian air in the western part of the country is traditionally a good sign.
Moving ahead to the December 21st - 31st timeframe, we see what that cold Canadian air reservoir does: it spills south and east. With continued warm air displacement into the Arctic Circle, below-normal temperatures have free rein to grind south and east into the United States and deeper into Canada. Above-normal temperatures are still anticipated to linger in the South and along the Eastern Seaboard, but this may be something that happens earlier in this period followed by colder weather later towards the end of the year.
Our last analysis of the CFS model comes for the December 31st - January 10th timeframe. It appears as though the CFS takes that MJO Phase 7 temperature composite to heart and continues to displace cold Arctic air southwards into Canada and the northern U.S., with much colder than normal temperatures seen across the continent.All in all, I see the CFS model as pretty conducive for a cold air event in this December 20th - January 10th timeframe, particularly with such a formidable reservoir of Arctic air in western (and eventually all of) Canada. As our last bit, let's turn our attention to the long-range ECMWF model and see what it says.
Unsurprisingly, the long-range ECMWF is less keen to bring about a much colder weather regime to North America, but it does take the intriguing step of developing and intensifying that reservoir of cold air in western Canada. With anomalies seen deeper here than what the CFS shows, it does open the door for the risk of a more intense cold air event if we took this forecast verbatim. Of course, that would be too easy, and the ECMWF instead brings cold weather to the far north-central U.S. but maintains warmth in the South and East for the balance of the year and into early January.
To Summarize:
- The stratospheric outlook has changed since earlier this week, and it now appears more likely that the stratospheric polar vortex will be disrupted to some degree in the December 5-10 period
- Combined with changes in the global weather pattern (via the MJO) to a state more conducive for below-normal temperatures in the U.S., the chances of a colder last several days of December & colder opening days of January are rising
- Model guidance contains differences, but with both main models showing a formidable reservoir of frigid air accumulating in western Canada by December 25th, I am feeling optimistic that future forecasts for the northern third of the country (and potentially further south) will continue to trend colder
Andrew