Andrej / severe-weather.eu

The new seasonal forecast is just 2 days away, so I decided to look at the stratospheric forecast from the October run.

While it missed the SSW event in early November, I decided to take the weaker end of the members that indicated some weakening in this period.

A total of 25 members (practically half, ot of 51 total) dropped below 15m/s in the Oct 25th-Nov 6th window. 21 of those 25 (84%) have their proper Stratospheric Warming reversal from late December to mid-February.

In the other half of the members, 96% reached a reversal in the December-February period. The difference is that the members with the stronger November were focused on an SSW earlier in the season, while the weaker November members showed a reversal more in the second half of January and into the first half of February.

I will prepare updated plots as soon as new data is available.

Looking at what's ahead for the Polar Vortex:
The heat flux forecast suggests the tropospheric wave drive has peaked for now, and we are slowly returning to normal territory by mid-November.

Radiative cooling in the polar stratosphere is strong at this time of year. With no upward wave guide and sharp radiative loss, it naturally accelerates the polar night jet via thermal wind balance, allowing the #polarvortex to re-intensify.

I am excited to see what the seasonal SEAS5 data will show, as it will have the current SSW event in its model spin-up, so it will be interesting to see what it thinks will follow, and if it will still show another disruption in January.

Stratospheric Warming Update (Nov 2nd) Part 2:

It's hard to talk about impacts while we are still not sure of the wind reversal 3 days ahead... But in the end, if the wind ends up bottoming out at 1 or +1, it makes little difference down the line. The #polarvortex disruption itself is remarkable for this time of year in any way.

But I did another research run on the data and looked at all modeled sources, like CMIP6 models and seasonal hindcast model members, to extract a weather signal from early November SSW events in modeled projections.

I am going to write a more in-depth article on the researched impacts of an early November SSW. But this is *NOT a forecast*. This research shows the early November SSW signal when compared to the same years with no SSW. So basically it shows us the difference in the atmosphere and what such an event is trying to force, not what the actual outcome is going to be.

But as a glimpse, we can see an interesting signal for December, matched with a strong El Niño and a west QBO setup, for a colder trend over parts of the U.S. and Europe.

As said, this is the early Nov SSW signal alone, so what it's trying to do "in the background". This is also modeled ocean-atmosphere data, not real observations, because no such events appear in observational records this early in the specific global setup as we have in 26/27.

Stratospheric Warming Update (Nov 2nd).

We are still looking at a proper Polar Vortex flow collapse with a wind reversal and the earliest official Stratospheric Warming on record. You can monitor all this yourself in my Polar Vortex hub (including the 3D shape); link in the comments below. 👇

The models are still not in perfect agreement, which is also a bit unusual, given that the first event window is already between November 4th and 6th. There is a second reversal window between the 8th and the 11th, but that falls within the same event window (20 days).

Only the AI models are currently not on board for any reversal (the AI ENS is not far off), which could be by design. But the ECM is lately pulling back the reversal, and usually the GFS follows with some delay. So definitely a race down to the finish line with this event! 🙂

How did the CFSv2 PDF+spread corrected forecasts handle this El Niño? Not well.

I looked at the data and plotted the weekly ensemble means (7 days × 4 runs = 28-run mean) since April against OISST relative Niño3.4, which CPC also uses in its plots.

I saw that @BigJoeBastardi often uses these plots to analyze the current El Niño, so I ran an actual verification to see how well this performs, and to show that this system doesn't really work in a Super El Niño.

The double-corrected forecast was always behind with the peak of the event, and the values it projected as peak were breached 1-4 months earlier.
Bottom line, the correction turned a small warm bias into a large cold one.

The likely reason: the correction maps each forecast onto the 1991–2020 observed distribution for that month. The 2026 El Niño is beyond anything in this record, so the correction pulls the forecasts back toward past extremes, and the corrected runs fall short while the real El Niño keeps climbing.