Watching the polar vortex: what the models show for early November

Before looking at the models, it helps to look upwards: where exactly is the polar vortex, and how far above the layer where our weather happens?

Here we are looking at the stratospheric polar vortex: a large-scale winter circulation in which strong westerly winds flow around cold air above the Arctic. The stratosphere sits above the troposphere, the lower atmospheric layer where our everyday weather takes place. Source: NASA Ozone Watch.

The charts refer to 10 hectopascals, or 10 hPa. This is a pressure surface roughly 30 to 32 kilometres above the ground; its exact altitude changes with the state of the atmosphere. The key point is that these charts describe conditions high in the atmosphere, not the amount of snow we should expect here. Source: Japan Meteorological Agency (JMA).

An important role is played by planetary waves. These are very large-scale wave motions that, under suitable conditions, can propagate from the lower atmosphere into the stratosphere. There they transfer momentum and can slow the westerly winds. The resulting circulation changes can also warm the polar stratosphere. JMA: how stratospheric warmings develop.

As a result, the vortex can shift, become stretched out or divide into separate centres. This division is called a Split. The NASA charts below show a historical example. They help explain the process, but do not depict the current development. NASA: historical vortex evolution.

How to read the historical sequence: On 8 January 2009, the vortex is still fairly compact. By 22 January it has become elongated and shifted. The chart for 29 January shows two separate areas. The later images show further weakening. The colours depict the structure of the vortex — not snowfall amounts or temperatures at ground level. The six charts are spaced one week apart. Source: NASA Ozone Watch.

The access date and model run time are not the same thing. This matters particularly for the wind ensemble: when accessed on 26 October, the available run was still the one from 25 October.

Product and model run	Forecast time	Findings and limitations
ECMWF/IFS ensemble mean 26.10.2026, 00 UTC	03.11., 00 UTC +192 hours Temperature and geopotential at 10 hPa	Strongly distorted, two-centred pattern. A split scenario is possible, but this is not an objective split diagnosis or a calculated probability.
ECMWF Subseasonal 25.10.2026, 00 UTC	Early November Wind at 60° N and 10 hPa	Marked weakening. The ensemble mean remains positive over the period considered, while individual members turn negative.
NOAA/NCEP GFS 26.10.2026, 00 UTC	03.11., 00 UTC +192 hours Geopotential at 10 hPa	Disturbed, two-centred pattern. A single forecast, not confirmation that an event has occurred.
DWD medium-range outlook 26.10.2026, 10:30 UTC	Until 2 November 2026 Weather in Germany	High pressure predominates. Uncertainty increases from the weekend onwards. No separate assessment of ICON stratospheric output was made.

Direct links to the products: ECMWF upper-air chart · ECMWF wind ensemble · GFS upper-air chart · DWD medium-range outlook. These live product URLs may show newer output later.

An ensemble is a set of different model calculations. In the ECMWF wind chart, thin blue lines represent individual members and the thick blue line their mean. The red lines describe the model climate. Individual curves below zero therefore do not yet establish a reliable wind reversal. ECMWF: chart and explanation.

One more point of context: both ECMWF products belong to the same model family, IFS. They cover different forecast ranges, but they are not two independent models. That matters when comparing their results.

Past events help put today's charts into perspective. But they are not templates that the next winter will simply repeat. Comparing 2018 and 2019 makes that particularly clear.

February 2018: a cold spell followed the disturbance

On 12 February 2018 the wind used for the diagnosis reversed and the polar vortex split. This is documented by NASA, including a historical animation. In late February and early March, Britain experienced a pronounced spell of cold weather and snow. The Met Office report on that event describes, among other features, high pressure over Scandinavia and cold easterly winds. In this case, the lower-level weather pattern was favourable too.

January 2019: no comparable effect in Britain

There was also a major stratospheric warming in early January 2019. The NOAA records date it to 1 or 2 January, depending on the reanalysis. Yet there was no comparable effect on British weather. This is explicitly noted by the Met Office in its retrospective assessment explicitly. The example concerns Britain, not Europe as a whole.

For me, this comparison makes the key point tangible: the disturbance high up is one part of the picture. What happens at the surface also depends on the wider weather pattern. Two examples cannot tell us the probability of an event, or provide a forecast for November 2026.

For further reading, the NOAA SSW archive compares several reanalyses. The table used here covers 1958 to 2023 and was last updated in September 2023. Reanalyses combine past observations with weather models. The list records major stratospheric warmings, not necessarily split events or snowy winters.

Has the polar vortex already split?
The charts checked here do not establish that. They are forecasts for early November, not observations of a split that has already happened.

Does a split automatically bring snow?
No. Whether it snows where you are depends on the actual weather pattern, temperatures and precipitation. A stratospheric disturbance alone does not decide that. Met Office.

Can a split be followed by a very cold spell?
Yes, that is possible, but it is not automatic. Such splits have been known for decades and occur from time to time. What matters is whether the circulation in the lower atmosphere also changes in a way that allows cold air to reach central Europe. That can bring a period of very cold weather. A split alone does not tell us whether Germany will be affected, or how intense or prolonged a cold spell might be. This concerns a possible spell of winter weather, not a lasting change in climate. Met Office: possible effects on the weather · Charlton and Polvani: historical split events.

Why do the forecasts differ?
Models and ensemble members produce different possible outcomes. That is why striking individual charts are not enough: agreement between forecasts, their spread and the way later runs develop all matter too.