Polar Vortex Forecast: Live Maps, Latest Updates, and the Winter 2026/2027 Outlook

ECMWF IFS is the global model of the European Centre for Medium-Range Weather Forecasts, here at 0.25° resolution. It is generally regarded as the most skilful model in the stratosphere, especially for the timing and strength of Polar Vortex disruptions. It runs twice a day (00z and 12z), and its maps go 15 days into the future in 12-hour steps.

AIFS is ECMWF’s Artificial Intelligence Forecasting System, a machine-learning model trained on decades of ECMWF reanalysis data, here at 0.25° resolution. It runs four times a day (00z, 06z, 12z and 18z), and its maps go 15 days ahead in 12-hour steps. Use the model run buttons to step back through the last four runs. AI models are very skilful in the medium range, but their stratospheric fields can look smoother than ECMWF IFS at longer lead times, so AIFS is best read side by side with the IFS: when both show the same evolution of the vortex, the signal is more reliable.

GFS (Global Forecast System) is the United States global model, run by NOAA/NCEP, here at 0.25° resolution. It runs four times a day (00z, 06z, 12z and 18z), and its maps go 16 days ahead in 12-hour steps. Use the model run buttons to step back through the last four runs and see how the forecast has changed. At longer range the GFS can swing strongly from run to run, sometimes showing vortex disruptions that later disappear, but it is a very worthy addition, especially to compare how it agrees on a forecast with the ECMWF.

ECMWF and GFS side by side, both showing the same date and time. ECMWF runs only at 00z and 12z, so it is paired with the GFS run from the same time of day. When the two models show the same evolution of the vortex, such as a similar warming or displacement, the forecast is more reliable. When they disagree, the outcome is still uncertain, and the next runs will show which way the trend goes.

10 hPa is around 30 km (19 miles) high, and represents the core of the Polar Vortex in the mid-stratosphere. Sudden Stratospheric Warmings are officially defined at this level, by tracking any wind reversals.

50 hPa is around 20 km (12 miles) high, in the lower stratosphere, about twice the height of the jet stream. Changes that reach this level are closer to our weather, and are more likely to affect the jet stream and the weather patterns at the surface in the following weeks.

10 hPa (around 30 km or 19 miles) and 50 hPa (around 20 km or 12 miles) side by side. Vortex disruptions usually start higher up and work their way down over one to several weeks. When a warming or displacement at 10 hPa also shows up at 50 hPa, it is getting closer to the weather layer, where it can influence the jet stream.

The Polar Vortex as it really looks. The black contour lines show the height of the pressure level in decameters, and the tight ring of lines around the low center marks the Polar Vortex itself, with the strongest winds where the lines are closest together. Colors show the actual temperature, with the coldest air (purple, blue) usually trapped inside the vortex core.

A round Polar Vortex sitting over the North Pole is strong and stable. When it gets stretched, pushed away from the pole (a displacement) or split into two parts (a split), it is undergoing a disruption event, with or without a stratospheric warming wave.

How much warmer (orange, red) or colder (blue) than normal the stratosphere is for this date, compared to the 1990–2020 average. In winter, the air inside the Polar Vortex is extremely cold, often below −70 °C (−94 °F) at 10 hPa.

A rapid warming of tens of degrees over the Arctic within a few days is the signature of a Sudden Stratospheric Warming (SSW), usually caused by large atmospheric waves breaking into the stratosphere from below.

Pressure heights compared to normal (1990–2020). Blue means lower heights than normal, where the Polar Vortex is deeper and stronger. Orange and red mean higher heights than normal, usually a high-pressure ridge pushing into the Arctic, often from the North Pacific side (the Aleutian High).

A large red area over the pole means the Polar Vortex is weakened or displaced, with its cold blue core pushed toward one of the continents, which often brings colder weather below it.

Press play or drag the slider to run the forecast in 12-hour steps, up to 15 days ahead with ECMWF and AIFS, and 16 days with GFS. You can also use the arrow buttons, or swipe the map left or right on a phone.

The strength of the Polar Vortex, measured as the average west-to-east (zonal) wind around the 60°N latitude circle at 10 hPa. This is the standard index used to track the strength of the Polar Vortex. Black is the observed analysis, red is the ECMWF, blue is AIFS (ECMWF’s AI forecast model), and purple is the GFS forecast, green the GEFS ensemble mean.

The gray line is the long-term average, and the dashed lines the record highs and lows since 1958. Above the gray line means a stronger than normal Polar Vortex, below it a weaker one. If the wind drops below zero (solid black line), it reverses to easterly. That is the official definition of a major Sudden Stratospheric Warming event.

The average temperature of the polar cap (60–90°N) at 10 hPa, around 30 km (19 miles) high, from 15 days ago to 35 days ahead. Black is the observed analysis, red is the ECMWF, blue is AIFS (ECMWF’s AI forecast model), purple is the GFS forecast, and green is the GEFS ensemble mean. The gray line is the long-term average (1980–2020), and the dashed lines are the record warmest (red) and coldest (blue) values since 1958. The left scale is in °C, the right one in Kelvin.

A strong Polar Vortex keeps the polar cap cold, near or below the gray line. A sharp rise of 15 to 30 °C (27 to 54 °F) within a few days, often climbing toward the record warm line, is the signature of a Sudden Stratospheric Warming.

The chance of a Sudden Stratospheric Warming (SSW), from three ensemble forecasts: ECMWF ENS (50 members), AIFS ENS (51 members, ECMWF’s AI ensemble) and the American GEFS (31 members). Each member is a slightly different version of the same forecast, so together they show the range of possible outcomes. The top panel shows the 10 hPa zonal mean wind at 60°N: the thick lines are the ensemble means, the shading covers the middle 80% of the members, and the dotted lines follow the weakest member. Black is the observed wind, and the gray band is the normal range for the date since 1958.

The bottom panel shows the share of members with easterly winds (below zero) on each day. That reversal is the official threshold for a major SSW. For example, 30% means that 3 in 10 members reverse the winds on that day. The legend gives the chance that the winds reverse at any time in the forecast. Above 30% is worth watching, and above 60% an SSW becomes likely. A reversal in spring (March to May) usually marks the final warming, the seasonal end of the Polar Vortex, rather than an SSW.

Wave activity, or the “wave attack” on the Polar Vortex. Large weather systems in the troposphere push waves of energy upward into the stratosphere. This chart measures that energy as the heat flux at 100 hPa, about 16 km (10 miles) high, averaged over 45 to 75°N: the more warm air the waves carry toward the pole, the higher the line. Black is the observed value, the colored lines are the means of the three ensembles (ECMWF ENS, AIFS ENS and GEFS) with the middle 80% of their members shaded, and the gray band is the normal range for the date from ERA5 (1980 to 2020).

Strong, long pulses well above the gray band are the classic warning sign of a Sudden Stratospheric Warming: the waves break in the stratosphere and slow the vortex down, usually 1 to 2 weeks later. A single short spike rarely does much, while several days of strong flux, or repeated pulses, can weaken or even split the vortex. Weak wave activity lets the vortex stay cold and strong.

Past winters that started most like this one. We compare the wind at 10 hPa and 60°N over the last 30 days, plus the next two weeks of the ensemble forecast, with the same dates in every winter since 1958, and show the five closest matches. The black line is this season so far, the dashed black line the ensemble forecast, and stars mark the Sudden Stratospheric Warmings that happened in those winters. The gray band is the normal range.

Analogs show how winters with a similar start played out, not a forecast. The stratosphere remembers its state for a few weeks, but the waves that cause an SSW are not predictable more than two or three weeks ahead, so a similar start does not decide how the winter goes. Since 1958, about half of all winters had at least one major SSW. When we tested the method on every winter since 1959, the number of SSW winters among the analogs did not predict whether an SSW followed, so read them as history, not odds.

The same vortex wind index over the whole season, from August to April. The gray shading shows the usual range from past winters (the middle 40% and 80% of years), and the dashed lines the record highs and lows. The orange line is last winter. The wind normally strengthens through autumn, peaks in mid-winter, and turns easterly in spring, when the vortex breaks down for the summer (the final warming). It shows how this winter compares to historical data.

The last 8 ECMWF forecast runs in one chart, issued over the past three and a half days, with the latest in bold red and older runs in lighter colors. If the newer runs keep moving in the same direction, for example each one showing a stronger weakening, the forecast trend is becoming more reliable. Big jumps between runs mean high uncertainty.

The last 8 GFS forecast runs in one chart, with the latest in bold red. The GFS runs four times a day, so these are the runs issued over the past two days. Runs that agree with each other give more confidence, while runs jumping up and down mean the forecast is still uncertain. Comparing this chart with the ECMWF trend shows whether both models are moving the same way.

The last 8 AIFS forecast runs in one chart, with the latest in bold red. AIFS is ECMWF’s AI forecast model and runs four times a day, so these are the runs issued over the past two days, each looking up to 15 days ahead. Runs that agree with each other give more confidence, while runs jumping up and down mean the forecast is still uncertain. Comparing this chart with the ECMWF trend shows whether the AI model and the physics-based model are moving the same way.

How well did the models forecast the Polar Vortex? Each line is one forecast run of the 10 hPa zonal mean wind at 60°N, the standard index of Polar Vortex strength, drawn from the time the run was issued. The black line is what actually happened, taken from the start (+0 h) of every run and smoothed over 24 hours. The bold red line is the latest run, the colored lines are the 8 runs before it, and all older runs are drawn in light gray. Where the gray lines fan out above or below the black line, the model kept forecasting the vortex too strong or too weak at that lead time. The dashed line is the normal wind for the date (1958–2025), and below zero the winds turn easterly, the definition of a major Sudden Stratospheric Warming.

Single runs are the main deterministic forecasts: ECMWF IFS (00z and 12z, 15 days ahead), the American GFS (four runs a day, 16 days) and AIFS, ECMWF’s AI model (four runs a day, 15 days). The black line is each model’s own analysis, so every model is checked against its own starting point. Watch for systematic errors: a model whose runs keep dropping the wind too fast, or that jumps between very different outcomes from one run to the next, deserves less trust for the next disruption.

Ensembles run the same forecast many times with slightly different starting conditions. Each line here is the ensemble mean of one run: ECMWF EPS (50 members, twice a day), AIFS ENS (51 members, ECMWF’s AI ensemble, four times a day) and GEFS (31 members, four times a day). The mean smooths out the noise of single members, so its lines change more gradually than the single runs. A mean that keeps sliding in the same direction from run to run is a strong sign of a real trend, while a steady bias against the black line shows how far to adjust the next forecast.

Last 30 Days shows every run issued in the past month and the full length of the latest forecast, ideal to compare how each model handled the most recent changes in the vortex.

Full Season runs from September 24 to April 30 and fills in as the winter goes on, with every run of the season in one chart. It shows how each model handled the build-up of the vortex in autumn, every warming or disruption, and the final warming in spring.

Each gray line is one member of the GEFS ensemble, the United States GEFS ensemble model forecast, running about 5 weeks ahead. The black line is the ensemble mean, and the blue shading shows the usual range for each date. The more members that drop toward or below zero, the higher the chance of a Polar Vortex disruption and potentially a Sudden Stratospheric Warming in the coming weeks. Note that the GEFS tends to keep the vortex too strong at longer lead times, so a weakening may turn out stronger than shown.

These charts show the atmosphere over the Arctic from the lower atmosphere up to about 48 km or 30 miles (1 hPa), with time running from left to right. Everything left of the red dashed line is the GFS analysis of the past weeks, everything right of it is the 16-day GFS forecast. The pressure scale on the left and the height scale on the right show how high each layer is.

The same charts from the ECMWF model, the European forecast that is generally the most skilful in the stratosphere. Everything left of the red dashed line is the ECMWF analysis of the past weeks, everything right of it is the 15-day ECMWF forecast, updated twice a day. The free ECMWF data only goes up to 10 hPa, about 32 km or 20 miles, so these charts stop lower than the GFS ones, and the layers between 50 and 10 hPa are filled in smoothly between the two levels. Warmings that start near the top of the stratosphere show up here a few days later, once they reach 10 hPa. Switch between ECMWF and GFS to see whether both models agree on how far a disruption works its way down.

The average west-to-east wind at 60°N at every height, in m/s. Dark red means strong westerly winds, the jet that forms the edge of the Polar Vortex. The winds are strongest near the top of the stratosphere and weaken toward the ground. Blue would mean the winds have reversed to easterly, which happens during a Sudden Stratospheric Warming, usually first at the top and then working down.

The wind compared to normal (1990–2020), in m/s. Blue means weaker than normal, red stronger. Blue areas spreading down from the top show a weakening Polar Vortex working its way down toward the lower atmosphere, which can take one to several weeks. A strong, long-lasting red signal means a stable Polar Vortex that keeps the cold air locked in the Arctic.

Polar cap (60–90°N) average temperature compared to normal. Strong red areas appearing at the top of the chart mean the stratosphere is warming quickly, as in a Sudden Stratospheric Warming, when parts of the polar stratosphere can warm by 30 to 50 °C (54 to 90 °F) within a few days. Blue means a colder than normal and usually stronger vortex.

The same temperature anomaly, scaled by how unusual it is for this time of year, in standard deviations (σ). This makes different heights and seasons easier to compare. Values above +2 or below −2 occur only a few percent of the time, so they mark truly unusual events.

Polar cap (60–90°N) heights compared to normal, in meters, from the surface up into the stratosphere. Orange and red mean higher heights and a weaker vortex, blue lower heights and a stronger vortex. When the red moves down from the top toward the surface, it often brings a negative Arctic Oscillation (AO), a weaker and wavier jet stream, and a higher risk of cold outbreaks into the mid-latitudes in the following weeks.

The famous “dripping paint” chart. The same height anomaly, scaled by how unusual it is (standard deviations). This view became widely known from research on how stratospheric anomalies descend into the lower atmosphere (Baldwin and Dunkerton, 2001). Watch whether anomalies above +1 or below −1 reach the surface: red favors a negative AO and colder spells, while blue favors a positive AO, with milder and stormier westerly flow.

Click the chart to open it full size in a new tab.

All charts load in full resolution, so you can pinch to zoom in on the details. Press and hold a chart to save it or open it on its own.