A Record Arctic Cyclone Made Sea Ice Retreat When It Should Have Been Freezing — And Heat Alone Can’t Explain It

By October, the Arctic Ocean is normally switching gears.

The long summer melt season is ending. Temperatures are falling. Open water begins freezing. Sea-ice concentration increases and the winter ice pack starts rebuilding.

But in October 2022, something unusual happened in the western Arctic.

A remarkably powerful cyclone swept into the Canada Basin during the autumn freeze-up. Instead of simply continuing its seasonal advance, sea ice in parts of the region retreated, became thinner and fragmented into smaller floes.

Now scientists have taken a detailed look at what happened — and the storm’s effects cannot be explained by atmospheric heating alone.

The peer-reviewed study, published on August 11, 2026 in Geophysical Research Letters, concludes that an unusually large reservoir of heat in the upper ocean played an important role. But even after accounting for atmospheric and oceanic thermodynamic effects, part of the observed ice response remains unresolved.

The missing piece may have been mechanical.

Monster waves generated by the cyclone appear capable of breaking the ice apart, redistributing floes and exposing more open water — potentially disrupting the Arctic freeze-up from below, above and sideways at the same time.

The Arctic was supposed to be freezing

That timing is what makes the event particularly interesting.

Arctic sea ice reaches its annual minimum near the end of summer. As autumn darkness returns and the ocean loses heat, the surface begins freezing again.

October therefore normally brings increasing sea-ice concentration across much of the Arctic.

The October 2022 cyclone temporarily disrupted that seasonal progression.

Researchers Xingkun Xu, Takuji Waseda, Tsubasa Kodaira, Takehiko Nose and Kazutaka Tateyama examined the storm using atmospheric reanalysis, satellite observations, ocean-model data and sea-ice measurements.

They found that the cyclone was associated with local reductions in both sea-ice concentration and sea-ice thickness, particularly around the marginal ice zone — the dynamic boundary where consolidated Arctic ice meets increasingly open ocean.

In other words, while the Arctic was entering its refreezing season, the storm temporarily pushed part of the system in the opposite direction.

A 30-year record Arctic cyclone

This was not an ordinary low-pressure system.

The storm was an extratropical cyclone — a large low-pressure system driven primarily by atmospheric temperature contrasts and upper-level dynamics rather than the warm-core processes that power tropical cyclones.

The researchers identify the storm as the most intense October cyclone in the western Arctic during the previous 30 years, based on its exceptionally low pressure and powerful winds.

Its minimum central pressure reached approximately 955 hPa.

Such an extraordinarily low pressure does not, by itself, mean the storm was a bomb cyclone. That term describes a cyclone undergoing sufficiently rapid pressure deepening — known scientifically as explosive cyclogenesis — rather than simply a storm that reaches a very low minimum pressure.

On October 7, while centered near 75°N and 176°W, the analyzed cyclone had a central pressure of about 957 hPa. Ten-minute sustained winds reached roughly 25 m/s, while significant wave heights climbed to around 8 meters.

The cyclone then moved northeast toward the packed-ice zone.

By October 8 it had weakened to around 964 hPa but was moving into an area with nearly complete sea-ice cover. Winds remained around 20 m/s and significant waves were still approximately 5 meters high.

For a thin, developing autumn ice cover, that combination of wind, waves, warm humid air and ocean heat can be extremely disruptive.

First suspect: heat from the atmosphere

A powerful cyclone can transport warmer and moister air into the Arctic and dramatically alter the exchange of energy between the atmosphere and the surface.

So the obvious first question was:

Did warm air simply melt the ice?

Not nearly enough.

The researchers calculated the anomalous surface energy budget and converted the excess energy into a theoretical “melt-equivalent” reduction in sea-ice thickness.

The maximum atmospheric contribution was approximately:

1 centimeter.

The atmospheric anomaly peaked on October 6 — about one day before the cyclone reached its maximum intensity.

Across October 3–9, the cumulative atmospheric melt-equivalent estimate was around 3 cm.

Clearly, the atmosphere mattered.

But it wasn’t the whole story.

Then scientists looked beneath the ice

The ocean contained considerably more energy.

The researchers examined heat stored within the upper 25 meters of the ocean and found a pronounced positive ocean-heat-content anomaly associated with the storm.

In places, the anomaly reached approximately 14,000 kJ per square meter per day.

Unlike the atmospheric signal, which peaked before maximum cyclone intensity, the oceanic heat anomaly peaked around the cyclone maximum and persisted for several days afterward.

If all of that anomalous upper-ocean heat had been transferred to the ice — an intentionally extreme upper-bound assumption — it would have represented the energy required for as much as:

14 centimeters of melt-equivalent ice reduction.

That is roughly fourteen times the maximum daily melt-equivalent contribution calculated from the atmospheric surface-energy anomaly.

But there is an important caveat.

The 14 centimeters does NOT mean the ocean melted 14 cm of ice

This distinction matters.

The researchers are not claiming that they directly observed 14 centimeters of sea ice being melted by the ocean.

The figure represents an upper-bound thermodynamic potential: the amount of melting that could theoretically occur if all the anomalous heat stored in the upper ocean were transferred efficiently into the ice.

Real ocean-to-ice heat transfer is much more complicated.

Only some of that heat may actually reach the ice. Temperature stratification, turbulent mixing, currents and the physical condition of the ice all affect how efficiently energy is transferred.

The scientists therefore tested several lower heat-transfer efficiencies.

Even when only 10–25% of the anomalous ocean heat was assumed to reach the ice, the oceanic contribution remained comparable to or greater than the anomalous atmospheric contribution during the cyclone.

Something important was happening below the ice.

But heat still didn’t completely explain what satellites saw

This is where the study becomes particularly interesting.

The researchers compared their thermodynamic estimates with sea-ice thickness observations derived from SMOS satellite data.

They did not completely match.

Neither atmospheric heating nor the more realistic partial-transfer ocean-heat scenarios fully reproduced the observed reduction in sea-ice thickness relative to climatology.

That leaves what the authors describe as an unresolved residual.

It does not mean the ice disappeared through some unknown source of energy.

Rather, it means that a calculation focused on thermodynamic processes — heating and melting — does not capture everything that can happen to sea ice during an extreme storm.

Ice can move.

Ice can separate.

Ice can pile up.

And, critically, ice can break.

The missing piece may be giant waves

An eight-meter sea is not merely transporting heat toward Arctic ice. These extreme ocean waves can also become a powerful mechanical force, repeatedly flexing, fracturing and redistributing vulnerable sea ice.

Ocean waves entering the marginal ice zone bend large ice floes repeatedly. Under sufficiently energetic conditions, those stresses can fracture the ice into progressively smaller pieces.

The result can transform a relatively continuous ice cover into a field of fragmented floes separated by cracks and open-water leads.

And that transformation can trigger several additional processes:

• smaller floes can be moved more easily by winds and currents;
• more open water becomes exposed to the atmosphere;
• waves can penetrate farther into the ice field;
• the total perimeter of ice exposed to water increases;
• lateral melting around individual floes can increase;
• ocean-atmosphere energy exchange can intensify.

That means a cyclone doesn’t necessarily have to “melt” all the missing ice vertically.

It can physically dismantle and redistribute the ice cover.

A possible storm–wave–ice feedback

The event illustrates a potentially powerful chain of interactions:

The new study does not quantitatively prove every step in this feedback.

That’s important.

The researchers specifically describe wave-induced breakup as a plausible unresolved mechanism, rather than a measured explanation for a precise amount of ice loss.

But observations from ships provide intriguing supporting evidence.

Ships saw the ice shattered after the storm

Researchers had another source of information that satellites and models could not provide: photographs and radar observations from ships operating in the region.

Images taken before and after the cyclone showed a clear change.

Following the storm, the local ice cover appeared substantially more fragmented, with numerous individual floes separated by narrow leads and patches of water.

Shipboard radar also detected fragmented ice and features consistent with wave activity within the ice field.

The observations cannot tell scientists exactly how much ice was broken by waves.

But they are consistent with the idea that the cyclone was doing more than warming the ice.

It was physically rearranging the Arctic surface.

The models struggled during the storm too

There is another important finding buried in the study.

The researchers compared sea-ice concentration represented by the ERA5 reanalysis with observations from the AMSR2 satellite.

During the cyclone, discrepancies reached as much as approximately 20%.

After the storm passed, the difference fell below roughly 5%.

That matters because extreme events expose weaknesses that may be much less obvious during ordinary conditions.

Operational sea-ice models must somehow represent an extraordinarily complicated boundary where atmosphere, ocean, waves and ice are interacting simultaneously.

And sea ice isn’t simply a stationary sheet whose thickness changes with temperature.

It fractures, drifts, rotates, compresses, spreads and collides.

The authors argue that better representation of wave–ice interactions and dynamic ice processes will be necessary to improve simulations of sea ice during severe Arctic weather.

Are Arctic cyclones hurricanes?

No. Despite sometimes reaching extremely low central pressures and producing violent winds and waves, Arctic cyclones are fundamentally different from hurricanes and tropical cyclones.

Hurricanes derive much of their energy from warm tropical ocean water and latent heat released by deep convection.

Arctic cyclones are generally extratropical low-pressure systems powered largely by contrasts in atmospheric temperature and pressure.

But central pressure alone can still become remarkably low.

A pressure around 955 hPa is comparable to the central pressure of some hurricanes — even though the physical structure and energy source of the storms are very different.

Why autumn storms matter so much

A major cyclone striking thick winter sea ice is interacting with a very different surface from one crossing newly forming autumn ice.

During freeze-up, large parts of the marginal ice zone can consist of young, relatively thin ice.

That makes the timing of an extreme storm important.

Young ice can be more vulnerable to deformation and wave-induced breakup, while open-water areas allow large waves to develop before they reach the ice edge.

The authors therefore caution that their findings are most directly applicable to Arctic cyclones occurring during the autumn freeze-up and early ice-growth period.

Does this mean Arctic sea ice stopped freezing?

No.

This was a regional and temporary disruption of the seasonal freeze-up — not a reversal of winter across the entire Arctic Ocean.

Sea-ice concentration resumed its seasonal increase after the cyclone.

The significance of the event lies elsewhere.

It demonstrates how an extreme storm can temporarily overpower the normal seasonal direction of change in a vulnerable part of the ice pack.

October says: freeze.

The cyclone says: not so fast.

And that is the real anomaly

The remarkable part of this story isn’t simply that an Arctic storm broke sea ice.

Scientists have known for years that powerful waves can fracture ice.

It is the combination of circumstances:

the strongest October western-Arctic cyclone in three decades arrived during freeze-up, generated huge waves and abnormal oceanic heat conditions, fragmented the developing ice cover and produced an ice response that thermodynamic estimates alone could not fully reproduce.

Instead of thinking about Arctic sea ice as a giant thermometer — warmer means melt, colder means freeze — events like this reveal a much more chaotic system.

Atmosphere, ocean, waves and ice are constantly exchanging energy and momentum.

Sometimes a sufficiently powerful storm can reorganize that system in days.

And in October 2022, for a brief period in the western Arctic, winter was arriving…

but the ice went backwards.