Arctic ocean moorings shed light on winter sea ice loss



The eastern Arctic Ocean’s winter ice grew less than half as much as normal during the past decade, due to the growing influence of heat from the ocean’s interior, researchers have found.

The finding came from an international study led by the University of Alaska Fairbanks and Finnish Meteorological Institute. The study, published in the Journal of Climate, used data collected by ocean moorings in the Eurasian Basin of the Arctic Ocean from 2003-2018.

The moorings measured the heat released from the ocean interior to the upper ocean and sea ice during winter. In 2016-2018, the estimated heat flux was about 10 watts per square meter, which is enough to prevent 80-90 centimeters (almost 3 feet) of sea ice from forming each year. Previous heat flux measurements were about half of that much.

“In the past, when weighing the contribution of atmosphere and ocean to melting sea ice in the Eurasian Basin, the atmosphere led,” said Igor Polyakov, an oceanographer at UAF’s International Arctic Research Center and FMI. “Now for the first time, ocean leads. That’s a big change.”

Typically, across much of the Arctic a thick layer of cold fresher water, known as a halocline, isolates the heat associated with the intruding Atlantic water from the sea surface and from sea ice.

This new study shows that an abnormal influx of salty warm water from the Atlantic Ocean is weakening and thinning the halocline, allowing more mixing. According to the new study, warm water of Atlantic origin is now moving much closer to the surface.

“The normal position of the upper boundary of this water in this region was about 150 meters. Now this water is at 80 meters,” explained Polyakov.

A natural winter process increases this mixing. As sea water freezes, the salt is expelled from ice into the water. This brine-enriched water is heavier and sinks. In the absence of a strong halocline, the cold salty water mixes much more efficiently with the shallower, warm Atlantic water. This heat is then transferred upward to the bottom of sea ice, limiting the amount of ice that can form during winter.

“These new results show the growing and spreading influence of heat associated with Atlantic water entering the Arctic Ocean,” added Tom Rippeth, a collaborator from Bangor University. “They also suggest a new feedback mechanism is contributing to accelerating sea ice loss.”

Polyakov and his team hypothesize that the ocean’s ability to control winter ice growth creates feedback that speeds overall sea ice loss in the Arctic. In this feedback, both declining sea ice and the weakening halocline barrier cause the ocean’s interior to release heat to the surface, resulting in further sea ice loss. The mechanism augments the well-known ice-albedo feedback — which occurs when the atmosphere melts sea ice, causing open water, which in turn absorbs more heat, melting more sea ice.

When these two feedback mechanisms combine, they accelerate sea ice decline. The ocean heat feedback limits sea ice growth in winter, while the ice-albedo feedback more easily melts the thinner ice in summer.

“As they start working together, the coupling between the atmosphere, ice and ocean becomes very strong, much stronger than it was before,” said Polyakov. “Together they can maintain a very fast rate of ice melt in the Arctic.”

Polyakov and Rippeth collaborated on a second, associated study showing how this new coupling between the ocean, ice and atmosphere is responsible for stronger currents in the eastern Arctic Ocean.

According to that research, between 2004-2018 the currents in the upper 164 feet of the ocean doubled in strength. Loss of sea ice, making surface waters more susceptible to the effects of wind, appears to be one of the factors contributing to the increase.

The stronger currents create more turbulence, which increases the amount of mixing, known as shear, that occurs between surface waters and the deeper ocean. As described earlier, ocean mixing contributes to a feedback mechanism that further accelerates sea ice decline.

Accelerated currents have practical implications in the Arctic. Ship captains need accurate maps of currents for navigation. Since currents move sea ice, oil and gas extraction activities also need information about currents.


This second study was described in a scientific paper published in the Geophysical Research Letters.

Additional co-authors for these papers include Ilker Fer, Matthew Alkire, Till Baumann, Eddy Carmack, Randi Ingvaldsen, Vladimir Ivanov, Markus Janout, Sigrid Lind, Laurie Padman, Andrey Pnyushkov and Robert Rember.

From EurekAlert!

116 thoughts on “Arctic ocean moorings shed light on winter sea ice loss

    • They throw data and make assumptions. Not very scientific. They say if the water is warming, it proves man made global warming. If it is cooling, it is just a anomaly.
      The Atlantic sea water that reaches the Arctic waters are mostly determined on the Gulf Streams salinity.
      The more salinity the more the movement of the current. This salinity drops at the Arctic/Atlantic boundary and shifts towards the south, this is what causes the Gulf Stream current. If the salinity falls the current slows and more Atlantic water enters the Arctic.
      Being that is the main contribution of heat in the Arctic waters, we can assume that less saline water is in the gulf stream there by allowing more heat to enter the Arctic.

      • There is very likely a regular process that drives ice area reduction in the Arctic ocean. There is STRONG anecdotal evidence of greatly reduced ice area from the early 20th century and from the 19th century. These reports suggest that there had to have been intervening periods of ice expansion.
        It is important to remember the physical realities of the seasonal variation at those latitudes. Virtually no sun for 3 months means very high heat loss from open water. It is an area of regular high winds which increases the mixing of the surface layer.
        An objective view of the climatic forces at work on the mixing bowl of salt water at the top of the world indicates an oscillating balance between the open water, heat losing mode and an ice covered and less distubed heat retaining mode.
        We hit the bottom limit in 2012 and are now in maximum heat shedding mode.

    • Your link says no.

      “We don’t believe the volcanoes had much effect on the overlying ice,” Reeves-Sohn told LiveScience, “but they seem to have had a major impact on the overlying water column.”

      • If you’re replying to me, then I know they won’t directly melt the ice, but the article makes clear they are injecting vast amount of heat into the water column. As this heat is coming from the deep, it will be mixing up the stratification, which is precisely what the top post is claiming to find, without attributing anything to a substantial source on the doorstep (so to speak).

        Unlike some, before I post I do read, and yes I did search out the original article, but summaries work better for some than the dry verbiage of a paper.

    • Of course geologic process affect sea ice, in many ways, even during the toasty Early Pliocene (4.5 Ma), with 400 ppm CO2 and boreal forests around the Arctic Ocean:

      On the causes of Arctic sea ice in the warm Early Pliocene,Iceland%20Sea%20at%204.5%20Ma.

      Influence of Sea Ice Expansion on Pliocene Arctic Climate

      Despite the relatively high Early Pliocene atmospheric CO2 concentrations of 380–400 ppm17,18, the changes in North Atlantic and Arctic paleoceanography caused sea ice to occur as far south as the Iceland Sea at 4.5 Ma. This development was likely controlled by the emergence of a modern-like EGC, which was established as a consequence of a surface water flow reversal across the Bering Strait14. The changed paleoceanography was crucial to allow the import of fresher, cooler Arctic water and sea ice into the Iceland Sea. Whether sea ice was exported directly from the Arctic, or the fresher, cooler water favored local sea ice formation in the Iceland Sea yet remains ambiguous. Nevertheless, the effects of a changed EGC and appearance of sea ice along its pathway are also recognized from the dinoflagellate cyst turnover in the Iceland Sea14 and the onset of biosiliceous sedimentation in the Labrador Sea15 at 4.5 Ma. Around that time, small IRD amounts are recorded at ODP Site 907, indicating that Greenland did have ice caps or small ice sheets that could produce icebergs and ice rafted detritus (IRD)38, albeit in volumes considerably lower than in the Late Pliocene and Quaternary. Tectonic uplift made elevated plateaus available in Greenland during the Early Pliocene39 where glaciers and ice caps could nucleate and eventually expand into a large, IRD producing ice sheet. Such large ice sheet started to deliver considerable IRD into the North Atlantic during the Late Pliocene40, when the zonal gradient in the Nordic Seas was already established. In fact, the cooler water and more substantial sea ice presence in the Iceland Sea after 4.0 Ma and in the Late Pliocene16, may have contributed to the gradual expansion of continental ice in Greenland. A more substantial and long-term sea ice presence along the East Greenland coast acts to thermally isolate Greenland from relatively warmer Atlantic waters41 and reduces heat advection42, as well as providing a sea ice-albedo feedback and inhibiting ocean–atmosphere heat exchange. These combined effects, together with the tectonic uplift of the circum-Arctic land masses43 all provide positive feedbacks for expansion of the Greenland Ice Seet (GIS).

    • Not just volcanoes, but soot from China, Russian icebreakers and Siberian industrial development warming rivers flowing into the Arctic Ocean. Cleaner sky over North America and Europe since the 1970s might help heat ice, too.

  1. 1) :
    Quote: “The moorings measured the heat released from the ocean interior to the upper ocean and sea ice during winter. In 2016-2018, the estimated heat flux was about 10 watts per square meter,“

    This is the first time I have seen actual measurement of heat flux rising from the interior. Generally this figures tends to be assumed to be in terms of micro watts and therefore gets ignored as an influence in climate explanations. It seems that this is not so and that there can be locations where large energies can be released from the interior in specific locations which would have considerable influence on the climate.
    A case in point could well be involved where El Niño and La Nina are concerned.

    Quote: “In this feedback, both declining sea ice and the weakening halocline barrier cause the ocean’s interior to release heat to the surface, resulting in further sea ice loss. “

    I suspect the hypothesis here is back to front; the ‘cause’ being the 10 Watts/sq.m rising from the interior. It seems that there is a bit of mushy conjecture here.

      • So if there is a 10W/mm^2 increase in upward heat flow, OHC must be dropping. This will also lead to more LWIR radiated out to space. Earth’s energy enters in the tropics and leaves via the poles.

        The mechanism augments the well-known ice-albedo feedback — which occurs when the atmosphere melts sea ice, causing open water, which in turn absorbs more heat, melting more sea ice.

        “Well known” *assumed* albedo feed back , which carefully ignores outward flux and changes in evaporation the thermal insulation of ice cover.

        When these two feedback mechanisms combine, they accelerate sea ice decline.

        Well why done we SEE an acceleration then ?? The first OMG arctic sea ice deline was in 2007 summer minimum. Despite an even more OMG, OMG minimum in 2012, there was a 65% INCREASE one year later. In 2018 and 2019 the summer min. was identical to that of 2007.


    • The micro-watts probably is referring to the heat flow from the interior of the earth.
      In this article the seem to be talking about heat flow from a warmer layer of water that is a few meters below the sea surface.

  2. So we have a study from 2003 to 2018, which then shows a change from 2016 to 2018. The study is of an extremely complicated and chaotic earth system, namely ocean AND atmosphere currents and their interaction. This 15 year study, with 2 years of anomaly, can correctly model the arctic ice system? Forever more? Not. This type of study, while contributing actual scientific data, cannot possibly detect all of the historical permutations of chaotic systems interacting, and therefore cannot possibly predict future such interactions. The only useful mark of the solid – liquid – gas phase of the hydrosphere of our planet is sea level, and this must also, at a minimum, be corrected for the position of drifting continents (and just for fun add in some of IanH’s volcanos).

  3. Hey look on the bright side all the swimming bears can stay a bit warmer … oh wait the bears are extinct.

      • The Chartic data is a 5-day average not the daily data. Average melt for the remainder of the season would comfortably take extent to second lowest, today’s NSIDC area is already lower than all but two of the previous September minima.

      • Gary,
        Don’t confuse the Griffter with data! His comments are analogous to solar and wind energy output: Imaginative, Intermittent and Unreliable.


      Still being a climate change denier , hey griff.

      Current levels of sea ice are in the top 5-10% of the last 10,000 years.

      Currently nowhere near their lowest over any meaningful time span.

      “in the 41 year “ SO WHAT !! meaningless !!

      Starting at the anomalously high sea ice extent of the late 1970s, is childish, ignorant and anti-science..

      • For most of the last 10,000 year, Arctic sea ice has been a lot less than now.

        Biodata clearly shows that the Arctic was often genuinely “ice free” (no not a Wadham’s Ice free)

        Yet the world is still here.

        Polar bears, and other Arctic creatures are still with us.

        Reduced Arctic sea ice is a PLUS, not a negative.

        Unfortunately, doesn’t look like dropping much further than to 4 Wadhams each year..

        Sort of leveled out as the AMO starts to turn downwards.

      • Disruption of NH weather patterns, weakening of the polar jet resulting in prolonged blocking patterns for example or a collapse of the three cell pattern to the equable state.

          • What does her climate gibberish even mean?

            I was born and brought up in the Northern hills of England. I’ll tell you what it means – it means that the weather will be a bit less sh!t, equally sh!t, or a bit more sh!t on a daily and weekly basis.

            I still love the place though. As they say “mustn’t grumble”.

        • Do you have any evidence that any of that would actually happen, or that it would be a bad thing if it did?

        • “Disruption of NH weather patterns, weakening of the polar jet resulting in prolonged blocking patterns for example or a collapse of the three cell pattern to the equable state.”

          Loydo, how far back in the Earth’s history did you look to see if this, if it’s actually occurring, is an anomalous occurrence?

    • NSIDC shows yesterday higher than 2019 and 2012, and about the same as 2007, 2011 and 2016. It’s just weather.

      If CO2 affects sea ice, why did the Antarctic grow from 1979 to a record high in 2014? Why has it recovered from the two freak weather events which lowered it in Super El Niño 2016?

    • Second lowest for the satellite era, which started at a time when sea ice was at it’s highest levels in the last 100 years.

      Regardless, griff has been predicting the imminent demise of sea ice ever since the record low levels seen in 2012. Meanwhile the real world can’t even manage to get close to the 2012 record for a second time.
      Forcing griff to concentrate on meaningless single day readings.

    • Arctic conditions in context NSIDC:

      “From July 27 through August 8, 2020, extent declined 470,000 km² (181,000 mi²), which is less than half of the average 1981 to 2020 extent loss of 950,000 km² (367,000 mi²) during the same period. After August 8, the rate of loss increased again due in part to melt in the Beaufort and Chukchi Seas, though the loss rate was still slower than average.”

      As of August 17th Sea ice extent stood at 5.15 million km² (1.99 million mi²).

      September 2007 sea ice extent was 4.28 million km² (1.65 million mi²).
      September 2012 sea ice extent was 3.41 million km² (1.32 million mi²).

      ……place your bet’s 🙂

    • Go easy on Griffie. At least this time he stated the length of his record, 41 years. He’s making progress.

      Hate to point this out to you though, Griff. Earth history goes back a lot further than 41 years.

  4. Polyakov and his team hypothesize that the ocean’s ability to control winter ice growth creates feedback that speeds overall sea ice loss in the Arctic. In this feedback, both declining sea ice and the weakening halocline barrier cause the ocean’s interior to release heat to the surface, resulting in further sea ice loss. The mechanism augments the well-known ice-albedo feedback — which occurs when the atmosphere melts sea ice, causing open water, which in turn absorbs more heat, melting more sea ice.

    The earth’s climate has been reasonably stable for billions of years. It’s difficult to believe that now we have discovered yet ANOTHER positive feedback system controlling it. Positive feedback is highly unstable. Just a tiny bit of offset results in an an increase in the disturbance, creating an even bigger offset. The size of the driving feedback cascades even larger and the limit of control is very quickly reached. If your room thermostat is inadvertently changed to give positive feedback, it doesn’t take days, or weeks for a problem to be revealed, within minutes your furnace will be blasting maximum output IN THE WRONG DIRECTION. It takes only the tiniest perturbation to start the cascade.

    If our climate contained all of the sources of positive feedback that are postulated by the “climate scientists” of today, we would have had a Venus climate for all these billions of years, and wouldn’t be here to speculate about positive feedback.

    • “In this feedback, both declining sea ice and the weakening halocline barrier cause the ocean’s interior to release heat to the surface, resulting in further sea ice loss.”

      Lets think about this one for a bit.
      A *Positive* feedback? OK.

      “cause the ocean’s interior to release heat to the surface”
      Releases heat. More heat released than otherwise. In the arctic, that can only mean *net* cooling. Heat is released from Earth to outer space at the polar regions.
      So longer term, this is a net cooling, and therefor a *Negative* feedback.

  5. What is ‘normal’?
    And how often does ‘normal’ actually occur, and for what duration?

  6. Some things the authors didn’t or couldn’t comment on.

    Water vapor. Humidity levels should be increasing in the areas where the is emanating.

    LWIR. Warmer water should be emitting more radiation. Has this shown up in the areas being studied.

    Origin of the heat. Neither land nor CO2 can heat deep water. Where did the heat originate and more importantly, when and why.

    • Good questions. Fingers-crossed that Dr. Roy Spencer reads this thread. I think he would take a shot at the answers.

  7. Griff,
    Click your heels together three times and say:
    “warmer is better, warmer is better , warmer is better”
    That should relieve your anxiety.

  8. Polyakov et al. (2003) concluded:

    “The composite temperature record shows that since 1875 the Arctic has warmed by 1.2°C, so that over the entire record the warming trend was 0.094°C decade−1, with stronger spring- and wintertime warming. The Arctic temperature trend for the twentieth century (0.05°C decade−1) was close to the Northern Hemispheric trend (0.06°C decade−1). The oscillatory behavior of Arctic trends results from incomplete sampling of the large-amplitude LFO. For example, the Arctic temperature was higher in the 1930s–40s than in recent decades, and hence a trend calculated for the period 1920 to the present actually shows cooling. Enhancement of computed trends in recent decades can be partially attributed to the current positive LFO phase.”

    “We speculate that warming alone cannot explain the retreat of Arctic ice observed in the 1980s–90s. Also crucial to this rapid ice reduction was the low-frequency shift in the atmospheric pressure pattern from anticyclonic to cyclonic. Positive and negative LFO phases of the SAT are shifted by 5–15 yr relative to those in the SLP record. The complicated nature of Arctic temperature and pressure variations makes understanding of possible causes of the variability, and evaluation of the anthropogenic warming effect most difficult.”

    • AO: Arctic Oscillation
      LFO: Low Frequency Oscillation
      SAT: Surface Air Temperature

      The latter in title of link.

      Should be obvious that small increase in air temperature can have little effect on Arctic sea ice. Lower winter minima occur despite always being below freezing in high years and low. Lower summer minima have not occurred since 2012 despite higher CO2 and often Arctic air temps.

      Meanwhile, Antarctic sea ice is at its 1981-2010 median.

      It’s all about the oceanic oscillations.

      • Most of the early melt this year was on the Siberian side of the Arctic, caused by that anomalous warmer air mass being pushed up from central Asia.

        Also caused the near record temperatures in some part of northern Siberia.

        But also pushed the sea ice to pack up in the Greenland Alaska region.

        Laptev, Kara and Siberian regions all melted early because of this, while Beaufort, Chukchi and the Greenland Sea are still quite full of sea ice.

        • July windstorms drove sea ice away from Alaska and NE Siberia (Beaufort and Chukchi Seas), spreading out and piling up the floes. Those weather events account for this year’s low ice extent, liable to come in second in the dedicated satellite era, ie since 1979, but still well above current summer minimum record holder 2012. The lowest years were caused by storms, usually late summer cyclones, ie 2007, 2012, 2016 and 2019. There were two in 2016. Dunno about 2011, another low year.

    • What also cannot be explained is the rapid increase in sea ice extent from the mid 1970’s to the late 1970s.

  9. 41 years of satellite records of polar ice. 16 years of records from this study, strangely enough they only discuss three of those years. The Earth is 4,500,000,000 years old.

    Why bother? CO2 is related to prosperity. More Prosperity, more CO2. Humans, well except those in California, will never choose laws and policies which reduce prosperity.

    And there was a Little Ice Age, and there was a Medieval Warm Period as well as several others, Roman, Minoan, etc. The ice changes. At one point it is theorized that the Earth was a giant iceball!

    Ice comes, ice goes. 1.3 Billion Chinese and another 1.2 Billion Indians all long for a new car, and lots of them are buying one. I might buy one myself one of these days, that new Lexus RC500 Convertible is gorgeous, as are the Jags.

    Drill Baby Drill….

    • Michael,
      Actually it is more like 47 years of satellite records but the first six years do not enhance the narrative and so are expunged.

  10. “The mechanism augments the well-known ice-albedo feedback”

    It really is sad when things that don’t exist, become well known.

    First off, because of the low angle of the sun, even at the height of summer, most sunlight is reflected.
    Secondly, they are talking about the formation of winter ice. THERE IS NO SUNLIGHT IN THE FIRST PLACE DURING THE WINTER.

  11. Oooops . . . the Prophet Al said all Arctic Ice would be gone by now . . . why are we so concerned as to be nit-picking over “eastern Arctic Ocean’s winter” ice extent?

  12. Five years ago I posted in WUWT an article on the Arctic Iris Effect which outlined the storage and release of heat via incoming warm Atlantic water in the Arctic. I believe much of the Arctic Amplification was mis-diagnose and blamed on changes in albedo while ignoring the heat released from Arctic’s warm Atlantic layer. The current slow down of intruding Atlantic water suggests after some heat ventilation, the Arctic sea ice will be returning within a decade or two

    • Which is no surprise given the cyclic nature of Arctic sea ice extent, driven by the ENSO, PDO, AMO, AO and other oceanic oscillations.

  13. I don’t (as usual) remember the details, but a few months back WUWT treated us to a paper that claimed increased melting in Greenland, Baffin etc. was causing more cold fresh water to enter the Arctic Ocean. The lighter fresh water layer stays between the salt water and the ice. The current study says that this fresh water layer is thinning and allowing the salt water to melt the ice from below.

    So there we have it. More fresh water (bad – because the Greenland ice cap is dying), or less fresh water (bad – because the sea ice is dying). We can pick the one that scares us most. Isn’t climate science a wonderful thing?

Comments are closed.