The Arctic Sea Ice Canary Refuses to Die

In late 2007 the Arctic sea ice area took an unexpected plunge and this event is largely responsible for triggering the Global Warming hysteria of recent years. This led leading warmist James Hansen to declare that man made global warming was now having a clear impact on the Arctic and the chorus warned of changing albedo and warming seas that may trigger an explosion of methane from melting permafrost and seabed methane hydrates. Everything was much worse than was believed before.

For the five following summers the Arctic Sea ice experienced similar higher than “normal” summer melt back culminating with “unprecedented melting” in the late summer of 2012 (Figure 1). But then last year, the high melt back did not occur and it seems unlikely to occur this year. Whilst it is too early to say, things appear to have returned to “normal”. That is good news, isn’t it? The Arctic Sea Ice canary’s refusal to die may turn out to be another nail in the coffin of global warming theories.

Figure 1 Detail for Arctic Sea Ice Area anomaly showing the significant summer melt back in the years 2007 and 2012. In 2013 and 2014 there are signs that the sea ice is returning to its pre-2007 state and there is little evidence for a terminal death spiral model favoured by warmists. The illustration is adapted from the full Arctic sea ice record as reported by Cryosphere Today.

This post could easily have been titled “Sea Ice for Beginners” since its main purpose is to provide the uninitiated with a perspective on global sea ice variations and what they tell us.

What is Sea Ice?

In the Arctic and the Antarctic the sea freezes in winter time when one or other of the poles points away from the sun and it is perennially dark and very cold. Seawater freezes at about -2˚C but the sea ice itself is fresh water ice hence the freezing process expells the salt resulting in the salinity of seawater surrounding sea ice to increase. This extra salty cold water is dense and it sinks to the ocean floor, a process instrumental in global thermohaline circulation.

In the Arctic, sea ice fills the Arctic ocean basin that covers the North pole and many shelf seas and bays at high latitude such as Hudson’s bay (Figure 2). In the Antarctic, the sea ice forms an apron around Antarctica. In the Antarctic, sea ice is not to be confused with ice shelves which are much thicker slabs of land ice that have slid out across the ocean. Ice shelves are not frozen seawater but are compacted snow fall. Ice shelves are dynamic, like glaciers, and from time to time ice shelves may fragment and break up.

Figure 2 Arctic sea ice as of 2 September 2014 is approaching its minimum extent and towards the end of September melting will be replaced by freezing. In winter time sea ice fills the whole Arctic basin. Note that there is already some ice on Lake Baikal and some northern Canadian lakes. I’m unsure if this is left over from last year or represents early freezing this autumn.

The Freeze – Thaw Sequence

In the Arctic, the sea ice begins to freeze in September following the autumn equinox since this marks the turning point of dark exceeding light. And it begins to melt in March following the spring equinox for the opposite reason. Sea ice area grows and shrinks on an annual basis as shown for the Arctic in Figure 3. The picture for the Antarctic is the same, but “seasonally” reversed. The maximum area of Arctic sea ice is about 14 million square kms and the Antarctic about 16 million square kms. The Antarctic sea ice sits at higher latitude and this combined with the larger area means southerly sea ice dominates the sea ice albedo effect.

Figure 3 The sinusoidal growth and decay pattern for Arctic Sea ice. Note how two years ago the late summer minimum was about 2 million square miles less than today. Chart from Cryosphere Today.

Sea Ice Data Sources

The satellite record of sea ice area begins in 1979 (35 years of data). The primary source for data is the National Snow & Ice Data Centre (NSIDC) but the primary reference web site is usually Cryosphere Today run out of the University of Illinois. Cryosphere Today provides all the charts and maps one could wish for but the presentation is tainted by a warming bias.

One other exceptional resource is the Watts Up With That sea ice page that has all the charts and maps from Cryosphere Today plus a lot more.

Sea Ice Nomenclature

There are two main measures of sea ice: AREA and EXTENT. In comparing data sets it is essential to know what measure is plotted. NSIDC provide this clarification:

Area and extent are different measures and give scientists slightly different information. Some organizations, including Cryosphere Today, report ice area; NSIDC primarily reports ice extent. Extent is always a larger number than area, and there are pros and cons associated with each method.

A simplified way to think of extent versus area is to imagine a slice of swiss cheese. Extent would be a measure of the edges of the slice of cheese and all of the space inside it. Area would be the measure of where there is cheese only, not including the holes. That is why if you compare extent and area in the same time period, extent is always bigger. A more precise explanation of extent versus area gets more complicated.

Furthermore, it is common practice to plot the anomaly rather than the measured values. This is an acceptable methodology that amplifies real world changes, but since it is an amplification it is also open to abuse. The anomaly is simply the arithmetic difference between the measured value and a reference value. As a reference value Cryosphere Today use the mean sea ice value from 1979 to 2008.

Arctic Sea Ice in Motion

The Arctic Sea ice is not a static sheet of ice but is rather in a state of continuous motion. The drift of the ice was first recognised when drift wood from Russia would turn up in Greenland. The drift of the ice was used by Nansen in his attempt to reach the North Pole by allowing the sea ice to enclose his vessel, Fram, in the hope that the drifting ice would take him close to the pole (see section on sea ice tails below).

WUWT have this map from the US Navy. If you check out the 365 day animation linked below you will see that the drift of the sea ice is extremely complex but that there is a strong tendency for ice to drift out into the N Atlantic. It was this drift that Nansen tried to use in his failed attempt to reach the North Pole.

Figure 4 The pattern of sea ice drift today, note that the gif is updated daily. This 365 day animation gives an idea of the complex dynamics of the Arctic sea ice.

Sea Ice Tales

The Finn Men of Scotland

Around 1700 AD there were many sightings of Finn Men in kayaks off the northern Orkney and Shetland islands of Scotland. One of them showed up in Aberdeen where he was captured and then died. His kayak is in a museum in the city to this day (Figure 5).

Figure 5 The frame of this Eskimo canoe is not whale bone but is made of timber from a european tree, Pinus Silvestris (Scots Pine) providing a strong clue that these Eskimos came to Scotland via Norway.

There is agreement that the Finn Men were Eskimos but no clear idea how they came to be so far out of their normal range. This lengthy account by David Macritchie from 1912 makes interesting reading for those who have time. To the various options considered I would like to add the possibility that sea ice, far more extensive than today, linked to The Little Ice Age, may have played a role in either bringing the Finn Men from Greenland via Iceland and the Faroes or bringing them to Norway from where they made their way to Scotland.

Figure 6 shows the drift ice index of Bond et al 2001 [1], annotated with historic events. The up arrow (close to the year 0) marks the arrival of Finn Men off Scotland. Bond’s data suggests that at this time drift ice came far south into the Atlantic. This could have provided fresh drinking water and shelter for the Finn Men on an ocean crossing from Greenland to northern Norway or from Greenland via Iceland and Faroes to Orkney. The former seems more plausible.

Figure 6 The North Atlantic drift ice index from Bond et al [1]. The drift ice index is based on the proportions of ice rafted mineralogical debris from discrete sources on Greenland and Iceland. Cycles 0 to 8 represent deep drift ice incursion to the North Atlantic. The drift ice index correlates with the 10Be record from Greenland and with known warming and cooling episodes: LIA = Little Ice Age; MWP = Medieval warm Period; DA = dark Ages; RWP = Roman Warm Period. The up arrow to the left indicates Eskimo sightings off Scotland.

Nansen’s Fram expedition 1893-1896

Nansen’s Fram expedition of 1893 to 1896 is another epic tale that can be read on Wikipedia. Nansen sailed his ship to a point close to the New Siberian Islands off the North coast of Russia where it became frozen fast in the ice in September 1893. She emerged in open water near Spitzbergen in August 1896, the three years of drifting, stuck in the Arctic sea ice, proving the southerly drift of the Arctic sea ice.

Figure 7 The Fram stuck fast in Arctic sea ice 1894. The windmill brought a smile. I’m not sure what it was for.

Ice Free North Pole 1987

Three nuclear submarines famously rendezvoused on the surface of the Arctic Ocean in May 1987.

Figure 8 One British and two US submarines at a slushy North Pole, May 1987.

The purpose of these sea ice tales is to illustrate the fact that the Arctic Sea ice is dynamic. We have only 35 years of satellite data upon which to make judgements about what is normal and what is not normal. It seems quite likely, as a result of cooling and warming episodes (Bond Cycles) during the Holocene, that summer Arctic sea ice may have in the recent past been much more and less extensive than today. Jumping to the conclusion that the melt backs of 2007 to 2012 is down to manmade warming seems unwise and lacks supporting evidence.

Sea Ice Hockey

Cryosphere Today provides the following picture of historic Arctic Sea ice extent (Figure 9). Note that Cryosphere are now shifting from area to extent. There are two key observations to make. The first is that satellite era data (starts 1979) somehow splices exactly with whatever data was used before which seems highly improbable and second the data from 1900 to 1960 are pretty well a flat line. And so this is a classic hockey stick where the invariant past gets replaced by a warming recent and present era.

Figure 9 A seasonal and annual sea ice reconstruction from Cryosphere Today.

It is highly relevant to ask how this chart has been compiled since to make a sea ice map would require thousands of instantaneous observations made across the whole of the periphery of the sea ice at least once per week if not daily. There are of course shipping records etc but compiling these into a reliable picture comparable with satellite measurements is a major challenge – though not impossible (see below).

The documentation provided by Cryosphere Today on how the above chart was compiled is wanting. An example of a gridded map is shown in Figure 10 where recognising the outline of Greenland and North America should allow your brain to fill in the rest (standing back from your computer will help). The numbers refer to the data sources (see here). Zero = open water and the stars I assume = presumed ice. It is difficult to accept this as a high precision or reliable record.

Figure 10 Gridded data sources used to make the Cryosphere Today historic sea ice reconstruction.

Another source for historic sea ice data comes from this amazing article by Frank Lanser in WUWT. The article displays many beautiful maps compiled by DMI (Danish Meteorological Institute?). An example is shown below. Simply scanning through the maps, that are mainly for the month of August, shows that Arctic sea ice extent has fluctuated significantly throughout the whole of the 20th Century.

Figure 11 An example of the many fantastic sea ice maps published by Frank Lanser on WUWT shows significant fluctuations in Arctic Sea ice at odds with the Cryosphere Today hockey stick.

Sea Ice Volume

The unexpected melt back of 2007 caught the IPCC unaware since the ice retreat was more rapid than their models predicted. This led to claims that things were much worse than presumed whilst in fact all this showed is that the IPCC models were wrong.

It was quickly realised that it was not just ice extent but the thickness of the ice that was important too giving rise to the excuse to burn a few million dollars more studying present and past sea ice volume. Volume had been declining for many years and in 2007 the ice reached a critical thinness and near disappeared altogether. This process now seems to be stabilised and if anything is being reversed with renewed ice volume growth. WUWT provide this image of sea ice thickness from the US Naval Research Laboratory.

Figure 12 Sea ice thickness. Note that this gif image is regularly updated. Anyone still not convinced about the dynamic nature of Arctic sea ice may want to view this 365 day animation.


The warmists analytical model begins with CO2 accumulation in the atmosphere leading to atmospheric warming (curiously absent for 16 years) leading to sea ice loss. The loss of sea ice reduces the albedo in the Arctic since ice is highly reflective and water absorbs much of the incoming radiation. Replacing ice with water will lead to accelerated warming of the ocean and a runaway positive feedback culminating in melting of permafrost and submarine clathrates and a tipping point for climate melt down. If Earth’s climate was so finely balanced we would not be here to tell the tale.

The late John Daly provides an alternative model on his archived web site. John notes little evidence for warming of the air in the Arctic during the period of ice wastage (Figure 13). He also notes that ocean currents transport a very large amount of heat into the Arctic Ocean and in particular the Gulf Stream (Figure 14). A small increase in the flow rate of the Gulf Stream could account for the observed thinning of the ice. This model appears to be consistent with the pattern of ice loss in the Barents Sea and around Novya Zemlya. And it also tallies with the model of Bond et al [1] that proposed periodic ingress of the Labrador current into the Gulf Stream allowing for drift ice to be transported much further south than at present. When this cyclical process repeats, Northern Europe will freeze.

Figure 13 The temperature record for Vardoe, Northern Norway courtesy of Clive Best who provides this excellent station temperature application for hundreds of climate stations world wide.

Figure 14 Map from John Daly showing water flows into the Arctic Basin.


To complete this beginners tour of sea ice data it is of course important to add a look at what has been going on at the other end of the world. The surprising reality for those unfamiliar with the data is that Antarctic sea ice area has been expanding slowly but steadily since the satellite record began (Figure 15). What is even more surprising is that journalists visit Antarctica to view the summer melt in the belief that this is evidence for global warming. Combining the Arctic and Antarctic data sets shows that global sea ice extent has been essentially uniform for the last 35 years. The global anomaly has in fact been mainly positive for the last 36 months. So why all the fuss about melting sea ice? You may well ask.

The warmist explanation for expanding Antarctic sea ice is accelerated melting of the Antarctic ice sheet resulting in freshening of seawater around the continent.

Figure 15 Sea ice anomaly from Antarctica. Source is Cryosphere Today.


  • The Arctic sea ice is an incredibly dynamic feature and it seems highly likely that its area and mass fluctuates as a result of natural processes on decade and century time scales. Shifting ocean currents is but one such example of natural cyclic change that may impact the Arctic sea ice.
  • Because of its dynamic nature it is unwise to use the Arctic Sea ice as a datum for differentiating natural from manmade climate processes.
  • The extensive melting of 2007 to 2012 was not repeated in 2013 and seems unlikely to be repeated this year. Melting in the Arctic has been compensated by freezing round Antarctica and global sea ice area has been remarkably uniform over the 35 year period of satellite measurements.

[1] Bond et al. Persistent Solar Influence on North Atlantic Climate During the Holocene. Science 294, 2130 (2001)

This entry was posted in Climate change, Climate change for beginners and tagged , . Bookmark the permalink.

16 Responses to The Arctic Sea Ice Canary Refuses to Die

  1. The existing scientific paradigm asserts a drop in temperature causes a physical phase change in sea-water from liquid to ice.

    But if you place your plasma physics hat on, you might think that an increase if solar protons into the polar regions, via the Birkeland currents, see Peratt, causes ice formation which then absorbs energy from the environment, thus causing freezing.

    Temperature in this case is an effect, not an affectation.


  2. And Gerry Pollack – Uni Washington – and EZ water theory. (A senior moment).

  3. Joe Public says:

    Thanks for your Primer, Euan.

    Would I be alone in wondering that if one of the variables moves in a direction unsupportive of a particular hypothesis, a different variable is ‘invented’? We have sea ice area, extent, and more-recently thickness.

    • Euan Mearns says:

      I sense a touch of cynicism. When lower troposphere warming is no longer an essential pre-requisite of lower troposphere warming. I can hardly wait for the Rhone glacier to begin a new advance, but I doubt I will live to see it 🙁

  4. Dominik Utz says:

    Hm.. very interesting discussion and article but on the cited Sea Ice Page (Watts Up With That sea ice page) there is also this nice plot giving the yearly average extents over the year averaged over the last decades. It is showing that the averages over each of the succeeding 10 year frames is always lower than the averaging over the previous 10 years. So even though this year is not a year of absolute minimum the curve is still below the 2000´s average.
    Also the minimum point of the 10 year averaged curves seems to be shifted slightly to the right, which would mean that probably the temperatures are slightly higher and the melting takes place slightly longer.
    All I want to say is that as one is not allowed to state that there is global warming because a year is exceptional warm also one is not allowed to do the opposite and state that there is no warming as this year is not as extreme as the previous one. And from the data of this graph it looks like this year beats still all the averages of the last 30 years of data…
    And yeap.. antarctica looks to be a very nice story… no clue 😉

  5. A C Osborn says:

    Just about everyone who has bothered to look into it knows that in 1979 Arctic sea ice was at a maximum and that it is cyclical. Analysis of older Sattelite data shows that prior to 1979 sea ice was a lot lower.
    Anecdotal evidence, especially from Admiralty Records shows that the ice ebbs and flows

  6. s.h. says:

    I was surprised to see this discussion of sea ice. I thought it was a red herring. I thought that one should distinguish between sea ice and ice.

  7. I suspect that if the data were available all of the Arctic would look like this:

  8. Luís says:

    Congratulations Euan, this is a superb post. I wish I had the time to dive this deep into certain things.

    Just a quick note on Figure 10. In 1990 graphical computer stations were still rare (and very expensive) and the software was not really there yet (ArcView was released in 1991). So Figure 10 may just be an inaccurate representation of accurate data.

    In a follow up post it would be nice to understand what “accelerated melting of the Antarctic ice sheet” really means.


  9. “The ship also included a windmill, which ran a generator to provide electric power for lighting by electric arc lamps.”

    • Euan Mearns says:

      Todd, its interesting information, but I don’t understand how anyone can bring themselves to write an article like that. The title:

      Greenland and Antarctic maps reveal “unprecedented” ice loss

      and the text:

      A series of maps published this week show Greenland and Antarctica are losing more ice than at any time since satellite records began.


      The satellite was launched in 2010 and has been collecting data on sea ice and ice sheets ever since.

      So we have 3 to 4 years of data upon which sweeping generalisations are made. We are in an inter-glacial period and these are characterised / defined by ice sheets melting. What else are they supposed to do?

      • Todd De Ryck says:

        Good points there Euan, do you read Jason Box blog?

        • Euan Mearns says:

          Todd, that’s also interesting information. My key observations are:

          1) The scale is in parts per billion (ppb) so we are dealing with around about 1.9 parts per million (ppm) methane. 10,000 ppm = 1%, so this is 0.00019% methane.
          2) The rise in Barrow is about 200 ppb or 0.2 ppm or 0.00002% since 1984.
          3) The annual cycle has a range of about 100 ppb or 0.1 ppm. So the annual cycle is scaled to about 50% of the overall rise since 1984.
          4) The Dragon breath spikes are of the order 0.1 ppm or 0.00001% – I ain’t scared of this dragon.

          Its fantastic data. A credit to the ingenuity of Man to invent such precise analytical tools. Plot the data zero scaled with % (or ppm) on the Y-axis to get a real picture. All the variance will likely fall within a completely horizontal line one pixel thick.

  10. Pingback: The Methane Time Bomb | Energy Matters

Comments are closed.