Is the Arctic really so much more interesting than the Antarctic?

This week NSIDC came with its yearly end of the season press release about sea ice. Their title was: Arctic sea ice shatters previous low records; Antarctic sea ice edges to record high. Kudos to NSIDC for mentioning the Antarctic record in their title. As many of you will have noted, the Antarctic sea ice reached about the same amount of sea ice as the record year of 2007. Wait a second, wasn’t 2007 also the year in which the Arctic sea ice reached a record low? Indeed, it was. So for the second time in a row the Arctic record low happened at the same time as the Antarctic record high. Is this a coincidence or is there some magical teleconnection operating between our poles?

Before we get back to that question, first some facts about the (media) attention that both the Arctic and the Antarctic are getting. Let’s do some simple ‘google’ and ‘google news’ searches for “Arctic sea ice” and “Antarctic sea ice”:

Arctic sea ice
Google: 4,220,000
Google news: 3970

Antarctic sea ice
Google: 967,000
Google news: 370

Apparently the Arctic sea ice is more interesting than the Antarctic sea ice, especially for the media. We all knew that: good news isn’t news, bad news is news.

In that sense the balance in the headline of the NSIDC press release is quite remarkable. Their press release itself mainly deals with the Arctic. Only at the end the Antarctic is mentioned:

As the Arctic was experiencing a record low minimum extent, the Antarctic sea ice was reaching record high levels, culminating in a Southern Hemisphere winter maximum extent of 19.44 million square kilometers (7.51 million square miles) on September 26. The September 2012 monthly average was also a record high, at 19.39 million square kilometers (7.49 million square miles) slightly higher than the previous record in 2006. Temperatures over Antarctica were near average this austral winter. Scientists largely attribute the increase in Antarctic sea ice extent to stronger circumpolar winds, which blow the sea ice outward, increasing extent.

NSIDC scientist Ted Scambos said, “Antarctica’s changes—in winter, in the sea ice—are due more to wind than to warmth, because the warming does not take much of the sea ice area above the freezing point during winter. Instead, the winds that blow around the continent, the “westerlies,” have gotten stronger in response to a stubbornly cold continent, and the warming ocean and land to the north.”

But I see this as a step forward. Back in 2007, the same end of the season press release didn’t mention the Antarctic at all.

The NSIDC press release attributes the increase in sea ice around Antarctic to stronger circumpolar winds. About the Arctic decline in sea ice they write:

Scientists attribute this trend in large part to warming temperatures caused by climate change.

So on the North pole we see the symptoms of a global changing climate, but at the Southpole there is a regional cause in play. Now it’s obviously true that there are large differences between the Arctic and the Antarctic sea ice. NSIDC summarizes the differences nicely:

Sea ice differs between the Arctic and Antarctic, primarily because of their different geography. The Arctic is a semi-enclosed ocean, almost completely surrounded by land. As a result, the sea ice that forms in the Arctic is not as mobile as sea ice in the Antarctic. Although sea ice moves around the Arctic basin, it tends to stay in the cold Arctic waters.

The Antarctic is almost a geographic opposite of the Arctic, because Antarctica is a land mass surrounded by an ocean. The open ocean allows the forming sea ice to move more freely, resulting in higher drift speeds. However, Antarctic sea ice forms ridges much less often than sea ice in the Arctic.

So it’s fair to say that the two regions are not easy to compare. Most climate scientists seem to agree that the Arctic sea ice is more vulnerable and therfore a better indicator for climate change. But we will see if this opinion will hold if times change and Antarctic sea ice starts decreasing.

Back to the question whether it is a coincidence that the record low in the Arctic is combined with a record high in the Antarctic. In the last decade there have been some studies about what is called the polar see-saw hypothesis, the idea that climates of the Northern and the Southern hemisphere are out of phase. There even is a wiki page although that one is rather short.

By far the most relevant and interesting paper I found is Twentieth century bipolar seesaw of the Arctic and Antarctic surface air temperatures by Petr Chylek, a wellknown scientist with a slightly skeptical profile.

Their abstract:

Understanding the phase relationship between climate changes in the Arctic and Antarctic regions is essential for our understanding of the dynamics of the Earth’s climate system. In this paper we show that the 20th century de-trended Arctic and Antarctic temperatures vary in anti-phase seesaw pattern – when the Arctic warms the Antarctica cools and visa versa. This is the first time that a bi-polar seesaw pattern has been identified in the 20th century Arctic and Antarctic temperature records. The Arctic (Antarctic) de-trended temperatures are highly correlated (anti-correlated) with the Atlantic Multi-decadal Oscillation (AMO) index suggesting the Atlantic Ocean as a possible link between the climate variability of the Arctic and Antarctic regions. Recent accelerated warming of the Arctic results from a positive reinforcement of the linear warming trend (due to an increasing concentration of greenhouse gases and other possible forcings) by the warming phase of the multidecadal climate variability (due to fluctuations of the Atlantic Ocean circulation).

Their paper suggests there is a strong anticorrelation between the Artic and the Antarctic with a peak of -0.89 when using a 17 year smoothing of the temperature data. Their figure 2a is quite interesting:

Figure 2. (a) De-trended Arctic (blue) and Antarctic (red) temperature time series smoothed by a 11 year running average (thin lines) or 17 year running average (thick lines), and (b) the AMO index [after Parker et al., 2007] annual values (thin line) and 17 year running average (thick line).












So the anticorrelation between the poles seems to have been there for some time. It’s not so easy to come up with a physical mechanism although the high correlation with the AMO seems to indicate an important role for the Atlantic Ocean. Chylek:

The high correlation of the polar de-trended temperature time series with the AMO index suggests that the variability of the Atlantic ocean circulation might serve as a link with the bi-polar temperature seesaw pattern. The observed seesaw pattern is consistent with the model of the interhemispheric ocean circulation that includes a strong upwelling along the Antarctic Circumpolar Current.

To conclude I think much more research should focus on this highly intriguing anticorrelation between the Arctic and the Antarctic. But this probably will only start when a shifting AMO really reverses the trend on both poles.



19 comments to Is the Arctic really so much more interesting than the Antarctic?

  • Elmar Veerman

    Note the word ‘de-trended’, which indicates that temperatures at both poles have increased! The graph you show here does not show the real T changes, something you fail to explain. Take a look at figure one of the paper. There Chylek et al. show the real T changes and construct a linear trend for both poles for the period 1900-2010. For the North Pole this is a rise of about 1 degree Celsius per century. For the South Pole it is about two degrees per century. The authors reason that greenhouse gases are the cause of this warming. Next, they subtract these trends from the graph and that gives them figure 2, the one you show. Which is certainly interesting. But it is a little bit strange that you forget to mention the T rise at both poles.

    And to get back to your question: why does the North Pole get more attention? I think that is mainly because it has a far bigger influence on the weather in the places where most people live. The missing ice at the NP likely caused the recent periods of extremely hot and dry weather in Russia and the US.

  • Guido

    I don’t think it is surprising that a google search for Arctic sea ice yields more hits than Antarctic; in the Arctic there is a clear downward trend over the past decades. Antarctic sea ice is at a maximum at the moment but without any trend. I had expected the difference in media coverage to be larger by the way. And of course you are totally right: bad news just attracts more attention.

    More importantly, I would second Elmar Veerman’s warning about looking at detrended data. Your end statement makes sense for the detrended data, but not if you look at the original time series. If the past holds for the future and you would stitch what happened in 1940-1970 to the end of the timeseries we have now, there will be no reversal of long-term trends. You end up with natural variability (indeed, very intriguing) on top of a trend that is not easy to explain if one dismisses the enhanced greenhouse effect.

  • Marcel Crok

    Hi Guido and Elmar,
    I think this is a pretty important statement in the article:
    “Furthermore, none of the AOGCMs used in the IPCC 2007 climate assessment report has been able to reproduce the early 20th century (1910–1940) Arctic warming followed by a sharp cooling period (1940–1970), although all models simulated the post 1970s warming trend [Gillett et al., 2008]. Consequently our understanding of Arctic climate, its internal variability, its drivers and responses is not yet complete.”
    As long as we don’t understand the early warming we have to be careful to be too certain of ourselves regarding both understanding the recent trend as well as what they call the underlying trend.
    But I agree of course it’s important to look not only at detrended data. My post became a bit long and this graph was the most relevant to show the anticorrelation.

  • Jos Hagelaars

    “Apparently the Arctic sea ice is more interesting than the Antarctic sea ice, especially for the media.”

    Well, maybe it has something to do with the amount of ice vanishing in the Arctic, ~500.000 km2 per decade, that draws more attention than the much smaller gain of ice in the Antarctic, ~100.000 km2 per decade:
    Or the difference in attention could also be related to the large difference in warming of the two areas on the globe, Giss data (as used in Chylek et al figure 1):
    Arctic 1970 – 2011 : +0.48 °C/decade
    Antarctic 1970 – 2011 : +0.08 °C/decade

    “Their figure 2a is quite interesting.”

    This de-trended graph is interesting when combined with figure 2b, which is omitted here. The Chylek et al paper speaks of a high correlation (and anti-correlation) of AMO with the temperatures in the polar regions. Their figure 2 raises some questions about this correlation not mentioned in the paper:
    – Why is there a peak around 1960 in the AMO index that does not show up in the temperatures?
    – Why does the temperature rise in the Arctic around 1970 lead the rise in the AMO index with several years?
    One could also wonder why Chylek et al did not provide any error bars in their figure 2a, taking into account the large uncertainty in the Antarctic temperatures before 1950 (see their figure 1).

    “I think this is a pretty important statement in the article.”

    This statement, in the introduction part of the paper, is partly based on Gillet et al 2008, which clearly states: “We find that the observed changes in Arctic and Antarctic temperatures are not consistent with internal climate variability or natural climate drivers alone, and are directly attributable to human influence.”

    Chylek et al 2010:
    Gillet et al 2008:

  • Pieter Zijlstra

    For some time now it is clear to me that the Arctic and Antarctic are completely different entities.
    The NSIDC webite is indeed a good start to illustrate and study the details.
    So I am getting sceptical if both areas are mentioned in one sentence.
    In this respect I regard title and start of this tread as an appetizer only.

    I would not discuss two records of different size (percentage of change) and kind (winter – summer)in one tread. That is to difficult (for me).

    A most interesting question is whether the thermodynamic equilibrium state (resulting in maximum thickness of the ice) can be achieved at the end of a winter. It seems that for the Arctic this state has not been reached for several decades now. Leading to the loss of multi-year ice.

    If a teleconnection is discussed between the two poles:
    – what about the time lag between start of the process and the endresult?
    – what could be the starting location
    – can it be a chain of process steps to transport energy from one location to the other?

  • Guido

    Marcel and other folks interested, please see these graphs I just made to put things a bit in perspective.

    Top graphs show AMO and temperature for both the Arctic (left) and Antarctic (right). Just like the paper discussed above it is all 11-year running mean from the same sources. Personally, I don’t think the Antarctic temperature for the first decades makes a lot of sense because it is based on one or two stations in Argentina; temperature records in Antarctica start in the 1940s or so.

    The left middle graph shows a scatter plot of the temperature anomalies. Correlation (R2) is 0.15 so not good but it is significant. Note that the relation is positive, so they are not in anti-correlation. That does change when you look at de-trended values as mentioned in the paper, I will return to that later.

    But since there is something the drives both the Arctic and Antarctic temperatures in the same way, let’s bring in the most obvious candidate: CO2 (middle right). If you use this record in combination with AMO values it is relatively easy to mimic the measurements, especially for the Arctic, see the two bottom graphs.

    Temperature anomaly Arctic = -4.2 + 1.8 * AMO + 0.014 * CO2 (R2 = 0.9)
    Temperature anomaly Antarctic = -7.8 – 0.8 * AMO + 0.022 * CO2 (R2 = 0.7)

    So indeed, they are inversely correlated with AMO if you subtract the trend (note the + sign for AMO in the Arctic and the – sign for the Antarctic). And transient (!) climate sensitivity for this simple linear experiment with no other forcings is 4 degrees per doubling of CO2 for the Arctic and 6 degrees for the Antarctic.

    A few conclusions from this simple experiment:
    – you are right in that the current rapid warming in the Arctic is partly natural (the same reasoning can explain the relatively constant global temperature for the past decade)
    – I don’t see how one can simulate the temperature records without a global forcing (like CO2)
    – surprisingly the climate sensitivity for the Antarctic seems higher (!)

  • Jos Hagelaars

    Hi Guido,

    Nice graphs! It is tempting to assume that the AMO is inducing some changes in atmospheric temperature. You should also consider it could just be the opposite. The AMO is based on sea surface temperatures and should be completely de-trended for the warming caused by greenhouse gases. One could question whether this can be done perfectly. E.g. see the references on:

    As I mentioned before, the Chylek et al paper shows a lag between AMO and the temperature. To me your graph in the upper left corner also shows that a change in temperature leads a change in the AMO. Your fit in the lower left corner shows a clear shift by a couple of years up to 1980. The AMO is something quite different than ENSO, which clearly leads ocean temperatures. See the discussion here:

    The oceans are warming due to the GHG warming, as can be concluded by the large increase of the ocean heat content since 1960/70. Some of this warming will probably be transported to the Arctic. As Mark Serreze puts it: “the Arctic amplification as presently understood has a suit of causes”. The change of the sea ice cover in the Arctic also alters the atmospheric heat flux. More information about this on page 90 of the Serreze review where the relation between AMO and the Arctic is also discussed:

  • Marcel Crok

    Hi Guido,
    Many thanks for preparing these graphs and the best fit regressions. I don’t know yet if this calculation
    of a transient climate sensitivity makes much sense.
    Both the global SST and the North Atlantic SST trend for the period 1900-2010 is around 0,053 per decade. So this indeed suggests there is a longterm trend that could be caused by a global forcing like CO2. Now let’s suppose this whole trend has been caused by CO2. Globally this would give a climate sensitivity of not much more than around 1 degree (assuming the relation is logarithmic so that the 40% rise in CO2 causes already 75% of the doubling effect). To get higher estimates we need the highly uncertain cooling of aerosols of course. Your calculation of 4 and 6 degrees of warming in the Arctic and Antarctic would suggest strong polar amplification, stronger than has been observed in the past.

    According to the latest book of blogger Bob Tisdale (Who turned on the heat?) which I am currently reading, the recent warming of the North Atlantic can be explained by AMO + ENSO. What happens if you add ENSO to the formula? A central point in his book is that CO2 is not a good explantion for the recent warming of the global oceans. The East Pacific (1/3 of the global oceans) has not warmed at all in this period. It’s difficult to explain why a global forcing like CO2 would not warm up such a large area of the global oceans.
    His main hypothesis is that the warming that has taken place is the result of ENSO. He shows that in this period (last 30 years) La Nina’s do not fully compensate for the effect of El Nino’s with warming of some oceans as a result.

  • Jos Hagelaars


    “The East Pacific (1/3 of the global oceans) has not warmed at all in this period.”

    This is really strange.
    The heat content of the total Pacific Ocean has increased since 1957 with about 0.13×10^22 J/year (to 2000 m). The heat content of the northern part of the Pacific Ocean as well as the southern part of the Pacific Ocean has increased since 1957 as has the heat content of the all the oceans of the world (0.42×10^22 J/year).
    OHC pentadal data:
    Besides that the slope of the MEI index from 1980-2010 is slightly negative.

  • Guido

    Marcel, Jos,

    Guess it comes down how one defines “recent warming”. Please have a look at these graphs with trends in temperatures over the ocean starting in 1900, 1950, and 1980. So if “recent warming” is 1980 then I see the point, but I also see a clear PDO signal. Anyway, even if you take 1980 as the start of your time series then 60% of the Pacific is warming, and the whole basis warms on average with 0.08 degree per decade. If you consider all ocean surface it is 75% warming with a rate of 0.11 degree per decade. Longer timeseries translate to a larger percentage of the oceans warming. You will always find somewhere a place that is not warming, and in the case of the Pacific basin since 1980 this can be explained.

    But let’s first see if we can agree on some other things. Marcel, you mentioned that with a 40% increase in CO2 we are already at 75% of the radiative forcing (RF) for a doubling of CO2. I think everybody agrees that the CO2-RF relation is not linear, but it is actually not far off. Following RF = 5.35 * (CO2 / CO2 pre-industrial) we are now with 390 ppm at 1.8 W/m2 while a doubling is 3.7 W/m2. In other words, we are “only” at 48% or so. Now, if you add other GHG and ozone etc we may be at 75% or so but then you also have to subtract aerosols, for which the direct effect is negative for sure.

    I am not very sure about the first half of the 20th century, but let’s repeat your calculation for the 1950 onwards period. Warming of the oceans is 0.5 degree since then. RF from -say- 310ppm to 390ppm is 1.8 – 0.5 = 1.3 W/m2. To calculate transient climate sensitivity: 0.5 / (1.3 / 3.7) = 1.4 degrees per doubling CO2. For the oceans, land of course is quite a bit higher. Again, taking other GHG into account would lower climate sensitivity, aerosols would increase it. If you start at 1940 you get exactly the same value.

    With regard to ENSO explaining the warming in the arctic, I don’t mind doing that but I already know the answer, because temperatures in the Arctic don’t really show a good correlation with ENSO. Will have a quick look later.

    Finally, I don’t trust the calculation I made for the South Pole a lot, there are just too many other things going on including ozone etc. But for the North Pole it would be foolish to dismiss the R2 of 0.9 based on two parameters that make physical sense and I doubt the the climate sensitivity will be more than 50% off. There is simply no way one can reconstruct the measurements with a low climate sensitivity.

    @Jos – I don’t know enough about the ocean to understand what the physics are wrt leading and lagging, but intuitively I would be surprised that a few years of anomalous temperatures in the Arctic drive the thermohaline circulation. But you are right, it is interesting that temperatures lead AMO.

    By the way, Marcel, you question whether these calculations make a lot of sense. I agree, but at this point these are exercises everyone can do for themselves (it helps if you can program but with Excel one can do quite a bit) and they are totally transparent. I’d be happy to share the code with anyone.

    For the Arctic case as well as for the global average I truly believe these kind of calculations are as useful as computer models. The take home message for me is that there is a lot of natural variability, which impacts some regions more than others, and operates on various time scales. But one needs a globally operating forcing to explain all observations….

  • Guido

    ps: RF = 5.35 * (CO2 / CO2 pre-industrial) should of course be RF = 5.35 * ln(CO2 / CO2 pre-industrial)

  • Guido

    With regard to AMO and ENSO (expressed as SOI because that is the only long-term dataset I could find, note that negative values are El Nino’s): here is the graph

    T vs AMO and CO2: R2 = 0.9
    T vs AMO and CO2 and SOI: R2 = 0.9
    T vs AMO and SOI: R2 = 0.6

    As before all based on 11-year averages.

  • Is the Arctic really so much more interesting than the Antarctic?

    Yes, much more interesting. Here’s why.

  • And yet another record has been shattered in the Arctic.

    Record dominoes 12: CT SIA anomaly

    Antarctic sea ice area has been dropping almost as fast as Felix Baumgartner, so a record CT global SIA anomaly is a possibility as well.

  • Guido

    Neven, volgens mij moeten we daar nog even geduld voor hebben. Maar tjeemig, deze grafiek is toch wel -hoe zal ik het zeggen- veelzeggend! bron

  • Marcel Crok

    Niet te snel Guido, we hebben hier met een (omgekeerde) hockeystick te maken en Lonnie Thompson (de slechtste data archiveerder aller tijden zit in het team, zie bv Zullen jij en ik samen eens naar de data achter Kinnard et al gaan kijken op een druilerige herfstavond?
    gr Marcel

  • Marcel Crok

    McIntyre heeft vorig jaar al een blik op Kinnard geworpen. Met de archivering zit het in zekere zin wel goed (data en source code zijn gearchiveerd), maar uiteindelijk ook weer niet zo goed, wat dan weer niet alleen de schuld is van Kinnard, lees bv:

    Een tweede post van McIntyre over de gebruikte O18-reeksen is echter nog interessanter. De bottom line is deze grafiek:

    Een simpele middeling van de gebruikte O18-reeksen laten een geleidelijke afname van de temperatuur zien door de eeuwen heen. De hockeystick van Kinnard kan hier dus niet door veroorzaakt zijn. Meer blog posts kan ik niet vinden.

    gr Marcel

  • De aangepaste grafiek die in beeld brengt hoe ver het minimum van 2012 afwijkt van de voorgaande 1450 jaar klopt om een aantal redenen niet, zie tweede comment onder eerdergelinkt artikel over de potentiële gevolgen van smeltend zee-ijs.

    In hoeverre de data van Kinnard et al kloppen, weet ik niet, maar ik denk zomaar dat als in de afgelopen 2000 jaar de Northwest Passage en Northern Sea Route allebei vijf op de zes jaar achter elkaar open waren geweest, dat we dat wel ergens in de geschiedenisboeken of proxies teruggevonden hadden.

  • Guido

    Marcel – point well taken 🙂 die was inderdaad wat snel, hoewel Neven misschien wel een punt heeft maar dat is moeilijk hard te maken lijkt me. Gaan we op een druilerige avond mee aan de slag, goed plan, maar dan kijken we ook even naar hoe het nou met die HADCRUT4 data zit want die trendanalyses kloppen voor geen meter, zal het even posten.

Geef een reactie

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>






Donate to support investigative journalism on global warming

My blog list