Sea Ice Update

This year witnessed a September minimum of Arctic sea ice which was only the 2nd-lowest on record. But the year’s minimum isn’t the surprising thing about this year’s sea ice. That would be the surprising lows observed during May and part of June, and now, it seems, during the most recent few days of October. Here’s the data, with 2016 in red:


Because the value dipped so often to it’s all-time low for time-of-year, the recent annual average is at its all-time low. Here are annual averages offset in time so that the most recent year is complete:


Meanwhile, at the other end of the Earth this year was middle-of-the-road, although it’s on the low end at the present moment:


The latest annual average is also on the low end, but not exceptionally so:


More important is the trend, which appears to be upward at the south pole but more strongly downward at the north pole:


Trend-wise, the southern hemisphere has gained about 2/3 of a million km^2 while the northern hemisphere has lost nearly 2 million km^2.

This blog is made possible by readers like you; join others by donating at Peaseblossom’s Closet.


20 responses to “Sea Ice Update

  1. Is the fourth chart mislabeled? Isn’t that Antarctic?

    [Response: Yes, it is. I’ll re-do the graph some time tomorrow.]

  2. David B. Benson

    The circumAntarctica sea ice trend is driven by the circumAntarctica wind trend, via the coriolis effect. A recent paper indicates that the circumAntarctica wind trend is in turn caused by changes in the Humboldt Current, sometimes called the Peru Current. This is an up welling of deep water, assisted again by the coriolis effect.

  3. I know the math gets more than a bit complicated, which is why I haven’t done it myself yet, but I have for years wanted to see a time series of annual reflected solar radiation based on difference between ice and ocean albedo, taking into account variation in radiation by date and latitude.

    This would be a more direct measure of the change in energy balance and put more weight on the summer months.

    [Response: You might find what you’re looking for here, as well as here.]

    • I’ve also thought of integrating surface albedo * insolation, but I got stuck at wondering whether it was actually a useful figure. When ice melts, two other major effects occur (that I’m aware of; are there others? likely): the ocean water can lose heat much more quickly via evaporation, and more clouds can form (precisely from that evaporation).

      Losing heat faster means the ocean doesn’t heat up as much as you’d expect just from the albedo change — but the air heats up more. Gaining clouds has a complicated effect: atmospheric albedo goes up, but so does heat retention in the lower atmosphere, and also the light become diffuse when it passes through cloud so the effective albedo of the surface goes down.

      It’s pretty easy to calculate numbers (if someone out there does the hard work of making the measurements), much harder to figure out what numbers make sense to calculate.

      • For 1970–2008 in the NH we have Flanner et al:

        I’d say it’s difficult to get super precise. You could try with reanalyses, but at least ECMWF ERA-I doesn’t do sublimation from ice properly. I don’t know about the others.

        We have satellite output from CERES-EBAF and Cloudsat-CALIPSO heating rates (product 2B-FLXHR-LIDAR). Neither of these have multiple-decade records and there are known problems, such as how Cloudsat fell out of the A-train constellation in 2011 and when it came back, it didn’t work at night any more.

        CERES is ultimately baselined using ocean-heat-content measurements, but there’s sparser coverage in the southern ocean and sea ice causes problems. I think we get a good enough rough answer from Tamino’s approach or the Flanner-like calculations.

        If anyone knows of more recent work on this I’d like to see it.

  4. methane madness

    Something to look forward to is the sudden collapse of the Filchner-Rhonne and Ross ice shelves of Antarctica. The resultant release of land based ice into the sea will lead to some spectacular Isostatic rebound, considering the land is depressed up to a kilometer my favourite outcome is a super volcano along the transantarctic mountain range, rivalling the Siberian traps event.

  5. Reblogged this on Don't look now and commented:
    More statistics that only fools would deny.

  6. It has been warm, it is warm and it will be warm in the Arctic, and that is likely a cause of the low sea ice extent. is a good place to follow global and regional anomalies
    For the Arctic start here:
    Just click on the map to step forward through forecasts and hindcasts.
    The forecast for the Arctic Ocean over the next seven days, is 4-5 C above normal.

    I don’t know if it is a rule, but Arctic sea ice lows seem to come 1-2 years after global temperature peaks. So what can we expect in 2017 or 2018?

    • Thanks Olof for this reference to Karsten’s work.

      An interesting detail: after having obtained from Roy Spencer a hint on a readme file concerning UAH’s 2.5° gridded data for the lower troposphere
      I managed to process it during the 2 last weeks. It’s amazing.

      As you certainly know, Nick@moyhu has a wonderful tool to view such data:
      allowing me for a spheric view on a grid within which each cell contains the OLS trend computed for its anomalies from dec 1978 to aug 2016.

      The highest trend was 4.68 °C / century; apart from a few cells near the South Pole and in Eastern Siberia (Kamchatka), nearly all 100 highest trends, all above 4.0 °C/cy, were found at latitude stripe 80N-82.5N (there is no data for 82.5N-90N, nor for 82.5S-90S).

      • When I write “The highest trend was 4.68 °C / century” I mean of course 4.68 °C above UAH’s baseline, i.e. the mean of the temperatures measured between jan 1981 and dec 2010.

      • Booh what a nonsense here! It was late yesterday evening (0:45 am here), too late to write meaningful things.

        A trend hardly could have to do with whatever baseline. Modifying the latter doesn’t change the former!

  7. The annual anomaly averages are less interesting than the minimums because we’re not expecting the winter ice cover to change by very much, in comparison to the summer cover. The Arctic Ocean gets no insolation for several months of the year: nobody’s expecting much of it to remain ice free in the winters. Seasonal changes (like this year’s May and October anomalies which you refer to) are more interesting, as are seasonal anomaly averages.

    • Actually, Nick, there is some discussion of a perennially ice-free Arctic Ocean:

      Recently, a number of studies found that it is indeed probable that no such instability exists for the loss of Arctic summer sea ice, whereas an instability might well exist for the transition from a seasonally ice-covered Arctic Ocean to an Arctic Ocean that is virtually free of sea ice throughout the entire year (1, 7, 9).

      That’s from 2009; not sure what the current state of the art is. But the paper I quote is here:

  8. skeptictmac57

    I often see deniers pop up on threads about declining arctic sea ice, claiming that the gains in sea ice in the antarctic make it all even and just part of some natural cycle (the mechanism of which they cannot explain).
    I like to remind them of this:
    “Land Ice
    Data from NASA’s GRACE satellites show that the land ice sheets in both Antarctica and Greenland are losing mass. The continent of Antarctica has been losing about 134 gigatonnes of ice per year since 2002, while the Greenland ice sheet has been losing an estimated 287 gigatonnes per year. (Source: GRACE satellite data)”

  9. Tamino and others here who have thought about sea ice prediction might find interesting a statistical analysis of prediction skill that Julienne Stroeve and I recently published in Polar Geography:
    “400 predictions: The SEARCH Sea Ice Outlook 2008-2015”

    There are some quantile and other regressions involved but the elevator message goes like this:

    “Results highlight a pattern of easy and difficult years, corresponding roughly to the distinction between climate and weather…. Thinning ice that is sensitive to summer weather, complicating prediction, reflects our transitional era between a past Arctic cool enough to retain much thick, resistant multiyear ice; and a warmed future Arctic where little ice remains at summer’s end.”

    Since the paper came out we’ve updated the analysis to include 2016, and have something in draft about that.

  10. More interesting than the Antarctic m² would be the m³ of ice there, which I understand have been declining quite significantly.

    • I’m not sure of a significant decline of Antarctic sea ice volume.

      What until now I managed to read about is rather that the volume increase there is less significant than the volume decrease in the Arctic (e.g. below, unfortunately behind paywall):

      • Ah. Thanks. No idea where I got that wrong idea from.

        The interesting part is that current climate models don’t yet capture that. Yet another straw-man “argument” for the deniers, I guess.

        NB: There are no paywalls in the age of sci-hub.

      • Thinking of land ice maybe?

      • Matthias Urlichs | October 25, 2016 at 1:47 pm
        NB: There are no paywalls in the age of sci-hub.

        “Error 500 from scihub22266oqcxt.onion.”
        What a terribly useful proposal… there is, as you certainly know, a good german expression perfectly fitting here: “Oberlehrer”.