RealClimate

Climate science from climate scientists...

You are here: Home / Climate Science / Arctic and Antarctic / Arctic sea ice minimum discussions

Arctic sea ice minimum discussions

by

Here is a continuation of the last Arctic sea ice discussion as we get closer to the 2011 minimum. All figures will update continuously.

JAXA Sea ice extent and area:




Cryosphere Today sea ice concentration:



Estimated sea ice volume from UW PIOMAS (updated every month):



Reader Interactions

143 Responses to "Arctic sea ice minimum discussions"

  1. Hope the next IPCC report has a better section on the retreat of sea ice. It has already retreated faster than all models under all scenarios; obvious bias in the models. Also, the IPCC chart comparing models was a display of the average extent during the months of July, August and September. The implication being that there can’t be any sea ice during those months before the arctic can be said to be seasonally ice free arctic However, sea ice reaches it minimum in September with August and October the next lowest months.

    Looking at trends in sea ice volume, it’s apparent that the arctic will reach a minimum of zero within the next decade.

  2. I take back some of what I said last month. I really like what Didactylos and Sphaerica (Bob) did with a monochromatic graph keyed on year, last month that started here.

    https://www.realclimate.org/index.php/archives/2011/07/arctic-sea-ice-discussions/comment-page-3/#comment-211290

    My favorite color would be light-blue = more distant past and dark-blue = more recent, and maybe on offset color for the current year. I can’t make a pattern out of the color schemes used above.

    I used to hope that people would wake up when they saw the trend in ice melt (How can they reconcile a huge decline in ice with no real warming trend? Yeah, I know, not all people in denial say there is no warming, but that zombie keeps walking around.), but that hasn’t been the case.

  3. The 2011 graphs so far look to be tracking almost as low as the 2007 minimum for area and extent and the downward trend in volume is still clear. Another good picture at PIOMAS shows the daily volume trend in a format more comparable to the JAXA graphs. If the volume trend persists, it will take only one slightly warmer than average summer to break all records for minimum extent and area. With the current solar cycle still looking like peaking on the low side, this is not too surprising but we can’t rely on that to save us long term.

    #2 Chris G: we may only really “wake up and see the ice melt” at the peak of the next solar cycle if this one tops out lower than average as expected.

  4. Jacobson: Soot’s the thing.

    “Soot emissions account for about 17 percent of global warming,….”

    “Jacobson says his calculations show controlling soot could reduce warming above parts of the Arctic Circle by almost 3 degrees Fahrenheit within 15 years, virtually erasing all of the warming that has occurred in the region during the last 100 years, a society release reported Wednesday.”

    That does not compute.

  6. http://www.msnbc.msn.com/id/44353322/ns/us_news-environment/

    New photographs taken of a vast glacier in northern Greenland have revealed the astonishing rate of its breakup, with one scientist saying he was rendered “speechless.”

    In August 2010, part of the Petermann Glacier about four times the size of Manhattan island broke off , prompting a hearing in Congress.

    Researcher Alun Hubbard, of the Centre for Glaciology at Aberystwyth University, U.K., told msnbc.com by phone that another section, about twice the size of Manhattan, appeared close to breaking off.

    But when they returned in July this year, they found the ice had been melting so quickly — at an unexpected 16-and-a-half feet in two years — that some of the masts stuck into the glacier were no longer in position.

  8. Can I ask for some help understanding the set of graphs? The first two appear to show not much in the way of trend at the peak ice but upwards of a 1/3 (!) loss in ice extent at the minima – yet 2007 may be a bit worse than 2011. However the trend line appears to show 2011 being significantly worse than all previous years including 2007? What am I missing?

    [Response: Different metrics. Arctic sea ice is not easily condensed to a single number, and each of the graphs show different aspects of the situation. The resolution of the apparent contradiction is that the ice is roughly equally spread out this year compared to 2007, but overall is thinner. I am not aware of any analysis that says that one metric is more indicative or predictive than another in any general sense. It’s just part of the complexity. – gavin]

  9. I have reservations about the volume numbers and ice thickness models. The piomas version 2 model that was introduced this spring shows considerably higher thickness values. The navy, for example (http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictnnowcast.gif) reports two to three meter thick ice all around the north pole. Russian and American science vessels recently met at the pole and reported actual thickness of a meter or less.

    Given that we currently have ~3 million square km of ice it would have to average more than 2 meters thick to reach the current piomas volume of a bit over 6 thousand cubic km. Given the thin and fragmented ice the Healy went through on the way to the pole I personally find that average doubtful.

    Given the disagreement between volume models and real-world measurements – admittedly limited spatially – I’m left wondering what the actual volume is. I’m looking forward to cryosat results once they’re done with the calibration phase.

    Note that there’s no good reading on arctic ice. If the models are all right wrt volume the ice is declining alarmingly. If the real-world measurements are closer and volume has already declined to 4,000 cubic km or less then we’re in even more trouble.

  10. #8–To elaborate a bit on Gavin’s response, the top graph is “extent,” which means the total area of map gridboxes containing more than 15% ice.

    The next is “area,” which attempts to quantify the area within those grid boxes which is actually ice-covered.
    Area is thus always going to be smaller than extent-well, for any realistic case, anyway. And the two track fairly well in general, though ice can spread out, which can increase extent but not area, or the reverse.

    For these measures, variability does change throughout the year; as you noted, the biggest change is in the warm part of the year (though the biggest change in Arctic temperature is the trend in the winter.)

    The third graph is volume, but the difference is greater than that, because it’s also graphing anomalies, not just the values themselves. It’s a useful way to display things in that it eliminates the annual cycle, which is so evident in the other graphs. You can also graph area and extent that way (and I know just where to find a couple of graphs like that–when I’ve posted this comment, I’ll go and fetch the links.)

    You can read more about these matters, and many other related ones, too, here:

    http://nsidc.org/arcticseaicenews/
    http://nsidc.org/arcticseaicenews/

  11. Sorry for the double link in the previous. Here’s the promised links:

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_N_min_to_date.png

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_UBN_min_to_date.png

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_VOL_min_to_date.png

    These bar graphs are all by L. Hamilton; be aware that they don’t all come from the same data source, so processing algorithms and therefor results may differ a little, apart from the fact that they are showing different metrics.

    They (and much, much else) can be found here:

    https://sites.google.com/site/arcticseaicegraphs/

  12. David Miller says:
    1 Sep 2011 at 12:39 PM

    Russian and American science vessels recently met at the pole and reported actual thickness of a meter or less.

    Can you direct us to a pointer on that? Over on Neven’s Arctic ice blog there’s speculation that ships visited the N Pole this year and didn’t gather that information. Sounds most unlikely that researchers on these ships would be so myopic.

  13. The Navy site says “system and web page are a demonstration and are not an operational product. NRL is providing the INFORMATION on an “as is” basis. NRL does not warrant or represent this INFORMATION is fit for any particular purpose ….”

    — I wonder if the report there showing ice thickness not as an average but as a navigation aid (as in, low hanging ice, don’t bang your submarine ….)

    Pure speculation. Perhaps someone knows.

  14. The three bar graphs (annual 1-day minimum CT area, UB extent & PIOMAS volume 1972 or 1979-present) that Kevin McKinney linked above are being updated frequently while the melt season continues.

    They’re intentionally simple, losing that “complexity” Gavin alludes to, but pretty easy for anyone to follow.

    I should have a new graphic within a day or two at Neven’s blog, comparing the five main area & extent time series.

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_N_min_to_date.png

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_UBN_min_to_date.png

    http://img.photobucket.com/albums/v224/Chiloe/Climate/sea_ice_VOL_min_to_date.png

  16. The latest data on NSIDC suggest that this year’s extent coud break the 2007 all time low. In any case, it will be very close. The Petermann glacier just lost a huge chunk of ice, according to blurb on Yahoo I just noticed. The comment thread was a pathetic display of ignorance, name calling and libertarian type ranting. Not very encouraging.

  17. Thanks for bringing this forward; i was just checking to look for it after reading about the Greenland ice sheet depletion.

  18. I do have one friend who is very much on the denialist bandwagon, and is a devoted fan of wattsupwiththat.com. I mentioned this issue about the thinning Arctic ice to him, and his reply was that the melting is being caused by “thousands and thousands of underwater volcanoes.” I have to admit, I hadn’t expected that reply – I thought he would just deny that any ice thinning was taking place. Intrigued, I pressed a little further, and asked if he thought there was an increase in the number of underwater volcanoes, and if so, why that might be. Again, he provided a sincere but entertaining response, saying that yes indeed, volcanoes were becoming increasingly active all over the world (he offered up the recent eruptions in Iceland as proof). The reason, he says, is that the planets are all lining up, a process that will come to a head in 2012, and cause the Yellowstone Super-volcano to erupt with disastrous results. He suggested I rent the video “2012”, a Hollywood blockbuster which until now I’ve managed to avoid watching:

    2012
    http://www.imdb.com/title/tt1190080/

    Well, there you go. We have a full explanation for the thinning Arctic ice, and it has absolutely nothing to do with CO2 emissions. And here I was needlessly worrying about AGW. Silly me.

  20. If the ice were melting as a result of underwater volcanoes, then we would expect the oceans to be warming from the bottom up. But actual (non-Hollywood) data shows that the oceans are warming from the top down, i.e., the heat is coming from above, not from below.

    Don’t expect evidence to change your friend’s mind, however.

  22. Candide – you friends needs medication – quickly. That is seriously detached from reality.

    [Response: I don’t know. Is the idea of planetary alignments causes mass climate change in 2010 really so different than the idea that keeps popping up (e.g. J. Curry) that the increase in CO2 in the atmosphere might not actually be related to human activities? You start medicating the astrology buffs and you’ll start medicating everyone. –eric]

  23. #9 David Miller says …

    NAVO’s PIPS 3.0, aka ACNFS, over 15 years in the making, is still a POS IMHO.

    The end user is NIC, and the best use of NAVO’s model is for ice edge detection, it’s meant for operational use only, meaning navigation in open waters.

    Read their own calibration report here;

    http://www7320.nrlssc.navy.mil/hycomARC/prologue.html

    http://www7320.nrlssc.navy.mil/pubs/2010/posey1-2010.pdf

    It overpredicts ice thickness by over a meter when compared to USACE CRREL (I use to work at CRREL way back in 1975) buoy data.

    It overpredicts fly over data by about 0.4 meters (Note: Not much of this type of data to compare against, if you were to ask me).

    In total it isn’t much of an improvement over PIPS 2.0, sorry but, Navy 0 Army 1.

    Have you been to the PIPS 2.0 site lately?

    It’s offline, after twice predicting a POLE HOLE where none existed, and I saved all those daily GIFS (from their 2nd failed attempt), to make a nice animated GIF of that absurd model.

    Further, I’m building a collection MODIS imagery where PIPS 3.0 shows 5+ meter thick ice, but there are actually open leads and much scattered/broken up sea ice, 5+ meters my AZZ!

    NOTE: You can stop with the Goddard type talk, now there’s a real ******** if ever there was one.

  25. I think what JohnN is missing is that the bottom graph is for volume anomalies which takes the thickness of the ice into account. An area covered with 3 meters of ice has 3 times the volume of the same area covered with 1 meter of ice. There has been a significant loss of thicker multi-year ice over the past few years so the volume has dropped more than extent or area.

  26. 18 Candide: http://www.imdb.com/title/tt1190080/ says: “the crust of the earth is becoming unstable” like on Venus.
    Follow the links and the continents re-arrange themselves in a few hours. All because the planets are lining up.

    Yet your friend can’t believe CO2 is indirectly melting the Arctic ice? I need one of Joe Romm’s head clamps!

  27. Comparing apples to apples (I think), the JAXA ASMR-E graph shows sea ice area as somewhat under 4 million km^2, but the cryosphere today graph shows sea ice area at 3.1 million km^2.

    Is there some way to reconcile those? Do they purport to measure the same thing?

  28. Just been looking at Prof Kevin Trenbeth latest findings..ie. for every 1 deg F. of ocean warming there is a 4% increase in water vapour in the atmosphere resulting in a 6-8% increase in rainfall. The spurious 6mm decrease in ocean height was the result of that riculously intense la-nina last year causing a hell of lot of rain to fall on land that is still perculating and meandering it’s way back to the oceans. Another study involving a tightly spaced convoy of 5 satellites 2 sec apart collecting data on cloud formation and extent, reports an over 50% decrease from the norm in arctic cloud cover. This is allowing even more sunlight to reach the remaining sea ice and cause accelerated warming not just from above but also below due the warming arctic ocean this is probably the ice area and extent graphs are so woeful this year and as I thought this year will most likely be the worst on record.
    One question I have that someone might know the answer to is….. due to 6-8% increase in rainfall over land could the change in river temp of all the thousands of swollen river waters emptying into the sea have any near future impact on the ocean’s temp. Logic says yes but I would like to have a rough idea as to what extent, whether it is negligible or yet another pos’ feedback system?

  29. Lawrence Coleman @ 28, do you references for those statements?

  30. [Response: I don’t know. Is the idea of planetary alignments causes mass climate change in 2010 really so different than the idea that keeps popping up (e.g. J. Curry) that the increase in CO2 in the atmosphere might not actually be related to human activities? You start medicating the astrology buffs and you’ll start medicating everyone. –eric]

    Well, the friend only needs to be medicated for a year or so, to contain his anxiety about imminent death until such time as the fear can be, er, superannuated.

  31. “Comparing apples to apples (I think), the JAXA ASMR-E graph shows sea ice area as somewhat under 4 million km^2, but the cryosphere today graph shows sea ice area at 3.1 million km^2.

    Is there some way to reconcile those? Do they purport to measure the same thing?”

    Yes, this is indeed apples to oranges, and no, they don’t purport to measure the same thing. See my comment #10 above.

    Extent is probably becoming obsolete as a measure, in that it makes less sense as a metric the less cohesive (or the more minutely fragmented) the pack becomes. (Just my opinion.) But area is harder to measure, and of course backward-compatibility in data is highly desirable.

  32. #3 “If the volume trend persists, it will take only one slightly warmer than average summer to break all records for minimum extent and area.” by Philip Machanik.

    Not exactly. A ‘cool’ or ‘warm’ summer doesn’t have much meaning in that region. Summer 2007 was actually on average somewhat cooler than normal, even during the flash melts in that season: http://ocean.dmi.dk/arctic/meant80n.uk.php . This was also during solar dip and onset of a very strong La Niña.

    Sunshine and wind are more important factors. And most of the melting appears to happen from beneath, what with ever warmer waters lapping around and underneath the pack.

    I do agree with the gist of the message. The pack could disappear any summer now. Volume decrease is accelerating, catastrophe theory needs to be incorporated in modelling the phenomenon ( http://en.wikipedia.org/wiki/Catastrophe_theory ).

    [Response: Worth linking to this comment. To expand on that point though, the DMI plot shows a current weather forecast analysis compared to a re-analysis product using a different model. There is no guarantee (or even much expectation) that they are directly comparable in regions that are relatively poorly sampled (such as the Arctic). Thus trends derived from comparing the two (without making any allowances for differing biases) are not going to be very reliable. – gavin]

  33. The NSIDC have an graph of total volume of ice and it down on 2007 as well.

    http://nsidc.org/images/arcticseaicenews/20110816_Figure5.png

    Joseph Romm often has articles stating that Arctic Sea Ice is in a death spiral.

    http://thinkprogress.org/romm/2011/07/08/262576/arctic-death-spiral-sea-ice-volume/

    The very last graph seem somewhat interesting. I believe that Richard Alley has stated that the models are running 100 years behind schedule and that warming in the Arctic is 4x faster than predicted.

  35. KAP (@20), the fact that the ocean would need to be warming from the bottom up is not a problem for the delusional, since they don’t even consider the water column between all these ‘new’ volcanoes and the ice. They simply assume that the warmth jumps straight from the volcanos directly to the ice.

    I’m not making this up: one of the more numerate denizens at WTFUWT actually calculated a ball park heat value emitted by these phantom volcanoes and declared that it would in fact be just enough to account for the 2007 melt…., except that he did not include the mass of the water column in his calculations.

  36. #31 — How is “sea ice area” for one apples and for the other oranges? What are they doing differently to measure ice area?

  38. JimCA, where are getting confused by trying to compare?
    Cite/point to the site definition; they do differ.

    JAXA: http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm

    “Definition of sea-ice cover (extent and area)
    The area of sea-ice cover is often defined in two ways, i.e., sea-ice “extent” and sea-ice “area.” These multiple definitions of sea-ice cover may sometimes confuse data users…..”

    Wattsup explained it at his blog back in 2008; I don’t recall how well.

  39. #36–Sorry, JimCA, I misread your original comment somehow. I was thinking one was for ‘extent.’

    However, you might imagine different area products as being, say Fuji apples versus Red Delicious (or something like that.) Significant differences still exist, even if they are all ‘apples.’

    One is the sensor used–IJIS uses the instrument on the Aqua satellite, which is why their data only goes back to 2002, while NSIDC uses data from a series of older instruments, suitably spliced together.

    Some products choose different resolutions, looking at 6.25 km2 grid boxes versus 12.5 k, and so on. This, too, can make a difference to the area result.

    Finally, the statistical presentation can make a difference. For example, the IJIS extent number is actually (I am told) a two-day moving average, whereas NSIDC extent presents a 5-day average. (This is all IIRC; I haven’t checked my memory on this.) So you’d expect NSIDC extent to be higher at this time of year, since it would be more affected by values ‘less far along’ in the melting process. (Haven’t checked if that’s the case just now, and of course other differences play into it, too.)

    If you want to specifically reconcile IJIS area and CT area, you’d want to search the ‘metadata’ that applies to each and see if you can infer what’s causing the difference you observe.

    It’s confusing, I know, and can be misrepresented by those naive or obfuscating folks who expect (or claim to expect) that these measurements should be an exact ‘truth.’ They are, as Gavin said, ‘metrics’–measurements–that are intended to give information about the ice. They are not the ice itself!

    One good thing about having all of these different metrics is the fact that, despite the differences of detail, they all show basically the same big picture. That gives us some additional reassurance that said “big picture” is not somehow the artifact of a particular way of measuring things.

  40. #38, Hank Roberts – Take a look at the RC article. There is a graph showing JAXA sea ice extent AND a different one showing sea ice area.

    JimCA is asking what the difference is between the JAXA data labelled ‘Sea Ice Area’ and the Cryosphere Today ‘Sea Ice Area’.

    ReCAPTCHA: Response rticatic

  42. Thanks for all the responses, everyone. I tried to be clear what I was asking, but should have been more explicit to avoid confusion.

    I think I see now how the difference might emerge, but then have to wonder if these folks talk to each other. Assuming there is a ground (so to speak) truth, it should be possible to perform spot checks on small areas to calibrate both, which presumably would reduce or even eliminate the discrepancy. But maybe that is easier said than done.

  43. Everyone, thank you for the responses. (And damn you captcha for eating the previous version of this message.) I tried to be clear in my question, but should have been more explicit to avoid confusion.

    I think I see how the discrepancy might arise, but then must wonder if those people talk to each other. It would seem to be straightforward to find small areas for which a ground (so to speak) truth can be found, then calibrate both series to that, in which case the discrepancy should mainly vanish. But maybe that is easier said than done.

  44. Any bets on another harsh winter for the CONUS this year? I’m betting on it happening again, caused by the lack of ice coverage in the Arctic.

  45. JimCA, they’re working on it. Examples:
    http://scholar.google.com/scholar?hl=en&q=arctic+sea+ice+satellite+area+estimation+pixel

    Your input could help:

    http://www.aip.org/fyi/2011/040.html

    “NASA Administrator Charles Bolden …. described problems confronting some earth satellite replacement programs, and starkly warned the committee “we are in dire straits as a nation when it comes to weather and climate prediction.” He was blunt in calling, as “dumb things” congressional attempts to defund a satellite program that would measure, among other data, shifting changes in the world’s climate. “I don’t do global warming, I do earth science,” he said emphatically.”

  47. #41–“It would seem to be straightforward to find small areas for which a ground (so to speak) truth can be found, then calibrate both series to that, in which case the discrepancy should mainly vanish. But maybe that is easier said than done.”

    I think it would be a moving target, since the differences are effective at different times, or in different situations. So the discrepancies wouldn’t stay vanished for long (although they might conceivably “phase in and out.”) Plus the data sets would lose that backward compatibility I mentioned. (The jargon is that they would become internally “inhomogenous.”)

    #42–I wouldn’t bet on it myself–I think there’s more to NH winter weather conditions than just the ice cover–but neither would I be terribly surprised if it came about as you expect. There is, after all, some support for it in the literature.

  48. #42 Rich Hendricks,

    I don’t know this CONUS of which you speak. ;)

    But if you’re talking about the very cold winter periods we’ve been having in the Northern Hemisphere recently: I suggest you google for ‘Judah Cohen Siberian Snowfall’, link.

    I’m researching this at present for a post on my blog, should post in the next week or so. Cohen makes a fairly convincing case that anomalous Siberian snowcover has a pivotal role.

  50. @ Candide — 1 Sep 2011 @ 8:38 PM Re sea ice melt and volcanoes

    see https://www.realclimate.org/?comments_popup=576#comment-90991 and https://www.realclimate.org/?comments_popup=576#comment-91077

    About the time I was posting that comment, I came across a skeptic who used a slightly different set of metrics, and calculated that a Vesuvius size eruption would melt an area of Arctic sea ice slightly larger than the ENTIRE state of Massachusetts!!!(assuming all the heat made its way through 4km of strongly stratified sea water) – then realized the relative scale of melt. The difference between average NSIDC minimum extent and current (or 2007) melt is ~2e6 km^2; the area of Massachusetts is 2.15e4 km^2. Yah think someone might have noticed 100 (give or take) Vesuvian scale eruptions, even if they were hidden beneath the Arctic ice?
    A mole! WHACK!!!