A Few Things Ill Considered

Arctic sea ice extent tracking below 2008

This is the headline on a press release from the National Snow and Ice Data Center (NSIDC) at the University of Colorado at Boulder. They are one of the goto sources for data about sea ice, either arctic or antarctic.

The final extent of arctic sea ice melting this year, as in any other single year, is a product of (large) local weather effects superimposed on (smaller) climate trends. That makes the final minimum extent very hard to predict, which has made it a bit of a horse race for the various scientific agencies that try to understand and model sea ice dynamics.


Not so hard to predict is what the denial-o-sphere’s reaction will be if the current track continues and the final melt comes in lower than last year. If you recall, last year was the second lowest extent on record and somehow the usual suspects managed to see this as sea ice recovering. This year, if last year’s “trend reversal” is re-reversed it will be their silence that will be deafening, not their hysterics.

The year on year change is not significant if you are interested in climatic changes. What is significant is the trend derived from an examination of the entire satellite record available:

From the press release:

During the first half of July, Arctic sea ice extent declined more quickly than in 2008, but not as fast as in 2007. As in recent years, melt onset was earlier than the 1979 to 2000 average. International sea ice researchers expect another low September minimum ice extent, but they do not yet know if it will fall below the 2007 record.

It is worth remembering that this is about ice extent (the area covered at least 15% by ice). If you are interested in the total amount of ice, you need to look at its thickness as well. NASA’s Jet Propulsion Labratory had a news release a couple of weeks ago that shed some more light on this other aspect, and the findings there are similarily dramatic.

From that article:

Arctic sea ice thinned dramatically between the winters of 2004 and 2008, with thin seasonal ice replacing thick older ice as the dominant type for the first time on record. The new results, based on data from a NASA Earth-orbiting spacecraft, provide further evidence for the rapid, ongoing transformation of the Arctic’s ice cover.

Unlike the ups and downs of extent, or temperatures, the average age of the ice is directly dependent on last year’s average age (and the years before that). If last year, nothing was older than first year ice, then this year, nothing could be older that second year ice. So looking at a point to point difference, like 2004 to 2008, while not the whole picture, is not the same meaningless exercise it is when looking at extent.

Putting extent decline together with thickness decline it is not unlikely that 2008 was in fact a record low for arctic sea ice volume.

It will be interesting to watch how this year continues to progress. Regardless of the final numbers, there is no evidence of anything but dramatic decline.

Comments

  1. #1 dhogaza
    July 25, 2009

    With the current high temperatures in the arctic other than north of the American mid-west (Hudson’s Bay has stubbornly lagged the 1979-2000 average as a result), and the recent steepening of the slope of the daily ice extent graphs from IJIS and NSIDC, it’s going to be an interesting stretch run …

  2. #2 barry
    July 25, 2009

    There was a bit of a kerfuffle when NSIDC showed less extent than other monitoring groups. Here are links to current Arctic sea ice extent time series for two other groups.

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

    ROOS – http://arctic-roos.org/observations/satellite-data/sea-ice/observation_images/ssmi1_ice_ext.png

    NSIDs estimate through last month showed more ice than the other two. The new state of affairs attracted no comments from the skeptiverse of course. They’d had their pound of flesh.

    The IJIS link is good for the spaghetti-ish graph that overlays the time series profiles since 2002. It’s clear from eyeballing that just how much weather impacts the growth and decline of the ice.

    dhogaza, it is going to be an interesting horse race. My naive bet was a 4.8 km/sq minimum in September (third lowest minimum on record). Currently my guess is looking a little high – but that could easily change. I’m wondering what kind of impact, if any, the growing el Nino will have on developments.

    ROOS hasn’t updated since 22/7. It will be interesting to see how their trend line looks when they do.

  3. #3 dhogaza
    July 25, 2009

    dhogaza, it is going to be an interesting horse race. My naive bet was a 4.8 km/sq minimum in September (third lowest minimum on record).

    A week or ten days ago, I would’ve agreed …

    Currently my guess is looking a little high – but that could easily change. I’m wondering what kind of impact, if any, the growing el Nino will have on developments.

    Well, here in the PNW, we’re facing near-record highs, and almost all of Canada is well above average except that little stubborn spot in Hudson’s Bay which is low (with surrounding areas down into the US being right about average). That ice will melt, regardless.

    One guy … a field tech or local researcher, I’m not sure … who posts at RC thinks that these high temps are going to lead to a huge ice melt over the next couple of weeks.

    We’ll see …

    ROOS hasn’t updated since 22/7. It will be interesting to see how their trend line looks when they do.

    Yes, indeed. It’s already interesting that their higher values are almost crossing 2008 …

    Meanwhile, I don’t bet on horses, or arctic ice. I prefer bridge :)

  4. #4 S2
    July 26, 2009

    If we do come in at somewhere between 2007 & 2008 I think that could mean that we would be more than 2 standard deviations below the trend line – for the third year running.

    I’m studying maths but I’m not yet good enough to be confident about what this would mean. If the three instances had happened at different points in the series I could cope with it – it would be unlikely, but not unlikely enough to invalidate the linear trend entirely.

    But three in a row? I wouldn’t know how to interpret that.

    Could it still be roughly linear, or are we faced with a step change or an accelerating, non-linear change?

    Any pointers would be appreciated – especially mathematical ones.

  5. #5 Harry Hensworth
    July 26, 2009

    The absolute best website I have seen on the issue of global warming is http://www.climatedepot.com . I wish all Americans could know about it.

  6. #6 dhogaza
    July 26, 2009

    The absolute best website I have seen on the issue of global warming is http://www.climatedepot.com . I wish all Americans could know about it.

    Morano makes Rush Limbaugh look like a science genius.

  7. #7 barry
    July 26, 2009

    S2,

    can’t help with the maths, but while it is generally contended there’s been a step change around 2007, there is debate about that and it’s not yet certain. We’ve a few more years to wait, I think, until that becomes settled.

  8. #8 coby
    July 26, 2009

    Morano does use lots of large fonts and bold face and colorful text, so he really must know what he is talking about.

  9. #9 crakar14
    July 27, 2009

    What i find interesting is that 2007 appears to be way below the 79-2000 average from April right through to July, 2008 appears to split the difference between the two whereas 2009 was almost at the average in May but then lost more ice quicker to finish (July) nearer 2008 levels.

    Any reason for the quicker and slower melt rates?

    Is there a date that we use to measure from one year to the next or do we just compare May to May or July to July?

    Coby,

    Yes i agree he does use funny fonts, hurts your eyes abit if you stare for too long. Not sure if this translates into depth of knowledge though.

  10. #10 T.H.
    July 27, 2009

    Hi there!
    Between 18. and 25. july the artic ice, according to satellite measurements from the IARC.JAXA information system, have melted nearly 850 000 square km.
    The melt rate in july is just now only 3.8 % behind the melting speed in the record year 2007.
    Total melted area for the first 25 days of july is 2,4 millions square km, only 96 000 square km less than in 2007.
    As what I have found out by looking at previous melting figures from the last years, the speed of the melting has never been so high in july as it has been the last two weeks.
    It looks like 2009 is coming closer to 2007 every day.
    The explanation is probaply very thin first-year ice and weathern patters that favorize melting in high speed.

  11. #11 dhogaza
    July 27, 2009

    What i find interesting is that 2007 appears to be way below the 79-2000 average from April right through to July, 2008 appears to split the difference between the two whereas 2009 was almost at the average in May but then lost more ice quicker to finish (July) nearer 2008 levels.

    Any reason for the quicker and slower melt rates?

    Well, the 2007 melts was made more dramatic by some unusual wind patterns that pushed ice south into warmer water in part of the arctic. The denialsphere has made a big deal of this being evidence that “it’s not really global warming”, kinda ignoring the fact that the reason why the ice was so easy to push around was because it was thin, first-year ice and that in past decades in those areas there would’ve been thicker, stronger multi-year ice.

    I’m not sure but a reasonable guess is that the big downtick in the 2007 graph is related to when the winds did their deed.

    As what I have found out by looking at previous melting figures from the last years, the speed of the melting has never been so high in july as it has been the last two weeks.

    It’s much warmer than normal in the arctic. 20C at the northernmost town in Norway? Are you kidding me? Normal July daily maximum temps there average < 10C.

    Since 2008 was quite close to the 2007 record minimum (much closer than to the 1979-2000 average), much of the arctic is covered with thin first-year ice, and much of the multi-year ice has thinned due to melting outdoing refreezing over the last three decades (on average).

    So a good hot spell melts ice more quickly than is typical, combine that with the ice being thinner than historically typical, and you can see it takes less time for the ice to melt to the point of disappearance. Of course it’s not as warm as 20C up where all the ice is, but it is considerably warmer than normal.

    The cooler than average weather we’re seeing in the american mid-west and east coast extends northwards – there’s a stubborn bit of ice left in Hudson’s Bay, still. But that’s diminishing quickly, too.

  12. #12 dhogaza
    July 27, 2009

    Actually we’re under the 2008 extent as of the moment, according to both NSIDC (check their July 22nd press release, and keep in mind that the 2009 melt has accelerated in the few days since they issued it) and IJIS.

    Not much under, though, a decrease in melting could easily see 2008 jump into the lead again.

    Is there a date that we use to measure from one year to the next or do we just compare May to May or July to July?

    NSIDC does month-month trends as well as annual maximum and minimum, but the minimum’s apparently considered the most important. They show the May-May etc graphs with a linear regression trend line in each of their monthly reports, which come online a few days after each month ends.

    BTW, the date on which freezing outpaces melting – the summer minimum – is a week or two later now than in the past, you can see this in the IJIS “spaghetti” graph.

    The annual winter maximum is decreasing, too, but not as much as the summer minimum. If you look at maps of sea ice, you can see that the winter maximum is partially constrained by topography – freezing land doesn’t make sea ice! Once the sea has frozen up to the northern shores of Siberia, Alaska, and Canada there’s nothing left to freeze so the decreasing trend is confined to those areas where the arctic sea blends into the Atlantic and Pacific oceans, rather than eurasia and north america.

  13. #13 dhogaza
    July 27, 2009

    The explanation is probaply very thin first-year ice and weathern patters that favorize melting in high speed.

    Unlike 2007 I think (in my case that means “guess”) the exceptionally warm temps over much of the arctic, not unusual wind patterns, will be sufficient to explain it (combined with first year ice, of course). We’ll see what the monday-morning quarterbacks (scientists) say after the season’s over!

  14. #14 barry
    July 27, 2009

    ROOS have updated their extent time series to 25/7, and 2009 has just dipped under 2008 sea ice extent for that day. That’s all 3 institues I know of under now showing the 2009 melt more than the same time last year.

    http://arctic-roos.org/observations/satellite-data/sea-ice/observation_images/ssmi1_ice_ext.png

    If any critics happen by, yes, we’re mainly talking about weather here. It’s by no means certain 2009 will end up below 2009 at summer minimum (or above the 2007 record, for that matter).

    7 weeks to go, +/- a couple of weeks, weather depending.

  15. #15 caerbannog
    July 27, 2009

    Sorry about the off-topic post here, but it looks like Anthony Watts sicced the DMCA on Peter Sinclair aka greenman3160.

    Youtube just pulled greenman’s excellent “Watts up with Watts” crock-of-the-week video and replaced it with this message:

    “This video is no longer available due to a copyright claim by Anthony Watts Surfacestations.org”

  16. #16 dhogaza
    July 27, 2009

    Yes, apparently Watts did, though Greenman’s video is clearly fair use.

    Tells you what kind of people are being dealt with here, doesn’t it?

  17. #17 S2
    July 27, 2009

    Thanks, Barry.

  18. #18 Steve L
    July 28, 2009

    I don’t have much time to spend on sea ice, unfortunately. The data must be around … has anyone seen a relationship plotted between Arctic temperatures in given months and the resulting sea ice melt? I tried at one point to develop a ‘cohort model’ (I’m a biologist) to see if I could solve ‘survival’ rates for ice of various ages. Then I realized how silly it was to expect something like that to work — you need to take into consideration where the ice of various ages is. That’s too hard, but if one can put air temperatures into areas with given ages or thicknesses of ice, maybe it could be useful. Has anyone seen such a thing?

    Coby, you’re in Vancouver, right? Too warm to sleep? I recommend you do a post on our poor Fraser sockeye this summer. Fraser River temperatures are forecast to break all time records probably within the next week (data go back approximately 100 years).

  19. #19 greenfyre
    July 28, 2009

    Sticking with the theme of the off-topic topic (some people’s kids :-) )

    Now Frere Pielke has put an oar in and Siegel takes him to task (as well he should)
    Roger Pielke Sr speaks on Climate Crock: Laugh or cry?
    http://getenergysmartnow.com/2009/07/28/roger-pielke-sr-speaks-on-climate-crock-laugh-or-cry/

  20. #20 John T
    August 1, 2009

    I’m wondering if the recent northern hemisphere volcanic eruptions over the last two years may not help stock up the arctic ice.

    I live in the high desert of Southern California and have a very wide horizon. It’s possible to see the Rockies (sometimes, weather conditions permitting) from about 140 miles away. . .barely. Every sunset and sunrise is a dirtyish orange/yellow color in a thick belt along the horizon contrasting sharply with a very white and much broader belt above that and the blue sky above. The effect is actually quite striking, sort of like a volcanic “vanilla sky”.

    This is caused, so I’m told, by the sulfur dioxide barfed up by a series of volcanic eruptions beginning in late February of 2008 in the Aleutians and culminating in the Mt Redoubt and Sarychev (Kamchatka) eruptions this year.

    I remember Mt. Pinatubo and how we had some truly drought-breaking rains in the several years that followed. Should it be expected that precipitation stands a good chance of increasing, perhaps dramatically, over the next couple of years?

    I’m not arguing over the long-term trend of the chart, but the next several years precipitation, if the Pinatubo eruption scenario holds, could show a temporary reversal of Arctic sea ice declination. That is, unless some other factor overrides the sulfur dioxide/precipitation connection.

    One factor that hadn’t occurred to me until just now is how do atmospheric sulfur dioxide levels from Pinatubo’s eruption period compare with those levels now?

  21. #21 Chris Lane
    August 4, 2009

    Am I right in thinking that the lessening extent of Arctic Sea Ice is an indicator that AGW is currently occuring?

    Is it also reasonable to conclude that any future growth in the extent of Arctic Sea Ice is an indicator that AGW is not occuring?

  22. #22 dhogaza
    August 4, 2009

    What’s important is *trend*.

    Here’s what the july-to-july comparison during the satellite record looks like.

    A clear trend, with wiggles.

    Same holds true for the september minimum, and to a lesser extent the winter maximum (which is partly constrained by the fact that the arctic sea is largely surrounded by eurasia and north america.

  23. #23 ScottM
    November 28, 2009

    It’s good to see that the sea ice is recovering nicely this fall and since it was another unusually cold winter, spring and summer in the midwest I have to wonder if the weather patterns that set that up, pulled the cold air from the arctic, slightly warming that region, and reducing the sea ice extent somewhat.

    Now that the ice is back above 2006 and 2007 and almost equal to 2008, I suspect we’ll resume the trend of increasing sea ice in the arctic to go along with the stable Antarctic ice extent.

  24. #24 ScottM
    November 28, 2009

    I’m sure it will be comforting to those fooled by the climate change fraud to know that the gatekeepers of the information used to perpetrate the fraud have been exposed as scam artists set to fit the available data to their pre-conceived conclusions. To know that the globe is only warming when you manipulate data and exclude those inconvenient to the truth, must be a tremendous relief.

  25. #25 skip
    November 28, 2009

    Hey ScottM:

    A previous denier poster who has since fled in disgrace attempted to use this this sea ice level point.

    Are you going to just be a hit-and-run raider or do want to stick around and learn something?

    skip

  26. #26 coby
    November 28, 2009

    ScottM,

    Speaking of fraud, no honest person can look at the graph of arctic sea ice extent in the post above and say “Now that the ice is back above 2006 and 2007 and almost equal to 2008, I suspect we’ll resume the trend of increasing sea ice in the arctic” as you do.

    The emails at the center of “climategate” are paleoclimate reconstructions and have nothing to do with the study of seaice or the modern temperature record.

  27. #27 Dappledwater
    November 29, 2009

    “To know that the globe is only warming when you manipulate data and exclude those inconvenient to the truth, must be a tremendous relief.” -Scott M.

    Well, the cryosphere seems pretty darned convinced. What were you saying about inconvenient truths?.

  28. #28 Matt Bennett
    November 29, 2009

    Skip,

    Looks like ScottM’s a ‘hit-n-run’ – I thought so, you could almost SEE his narrative shining through….

    Pity, there really is so much to learn.

    Matt