Sea Ice Update

I’ve said it before and I’ll say it again: one of the strongest evidences of global warming is the dramatic loss of sea ice in the Arctic. Northern hemisphere sea ice has taken a nose-dive recently, which has caused some speculation in the blogosphere that we might be headed toward a record-shattering melt season.

I’ve already made my prediction that this year’s September average sea ice extent from NSIDC will be 4.45 million km^2, which would be less than last year but not break the record minimum set in 2007. But there’s a large uncertainty in that figure, it could be higher or lower by nearly a million km^2. We can imagine three broad categories of possible outcomes: the extent could be more than it was last year, or it couldbe less but still not break the record set in 2007, or it could break the record. I’ll go so far as to say that there’s a 36% chance this year’s September average extent will be higher than last year’s, a 27% chance it’ll be less than last year’s but not break the record, and a 37% chance we’ll see a new record low. The recent nose-dive, although interesting, doesn’t motivate me to change my prediction.

This last winter, the extent of Arctic sea ice was greater than in other very recent years. That’s not such a surprise, after all it gets cold up there especially in winter. Nonetheless, some of the fake skeptics crowed about it. But although it covered enough extent for some to say it had reached “normal,” the ice itself wasn’t normal. It was thinner than before, so its total volume was still much reduced. Much.

There are three commonly used measures of Arctic sea ice. One is extent, which is the area covered by at least 15% ice. Another is area, simply the area of sea ice, so that a region with only 15% coverage only contributes 15% of its area to the total. The third is volume, which is the total volume of sea ice. Of the three, only volume takes into account how thick the ice is, and only volume reflects the extreme thinning which Arctic ice has experienced.

We have many sources for sea ice data online, and for each of the metrics there’s a source of daily data. Daily extent is reported by JAXA, although unfortunately their data don’t begin until mid-2002. Daily area is reported by UIUC through their website Cryosphere Today. Daily volume data are available from PIOMAS. What does each one show?

Let’s look at extent data first. The usual way to graph it is to overlay each year’s data on top of the others to show how the annual pattern has changed:

I’ve highlighted the 2012 data in red (and made the line thicker). Although during March and April 2012 ice extent was higher than average (but “average” reflecting only data since mid-2002!), it has recently taken a nose-dive and is now near the minimum for this date.

A representation I prefer is to show a time series plot of anomaly, which is the difference between a given value, and the average for the same time of year. This removes the annual cycle (or at least, its average) from the data. It also focuses attention on the trend, which is the truly important thing to note:

Sea ice extent has obviously trended down, even if we use only one decade of data (which is all JAXA offers). Yes, that downward trend is statistically significant. The nose-dive in recent days is obvious, at the end of the data.

Global warming should have its greatest impact on Arctic sea ice toward the end of summer, when the melt season culminates. This motivates us to study the annual minimum extent of the ice:

Again there’s a clear decline, and again it’s statistically significant even with only 10 data points. The summer minimum extent of Arctic sea ice is decreasing at nearly 150,000 km^2/year. Strong evidence.

Although extent data from JAXA don’t begin until mid-2002, area data from UIUC start in 1979:

I’ve highlighted 2007 (the record-breaking year) in green, last year in blue, and this year in red. Last year (2011) actually broke the 2007 record (by a hair’s breadth) for minimum area in the UIUC data, but not for minimum extent in the JAXA data. The recent nose-dive is there too, showing that area isn’t just nearly the lowest on record for this date, it is the lowest.

Again, I prefer an anomaly plot:

With over three decades of data instead of just one, the downward trend isn’t just obvious, it’s not just statistically significant, it’s in your face.

The annual minimum shows a similarly undeniable decline:

As undeniable as the decrease in extent is, as dramatic as the decline in area is, the decline in Arctic sea ice volume is nothing short of astounding. The year-on-year plot doesn’t really show how steady the loss has been because it doesn’t show which year is which:

The anomaly plot does show how steady the loss has been:

Do I need to say it?

The summer minimum sea ice volume hasn’t just declined, it has decrease by 75% during the time span covered by the data:

It wasn’t that long ago that Anthony Watts made quite a fuss over claims that we had lost 75% of Arctic sea ice, calling the idea “ridiculous” and “patently false.” He was pretty thoroughly raked over the coals, because the statement was about summer minimum sea ice volume — which has declined by 76%. In my opinion, for Anthony Watts it’s the truth itself that is “ridiculous” and “patently false.”

A few more predictions: If this year’s minimum is higher than last year’s, then the fake skeptics will crow loud and long about “recovery” of the sea ice even though the trend remains undeniable. They have to take anything they can get — and when it comes to Arctic sea ice, they’re desperate. Anthony Watts will probably mention Mark Serreze from NSIDC and ridicule his notion of an Arctic sea ice “death spiral,” even though a death spiral is exactly what we’re witnessing.

If this year’s minimum is less than last year’s but doesn’t break the 2007 record, the fake skeptics will also talk about “recovery” and how there’s nothing to worry about.

If this year’s minimum does break the 2007 record, the fake skeptics will adopt one of two strategies. The dumber ones will talk about how it’s all because of something — wind, ocean currents, the Arctic oscillation, solar ultraviolet, leprechauns — anything except global warming. The smarter ones won’t breathe a word about it.

What will this summer bring? Will we break the record for the summer minimum again? In any of the measures? In all three? I don’t know. In addition to trend, sea ice shows fluctuations. The uncertainty in predictions is sizable. But this much you can expect: the trend will continue. Because of global warming.

I’ve said it before and I’ll say it again: those who deny that Arctic sea ice decline is powerful evidence of global warming aren’t skeptics. They’re fake skeptics.

151 responses to “Sea Ice Update

  1. Looking over the graphs my gut feeling tells me that there will be less ice then in 2007. The steady reduction in volume since must have an effect on extend and area some time.

    A lot of details can be found here:

  2. I read that post and comments over at WUWT, pathetic, they’ll take anything they can get these days, but then desperate times call for desperate measures, the business of denial is getting harder and harder, what with the relentless ice loss, global temperature rise and sea level rise.

    I was tempted to point out that the “amount” of solids is normally measured using weight, so actually the summer minimum has decrease more than 75% as the new ice is not as dense as old ice … I was tempted for all of 2 seconds that is, then thought better of it.

  3. If this year’s minimum does break the 2007 record, the fake skeptics will adopt one of two strategies.

    You forgot one: the figures are mistakes or fakes. ;)

    Great article, as always. My guess is that volume will break the record even if area and extent do not, even though they will still be very close.

  4. Halldor Bjornsson

    The decline in sea ice is incontrovertible. One of the issues is whether the thinning of the sea ice (and a higher fraction of young ice) is making the system more erratic. If you look at your UIUC anomaly chart, the variability in recent years looks different than, say the decade from 1995- 2005. One of the interesting things to note in this regard is a buildup of relatively low salinity layer (freshwater anomaly is the oceanography jargon, – although it is still quite salty) in the Arctic. Such anomalies make it easier to form ice during winter. Current thinking is that this is wind driven and if (when?) winds change the anomaly might propagate out the Fram Strait into the Nordic Seas.(

    Last time that happened was in the late 60’s and it lead to anomalously cold conditions along Greenland and down to Iceland. It will be interesting to see if we are in for a replay of another “Great Salinity Anomaly”, but how it will play out in a warming world is anybodies guess…..

  5. Whether there is a record Arctic sea ice low in September will depend on weather, which makes ~300,000 km2 noise on the climate signal. Decreasing albedo early in the summer as seen this year leans the odds toward a record low. Incidentally, I believe the axis in the volume plot should be 1000 km3, not 1,000,000.

    [Response: Right you are, it’s thousands not millions of cubic kilometers.]

  6. Death spiral graph:

    • That’s a very cool (sorry!) visualization.

    • Yay! Not for the answer shown there, but because someone has carried out an idea I was about to suggest — plotting sea ice information on a polar plot rather than cartesian. The figure does look useful, as I’d hoped. Much easier to see things with strong annual cycles this way.

    • Fascinating topography. Almost a 3D reniform. In fact, it would appear that it will go from reniform to cardioid the first year that there is no summer sea ice.

      I wish that I had a better handle on the mathematics of the geometry – it’d be an interesting way to integrate the future.

      I’ve been playing around with graphing the volume data in a similar fashion, and curiously the pattern is ovoid up to about the turn of the century/millenium, after which the reniform shape starts to appear in the ice volume parameter. I currently don’t have a computer on which I can easily graph the data in 3D, so if someone here could whip it up quickly, I’d like to see. Especially if it gives a hint of when the sea ice volume version of the 3D cardioid might appear…

      • Dang, I only meant to italicise the word ‘volume’.

      • One of the things that strikes meas I have played around with these radar graphs is the consistency of trends in symmetry, and in changes thereof, as the years pass.

        Any denier of global warming who would rather evoke one or more choices from a lottery of alternative transient phenomena (such as wind…) needs to address the underlying reason why the Artic ice architecture is changing according to a very distinct and temporally extensive signature.

    • TrueSceptic

      Sorry about the stupid question Daniel, but where do the months begin and end? Does each radius represent the start of the month with that label?

      • I didn’t make the graph. Sorry for not pointing that out earlier. It is Jim Pettit from who made it.

        Each cricle represents a sea ice area. Zero sea ice area is in the center of the graph (origo).
        Each line (radiating fron the centre) represents a mounth.
        The coloured circles demonstrates how sea ice area changes throughout the year. In between the october and the september line you see that the coloured lines all approach the centre (low sea ice area) because that’s (middle september) when the area is smallest.

      • I realize I never answered your question. As you write, a line represents the first day of that mounth.

      • Thanks. The meaning of the graph was obvious apart from the position of the months, which you have now explained.

        I asked you because you posted the link to it. ;)

    • Beauty! In polar coordinates no less.
      Only a very minor adjustment between day of year and degree.

    • My two cents on colors over time: Make a consistent change in the RGB values over time. It would be much more visually striking.
      Something like:
      r = the 5-year running mean for day-of-year
      theta = day-of-year * (360 / 365.24)
      color =
      R: 128 * (days_since_start – days_total) / days_total
      G: 128 * (days_since_start – days_total) / days_total
      B: 255

      Many variations are possible. This would produce a roughly spiral shape that went from pale blue in the most distant past, to pure blue in the most recent. Probably have to adjust the running mean, tradeoffs between a smooth graph and being able to see recent measurements.

    • Great way to visualize the data. Here’s an animated version:

  7. Sea temperatures are high at the moment. See: ( In “Select Parameter”, click ‘anomalies’. )

    H/T Jason Box on Twitter

  8. I applied a quadratic fit to the September average extent data from NSIDC and of course was able to make the same prediction as tamino. What I found interesting is that when I performed this same exercise to earlier years, as though I were attempting the prediction in earlier years (“retrodicting”?), the method overestimated the following year’s value in 7 of the previous 11 years. (I stopped going back in time at 11 years because I got lazy and the available years of data is ever-decreasing).

  9. The Intrade Market “Arctic sea ice extent for Sep 2012 to be less than 4.3 million square kilometres” is a “low volume” market (no pun intended), but it is predicting a 48% chance of a new record now.

  10. If this year’s minimum is higher than last year’s, then the fake skeptics will crow loud and long about “recovery” of the sea ice…

    Thanks to “Cadbury’s Law,” as stated by Russell at the Rabett hutch:

    “Since the melting point of ice is constant, climate change cannot happen.”

  11. Two thoughts. First, to make the overlaid graphs clearer, you might consider using a greyscale gradient, so the first year’s data is at 10%, the next at 20%, and so on until the current year, at 100% black (or so other color gradient, for example from pure blue to pure red).
    Second, you (or someone with a strong constitution) could mine WUWT to demonstrate that they accept arctic ice loss – just look at the predictions they give for each year, and note how they decline precipitously. What better proof that they accept this, while claiming not to? Also note, of course, Steven Goddard’s (IIRC) late-season prediction of a few years ago, that was close to eventual values, but done much later in the season than other estimates.

  12. Well, Tony has already started on it’s the currents not the temperature. It would help to have some support. There are some people over there who can listen. Just ignore the rest and soldier through it

    • Rattus Norvegicus

      Eli, I would love to help you but the “denizens” over there will never understand the difference between “proximate cause” and “ultimate cause”. It just ain’t worth it, although I have put in a bit of an effort here albeit on a different subject….

    • If wind can blow the sea ice completely from the Arctic, or ocean currents can sweep the sea ice completely from the Arctic, and if this is not due to global warming, then what is happening?

      And has it happened before? If so, where is the evidence for this phenomenon, and what are the global implications of such a profoundly strange occurence? And why is this not on the blunt side of Ockham’s razor?

      I’m a bit allergic to WWWt after a fruitless engagement with David Archibald’s nonsense about sea level rise, but if one of the hardy spelunckers here could ask the Denialati there such questions I’d be curious to hear about the answers – at the moment I’m not inclined to spare a tab and further crowd my RAM…

    • I would try to help, but am banned from the site because I cast aspersions on one of the opening posts.

      Watts quotes a lot of papers which show that wind currents due to changes in the Arctic Oscillation are driving the decline of Arctic Ice Extent. He neglects to mention that models show that global warming due to GHG’s will shift the AO poleward, and cause the very wind currents that are observed.

      • Or that thinner , more broken ice is much easier for the wind to move, or that there is also significant melting in place, as determined by various observations (not to mention common sense.)

        Though give him credit for at least allowing Dr. Meier to do that guest post discussing some of these things.

  13. I’ve said it before and I’ll say it again: those who deny that Arctic sea ice decline is powerful evidence of global warming aren’t skeptics. They’re fake skeptics.

    Well put, Tamino.

    • Yeah, I see the point. Would be nice to see a presentation w/the same virtues in terms of end-end connection of annual cycles but without the area fault.

    • Mmm–but I don’t think that counts as ‘ill-proportioned.’ Perhaps one needs clarity that the area ‘swept’ by the annual trace is not reflective of the area of the ice itself–a natural, but incorrect, conflation. Rather, it’s a picture of how the area changes throughout the year(s). As the comment indicates, the *radius* represents area.

    • And, since it deals in decadal averages, not individual years, it doesn’t really ‘spiral’ either–each decade forms a discrete ring.

      Though given the variability, it would probably just be spaghetti if you did go annual.

      • To make it spiral you could do an average for a given day of the year using a smoothing kernel which advances 10 days for every day you advance in time. The result would be smooth, show a decadaly averaged annual cycle, and advance 10 years per cycle. The end points might be a bit messy.

    • Death Spiral graph looks quite OK, but the point dboström makes annoys me too. Could, say, running 5-year (or whatever) averages on individual dates (or close-by (3-7 days?) dates to get more data points in the mix) make it look less spaghetti? Of course then one would have to apply different color for dates after 2009 because the amount of data points decreases getting closer to present. It’s simple enough to get the circular coordinates by dividing the date of year field by 360 degrees and converting to radians, but still there would be some work (at least for me) in getting the formulas correct within the tablebase. Well, probably it would still look somewhat spaghetti at least near dates when there’s least variance between years. But it would be nice to see how this method would pan out.

      • So, I guess my previous could be stated as “What is the purest spiral that can be computed from this sea ice data?”, if one wants some mathemathical challenge to it too. The trouble here would be to get the function display also the variation in winter (as the function would be a spiral, it should decrease also in winter though on data this isn’t very clear). This function could then be called the “19th-century-style scientific description of the disappearance of the polar ice on a warming planet” to get it sounding more like natural science ;-). And if there would be no error bars the requirement of absolute proof (as was common in the 19th century) would be met (to conclude the joke). .

  14. Sceptical Wombat

    Can anyone tell me how fine the grid is that is used to determine Sea Ice Extent? For example are we talking about a square kilometre or a hundred square kilometres having 15% ice cover?

  15. I see that yet again, we’ve (Tamino and I) got almost identical statistical guesses for September’s average extent. Mine is 4.39 million km^2, with 1 sigma = 0.5. In other words, I wouldn’t be surprised by a new record, but would be by the ‘return to normal’ that some sources like to claim.

  16. Earthfriend

    Excellent post, as usual! However, I disagree with the comment on the strategies of the fake skeptics. I think it´s a smart strategy to blame a possible record minimum – if it happens – on the weather. In a way it´s true, because the weather will decide if the record will be broken this year. So it´s a smart strategy to try to take focus away from the trend. And in this case it is possible to do so with statements that are true, but misleading. Of course it is not very smart to actually believe there is no trend, or that the trend is caused by weather conditions, but that´s another matter.

    • Earthfriend, I agree with what you have written, except for this: “In a way it´s true, because the weather will decide if the record will be broken this year.” Well, partially, but not entirely. If a record is broken, it will also likely be due to the fact that there is a lot of single-year, thin ice–and that is not so much the result of weather as it is of volume trends. That is–climate change.

      • Earthfriend

        I totally agree with you on the reasons for a possible record minimum. The statement that “the weather will decide” is among those I call true, but misleading. Because it´s not the whole truth.

  17. Sceptical Wombat,

    I believe the grid is square @ 25km per side.


  18. Timothy (likes zebras)

    If this year’s minimum does break the 2007 record, the fake skeptics will adopt one of two strategies. The dumber ones will talk about how it’s all because of something — wind, ocean currents, the Arctic oscillation, solar ultraviolet, leprechauns — anything except global warming. The smarter ones won’t breathe a word about it.

    Mr Watts has already laid out his stall – he has a countdown to September, since he has found some prediction that was made a few years ago that the Arctic would be virtually ice-free in summer.

    Even if this year is a stonking melt year, and halves the current record minimum, there will still be enough ice left that Mr Watts will be able to claim that it isn’t virtually ice-free, that the prediction was wrong, and therefore all of AGW theory is baloney.

    • Mr Watts is one of those people who can’t really be trusted with the English language. His WUWT story picks on a quote from a 2007 National Geographic item that reflected on the unprecidented 2007 Arctic melt season.
      “Just last year two top scientists surprised their colleagues by projecting that the Arctic sea ice was melting so rapidly that it could disappear entirely by the summer of 2040.
      “This week, after reviewing his own new data, NASA climate scientist Jay Zwally said: “At this rate, the Arctic Ocean could be>/b> nearly ice-free at the end of summer by 2012, much faster than previous predictions.

      Watts’ denialism converts this quote into “…the Arctic will be “nearly ice free” according to a prominent NASA scientist in a National Geographic article on December 12, 2007.

      I do wonder if Watts ever properly read his source article or whether he was still reeling from the audacity of its title “Arctic Sea Ice Gone in Summer Within Five Years?
      The source article actually ends with comment from the “two top scientists” & their 2040 prediction in light of the 2007 melt season.
      Last year, Cecilia Bitz at the University of Washington and Marika Holland at the National Center for Atmospheric Research in Colorado startled their colleagues when they predicted an Arctic free of sea ice in just a few decades. Both say they are surprised by the dramatic melt of 2007.
      Bitz, unlike others at NASA, believes that “next year we’ll be back to normal, but we’ll be seeing big anomalies again, occurring more frequently in the future.”
      And that normal, she said, is still a “relentless decline” in ice.

    • Straw goalposts having the obvious benefit of being more portable than more substantive constructs…

  19. If they can do it, I can do it.

    Functionally, the sea ice is gone when walrus or seals or polar bears do not have adequate habitat for survival – regardless of how much sea ice remains.

    There, I just moved the goal posts.

  20. Timothy, the prediction you mention was not actually a prediction, but a simple if-then statement made in late fall 2007, commenting on the very steep decline to a new September minimum that year. *If* that rate continued, *then* the Arctic could be nearly ice-free by 2012.

    From the National Geographic article that apparently is the source for this quote:
    “Just last year two top scientists surprised their colleagues by projecting that the Arctic sea ice was melting so rapidly that it could disappear entirely by the summer of 2040.
    This week, after reviewing his own new data, NASA climate scientist Jay Zwally said: ‘At this rate, the Arctic Ocean could be nearly ice-free at the end of summer by 2012, much faster than previous predictions.'”

  21. What is the physical evidence that Arctic sea ice loss is related to global warming?

    Summer surface air temperatures in the Arctic have been above freezing through out the recorded period and the thermal mass beneath the ice has always been sufficient to melt all the sea ice.

    There is evidence that advection is causing some portion of the loss.
    Where is the evidence that increased thermal energy, during the melt season, is causing the rest?

    [Response: I see that you’re not a skeptic.]

  22. Tamino – Comments on the WUWT thread you linked to are, oddly enough, now closed.

    • Might just be because the post is more than 2 weeks old.

      • You may well be right – looking at older posts, that appears to be around the timeframe used.

        Still, it’s amusing, as the last two posts on that thread were likely sparked by Tamino’s post here…

  23. don fontaine

    Hudson Bay sea ice at UIUC would be interesting because annual min an max area don’t change (all frozen to all melted). What changes is the time spent frozen, say time spent more than 50% area frozen. Can wavelet analysis detect changes in the position and width of a pulse (say square wave) within an annual cycle?

  24. Tamino: I’ve said it before and I’ll say it again: Likewise, from an earlier post: Am I missing something here? The temperature goes up; ice melts. It’s melting virtually everywhere. It’s just one of the physical manifestations of a warming planet. Tamino: it’s all because of something — wind, ocean currents, the Arctic oscillation, solar ultraviolet, leprechauns… The leprechauns must be busy since they’re also melting glacial ice worldwide, causing temperatures in lakes and rivers to rise and snowpack to melt earlier and, well, you get the picture…or not. Skeptical Science (6-19) has a similar story to tell:

  25. One would expect offhand as the arctic ice thins and becomes more annual first year ice (younger ice) the area anomaly would begin to fluctuate and that appears to be what we are seeing. Eventually that fluctuation will switch to a new state, but this could go on for a while since there are so many gracious pathways of buffer capability in the large ice sheets and deep oceans we have. Floating ice is just be bellwether.
    Unfortunately, deep ocean is a lot more fluid than large ice sheets.

  26. Rob Honeycutt

    I always say this: None of us will ever convince the “fake skeptics” that this is happening. They would rather go down with the ship than admit defeat.

    But they continually present an opportunity to teach others about what is actually happening by addressing their insanity as frequently as possible. Every time any one of us takes the time to show them wrong, we’re not convincing that person (nor do we need to), we’re influencing a broader audience of readers. Those silent readers are your target audience. Respond to the fake skeptic but write for the silent reader whom you hope to influence. They are the foil against which we can communicate the overwhelming body of science.

    You do a great job of exactly this, Tamino.

  27. Generally, looking at the graphs,extent,area and volume there is a point about a week after the solstice when they get steeper, and for seven weeks or thereabouts we have the main melt season. Right now we’re 10-20 days ahead of ‘schedule’ depending on how you view the numbers. Given the right/wrong local weather over the next 10 days the ‘head start’ could be 15-25 days which I think opens up the possibility, given that systems tend to collapse under extreme stress, of an ice free arctic, at least by the 15% criteria.. Not likely but maybe 1 chance in 8? no 6.

  28. Daniel J. Andrews

    Hadn’t seen that sinking of Hy-Brazil before. How remarkably like fake sceptics they are.

  29. Harold Pierce Jr

    You are looking at the wrong month of the ice extent graph. The important month is May.
    It really doesn’t matter what happens in summer. After the sun falls below the horizon, freeze up occurs quite rapidly. By mid Dec or so the ice is about 3 ft thick, and the ice road trucking seasons starts.

    Don’t you watch “Ice Road Truckers” on the TV. You should. You might learn something about the Arctic.

    • Harold Pierce Jr
      You say “The important month is May.” Why should May be important rather than this month?

      • Maybe because May shows the lowest decline? Just guessing here…

      • Harold Pierce Jr

        Note the tight range of sea extent in May. When the sea extent starts to drop in May, then I going to get a little nervous. That would presumably mean the winters are getting warmer.

        [Response: So you actually did pick May because it has the lowest rate of sea ice decline. Perhaps you didn’t consider that all you’ve done is give every reader here a textbook example of cherry-picking — deliberately ignoring all the evidence you don’t like, while assigning supreme importance to what you think supports your preference.

        Although your cherry-picking is blatant, it’s not really very smart. The data for May might look like a “tight range” to you, but to numerical analysis it looks like a statistically significant decline. Since sea ice extent in May has started to drop already, perhaps you should get nervous.]

      • Time for Harold to worry:

      • And there was me thinking it was because the nuts get gathered in May.

    • HPJr lives in his own mono-utopian little world, as evidenced by this comment he left at SkS:

      “I am organic chemist and quite well aware of this info you guys mention here most of which is pie-in-the sky nonsense!

      I stand by “Fossil Fuels are Forever”!

      Wouldn’t you really like to have a bright red Ford Mustang Conv. with a big honkin’ 5 L V-8?

      Go which watch “Mighty Ships” on the History Channel.”

      Moved onto Ice Road Truckers, have we?

      • Rob Honeycutt

        You gotta wonder how many organic chemists use the History Channel as their source of scientific data. //sarc

      • Alex the Seal

        I’m wondering which TV show he got his organic chemistry degree from.

    • Phil Scadden

      ‘Don’t you watch “Ice Road Truckers” on the TV. You should. You might learn something about the Arctic.”
      You could try turning your TV off and reading some science. You might learn something period.

  30. Ice Road Truckers (commercially initialized as IRT) is a documentary-style reality television series that premiered on History on June 17, 2007. It features the activities of drivers who operate trucks on seasonal routes crossing frozen lakes and rivers in remote arctic territories in Canada and Alaska, as well as Alaska’s improved but still remote Dalton Highway.

    Not sea ice, harold …

  31. Arctic sea ice is declining for every month, Herald:

    And very few of those ice roads are even in the Arctic (north of 66.562°), plus they are on fresh water lake and river systems, not sea ice.

    You should try looking at a map sometime, you just might learn something about the Arctic.

  32. Someone earlier mentioned the Economist this week and their article on “The Vanishing North” which is an interesting read. In a separate article they noted “But America has a unique distaste for evidence.” I don’t always agree with the Economist but they certainly got that right. If I were Canadian I might say, we see a lot of that here, eh?

    • I am a Canadian, living in Atlanta, and I do say that–though it should be noted that the current Canadian government appears to me to possess much the convictions of the Tea Party on climate change, plus the hypocrisy to cover up those convictions in the interests of pandering to a public more convinced of (though still not sufficiently worried about) the problem of GW.

      The Harperites are shutting down worthy research efforts in favor of buying expensive military hardware of dubious utility, and the role they have played at Rio, as in the last several conferences, has been null at best and obstructive at worst. They’ve shut down the advisory council on GW. And of course, the oilsands are, well, highly addictive.

      It’s been a long fall from the days of Kyoto.

  33. This reminds of David Griffiths, “Introduction to Elementary Particles,” at p. 27 where he states about the conservation of charge and energy, quoting Richard Feynman, that ‘whatever is not expressly forbidden is mandatory.’ Mandatory is kind of a tough sell; but it seems the High Arctic has bought it.

  34. Kevin MacDonald

    It turns out there’s no need to panic, Arctic sea ice isn’t declining, it’s re-normalising: It seems that the time may have come to declare that the arctic sea ice has in fact reached a “New Normal.”

    • Little doubt that this is part of the “New Speak” development. Next up “New” is dropped, it’s called “Normal” and when it’s lower it’s just a within X Std.Dev, he gets to choose what standard is [in his weather business], and if it’s up it’s “Recovery”. What’sup “New”?

      • Kevin MacDonald

        The problem is Arctic ice loss, in any metric, is accelerating; will it normalise long enough for “new” to be dropped?

        Personally, if it didn’t carry such nasty implications for the planet, I’d be looking forward to the forthcoming post, 2 or 3 years hence, where Watt’s gloats about Arctic sea ice recovering to 2007 levels.

    • Well… The “new normal” expression comes from the ARCUS report, not Anthony. Also, I don’t think it is meant as a reassuring statement. The “new normal” that we can expect for the future is in fact continuous ice loss.

    • Susan Anderson

      Forget history. Move the goalposts, we have a “new normal”. Forget … that which we are doomed to repeat if we forget. But … unless we remember there won’t be …

  35. KD: It turns out there’s no need to panic, Arctic sea ice isn’t declining, it’s re-normalising: It seems that the time may have come to declare that the arctic sea ice has in fact reached a “New Normal.”

    BPL: That’s why God made Chow tests.

  36. WUWT predictions for minimum sea ice is based on a readers poll?

  37. That is why I called it the New Abnormal at the end of last melting season.

    This melting season is about to end. We have to wait and see if new minimums will be reached on the two popular extent graphs from IJIS and NSIDC. With the current weather forecast I’d say the odds are against that happening, but this only underscores the new abnormal in the Arctic: despite adverse weather conditions this melting season is on a par with the 2007 freak melting season. If these general circumstances persist, the Arctic will be very close to becoming ice free by the end of summer before 2020. Sooner, if we get a melting season with the same weather conditions as 2007.

  38. muoncounter

    Times they are a-changing: Even la Curry recognizes ” … the recent decline of Arctic sea ice has played a critical role in recent cold and snowy winters.”

    Not long ago these would be heresies: a) an admission that sea ice has declined; b) an admission that warming -> sea ice decline -> cold/snowy winter. The abstract (the full PNAS article is paywalled) also suggests that the changing pattern isn’t the good old AO: ” … clearly different interannual variability than the classical Arctic oscillation.”

  39. The phrases ‘new normal’ or ‘new abnormal’ should not apply to anthropogenically driven plateau, since they are due to recent and reversible changes in anthropogenic inputs — unless they have triggered true phase changes. Which is precisely what we are trying to avoid.

  40. “The abstract (the full PNAS article is paywalled) also suggests (…)”

    PNAS articles are open access, just click in the PDF option, it leads here:

    Click to access 1114910109.full.pdf

    • Sorry, but still ending at the sign-in paywall.

      • muoncounter

        Here’s what looks like a prepub pdf. Almost reads like a textbook in arctic amplification:
        When highly reflective sea ice is replaced by open water during the ice melting period, there is a substantial solar heat input directly into the ocean, increasing the heat stored in the upper ocean. … Warming of the upper ocean retards the recovery of sea ice during the fall freeze-up. As a result, the ice coverage in late autumn and early winter for the past few years is significantly below the mean of 1979-2000, exceeding two standard deviation of ice variability.

  41. Something to consider when thinking of ‘new normal’ is that the old ‘normal’ was established very suddenly, my understanding is that to flash freeze a mammoth [with springtime vegetation in it’s stomach] fast enough to keep it edible for millenia takes a temperature of -150c, and that the cause of death seems to have been asphyxiation. My point being that a frozen eastern siberian- laptev sea, and yakutsk permafrost was established suddenly and its historical condition not the result of climate driven forces , but once established reinforced its own new normal.

  42. I think the flash-frozen mammal may be an urban legend, but I don’t know the details.

  43. Had a very interesting conversation with one of my friends. He has no academical degree, actually, he has been working in warehouse for a years. But, when I told him about Northern Passage (sailing trough NE and NW passages in a plastic yacht) in one year, he was so surprised about changes in the Arctic. Just he remembered all ice breakers needed to complete that way. Yes, he is now close to 50 year, he remembers what have been taught in his younger years and it is so strange to him. I know this is no evidence, but it is striking somehow to me.

  44. Bryson Brown

    I think the flash-frozen mammoth story appears in Velikovsky– and I’ve seen it in creationist literature too, where it’s aimed at lending credibility to the cartoon catastrophism that young earth geology requires. I’m inclined to think a dead mammoth lying on snow or ice at 40 below would cool pretty quickly in the wind… a serious calculation might persuade me otherwise, but flash-freezing doesn’t generally involve liquid nitrogen or even dry ice.

  45. Phil Scadden

    This is creationist stuff. Perhaps time to read
    Guthrie, R. Dale, 1990. Frozen Fauna of the Mammoth Steppe. for a more scientific treatment.

  46. On a flat frozen peatland a carcass that big would gather heat rather early in spring and sink to the level of permafrost quite fast. Probably the lowermost parts would be in the best condition. I do not remember the temperature range for the bacteria that start the decay but anyway the decay is very slow just above freezing. I’d guess people who know how to hang meat (or fish) to dry would know about this, but I do not know if the baby mammoth in question was bled to death, which might have happened by a pack of wolwes and an over-protecting mama mammoth and relatives guarding the dead offspring until the teeth of the wolwes couldn’t anymore bite the frozen meat.

  47. Let’s not forget that this winter skeptics were wailing about an increase in the winter ice cover (thanks to an unusual cold snap in the Bering Sea, but you couldn’t tell them that).
    Whether you are a skeptic or an AGW-persuaded, you cannot go by a few weeks data. Weather anomalies happen all them time. You must wait for the fat lady to sing! End of September

  48. Not sure if this was mentioned but the June sea ice area was the lowest on record for that month (NSIDC):

  49. I avoided reading Velikovsky for years as a result of the almost universal opprobrium that surrounds him, i regret that. The above statement of -150c was an error of memory it should have been -150f, this
    is in my view fairly even handed, that is it doesn’t contradict anything i’ve come to accept, from an almost lifelong interest in the subject.

    [Response: I remember Carl Sagan’s comment on Velikovsky’s work. At first he thought it was fascinating and insightful history and archaeology but of course the astronomical ideas were idiotic. Then he asked some historical/archaeological scholars who said that in their opinion, it was fascinating and insightful astronomy but of course the history and archaeology were idiotic.

    Velikovsky was a good writer. I don’t regard him as a reliable source.]

  50. It was that sort of view that kept me away from him for so long and it is a wonderful peice of rhetoric, but within this context [sea ice] and what the future possibilities are i think it’s important to establish whether the mammoths who seem not to have been fitted out [long hair not fur and about the same subcutaneous fat as an elephant] for arctic living, died in a sudden cataclysm when the forests of eastern siberia were flattened giving rise to prodigous amounts of wood filled ‘muck’ that only now begins to release the brewed methane, most worryingly from the continental shelf of the laptev, or not. In view of the warm southern waters flowing down the lena and the ob arriving in about ten days bringing high humidity and intense heat to the siberian coastline it’s too late to do much, but if it could be established that prior to the cataclysm this area had a climate similar to an english spring, which both legend and the stomach contents of the mammoths attest to we may develop some idea as to whats in store.
    To the best of my ability to find out only velikovsky predicted the abberant rotation of venus and its surface temperature and was almost universally ridiculed before his predictions came true. Also after reading akenatun /oedipus i read all the ‘theban’ plays and became convinced that not only was he right but that the other characters were tutenkamun his wife/sister [helen] who after his+his’brothers death in twelve against thebes was kidnapped/rescued by paris who was a prince/king of a city in the nile delta and pursued by two other princes/kings of other delta cities and a fight for her hand/ the succesion took place at troy. Of course this could all be wrong but would surely upset the applecart for historians and even i have difficulty contemplating venus’s recent arrival, mainly i think because the ‘herd’ has such a strong consensus but thats been wrong before.

    • Andrew Dodds

      I’d have to point out that Venus is not ‘Hot’ in reality – it is exactly the temperature it should be (i.e, once you’ve allowed for the greenhouse effect and distance to the sun). It is certainly not ‘hot’ in the sense that an object fresh from the oven is hot – an object just out of the oven is out of equlibrium.

  51. johnm33: Velikovsky was badly wrong, about almost everything. As far as the ‘flash-frozen mammoths’ go, start with Of course follow up with the sources cited. n.b. the Ted Holden mentioned was and is a big fan of Velikovsky’s.

    My favorite illustration of Velikovsky’s work is his argument that clouds have antigravity (see #3, and note that I’m referring you to his own writing). More generally, that there’s no such thing as gravity.

    Not sure what Velikovsky has to do with arctic sea ice.

    • Robert Grumbine
      The top link comes up 404, I didn’t bring up velikovsky i’ve only very recently read his catastrophe stuff,[and don’t know what to make of it] there are plenty of books/sites about the mammoths, my point is that if these beasts were flash frozen perhaps the east-siberian arctic was too and that established a new normal which is being rapidly altered. The center of the ice in the ice age seems to have been between greenland and the great lakes, ok the bering straight is now open but that does not preclude the possibility of the old pattern of snowfall returning, which seems increasingly likely as the gulf stream penetrates further into the arctic and its motion gets converted to heat under the ice.

      • thanks ligne, I manually typed rather than cut and paste.

        johnm33: Do read the article. if it were true that the mammoths were flash-frozen by such extreme event as you are thinking, that would certainly be quite interesting. After you read the article, you’ll see that it is unlikely to have been the case. Even before that, one could ask why, if we’re seeing ‘flash-frozen’ mammoths, there are few or no flash-frozen deer, caribou, rabbits, squirrels, people, … A massive regional event like you’re considering would have frozen everything, not just the most massive of creatures in the region.

        The Gulf Stream doesn’t go to the Arctic at all, and won’t under any conceivable climate change. The North Atlantic current makes it in to the Nordic Seas, but, again, there’s no plausibility to it passing in to the Arctic through the Fram Strait. (See any introductory oceanography textbook that mentioned the North Atlantic current and you’ll see what I mean.)

        The energy of motion, vis viva if you’re reading old enough texts, kinetic energy for modern, is trivial compared to the energy of melting ice, or of changing temperature of water. Water density is on both sides of the equation, left hand side is how much temperature rise in the water you’d get, and right hand side is the kinetic energy:
        C_p * \Delta T = \frac{1}{2} u^2

        Cp * DT = 0.5 * u^2

        U is order 0.1 m/s. Exceptionally strong — which excludes the North Atlantic Current in the Nordic Seas — is 1 m/s. Cp is the specific heat of sea water, approximately 4000. DT is the temperature change if you managed to convert all the kinetic energy to heat, and is about 1.25e-6 K. A micro-degree is sometimes important, but I doubt this is one of those times. Even for the Gulf Stream, at 1 m/s, it’s only 100 microdegrees, 0.1 millidegrees.

      • Thanks for that quite an interesting read but nothing new, pity the link I put up seems to have ended almost as soon as i looked at it ,spooky, but in it they mentioned that the upright frozen mammoths/rhino were all pretty much aligned north south, also that the smaller [than rhino] beasts were all washed/blown away and were found by miners and explorers amongst frozen forest debris smashed, if I remember correctly, on the northwest face of hills and mountains both in alaska and siberia. I’m not trying to make the case just looking for an explanation that includes all the facts.
        As to the speed of the ‘gulf steam'[sign of my early 60s education] I thought the 60th parallel moved at about 800kph and the pole at 0kph, so 800×1000/[60×60] -1 mps or is that wrong?

  52. Phil Scadden

    Johnm33 – “To the best of my ability to find out only velikovsky predicted the abberant rotation of venus and its surface temperature and was almost universally ridiculed before his predictions came true”.

    You need to improve your ability then if you are going to question science. “Broca’s Brain:_Reflections on the Romance of Science” by Carl Sagan would be a readable and accessible demolition of your assertions on Velikovsky. A what point exactly do you realise that something is a complete fantasy? If there was any grain of truth in his hypothesis (which break physics all over the show), then how do explain the ice-core record? A complete lack of catastrophe there.

  53. Johnm33 – visit WUWT, I am 100% certain that there are some there who are convinced that Velikovsky’s on the money.

    You’re not going to get traction here …

  54. Phil Scadden

    John, further to above, it might be better to think it through like a scientist. The ice cores record up to 800,000 years of temperature and atmospheric composition. They indirectly record a great deal more (precipitation, ray flux etc) .Their existence implies ice has been present in the core location for that length of time. Now if Velikovsky’s hypothesis was correct, what would you expect the record to show? (Would you expect them exist even?). What does conventional science predict – and what do the ice cores actually show.

  55. Tamino, I may have missed it somewhere in the comments, but in the post itself you didn’t mention a “loss percentage” for either extent or area, as you did for volume. If you still have the files, would it be possible to ask what those values might be?

    (Tried to work this up myself for NSIDC extent, but I’ve only got Excel, being a Bear Of Very Little Stats. After much cut-and-paste aligning of data, I realized that the days were out of order in the original files for some reason. Feh.)

    • Kevin,
      I have yet to find a statistical analysis I couldn’t do with Excel–including Monte Carlo studies. It should go without saying that the random numbers aren’t always random, but you can improve on the randomization to some extent.

      • I’m not trying to blame it on Excel, primarily. The undoubted problem is lack of knowledge/skill on the part of the operator. (Much more at home in Word!)

        Nonetheless, I think I have a reasonably time-efficient workaround to get the numbers I need. We’ll see if it pans out…

      • Seems reasonable, pending a little confirmatory poking around; my ‘quick and dirty’ estimate is that annual mean extent is down about 21% since ’79.

  56. …and AR4 had 2.7%/decade, whereas my “quick and dirty” version had 4.9%/decade. Given the radical collapse since AR4, that might be more or less consistent. (The latest AR4 data would have presumably been 2006.)

  57. OK, still flailing away at % decline numbers for extent and area. And it’s not (yet) working for me. Any help from the community would be appreciated. (And try not to snicker, OK? Dunning wanted to write about my stats, but Kruger insisted I didn’t meet the self-delusion criteria…)

    I started looking for area numbers. So I downloaded the Northern hemisphere monthly area files (“nasateam” version, FWIW), computed an annual mean from them, used the TREND function in Excel to fit a line (and oh, the hours messing with LINEST and SLOPE and heaven knows what!), then used the endpoints of that line as max and min values to compute a decline percentage–8.4%. That’s not *per decade*, that’s overall–much lower than the ‘quick and dirty’ method I’d used for extent (which just averaged monthly percentage declines as given by NSIDC.) Also not really in line with published values, I think.

    So I went back to sea ice *extent,* thinking that I could at least use consistent methods. Lo and behold, by this method the extent decline is about 5.7%. That’s at least consistent with the intuition that area should show greater decline than extent, but it’s still lower than I would think it should be.

    Of course, if I want to compare apples-to-apples with the volume decline figure in the post, I really need to redo this for the minima (much easier, luckily.) And that will account for the lower numbers to a considerable extent, since the decline is greatest at minimum (and that’s arguably the most important number, I suppose.)

    But I’d still like to understand the mismatch between ‘quick and dirty’ and the version outlined above. I could understand some discrepancy, but such a wide gap says to me that I’m quite likely missing something big–something conceptually important.

    Suggestions, anyone? Comments?

  58. One last comment on this, just for the record–I calculate the numbers for minimum extent and area as -38% and -39%, respectively.

  59. Philippe Chantreau

    Is it me or does CT’s last image show the Northwest Passage pretty much open?