Sea ice minimum kerfuffle

The summary of betting on sea ice refers. If you look in my comments, you’ll find any number of well intentioned people advising me that its time to close up the bets before I take a bath. But I haven’t. Anyone wanting to pile in is still welcome (if you can’t be bothered to look up the previous post the bet is simple: will this years Arctic sea ice extent minimum, as measured by the satellites, be less than last year? I say no).

i-1d2ce34f718bcf56cbda166133ef631c-N_timeseries.png The May 5th version of “Arctic sea ice news and analysis” provides some more fodder. I’ve ripped off a pic from them which I like. To me, it rather suggests no record this year. They somehow convince themselves that it does suggest a min. Hey ho, we’ll find out in a few months. Their section on “Estimating September extent based on past conditions” is cute. I like it; its a nice idea. Totally lacking, of course, is how would this idae have fared if applied last year, or the year before, or…

My personal opinion remains that we simply can’t forecast year-to-year variations with any degree of reliability. The long-term trend is clearly downwards, and there will be a new record sometime. But based purely on the behaviour of the sea-ice extent timeseries I still consider that new record this year is less than 50% likely. I also wonder if people aren’t in some danger of getting a teensy bit carried away publishing these “forecasts” which they don’t really believe. They are speculative prognosticaions, no more.

Andy Revkin (its him again!) covers this. In the comments, Bill Chapman is reported to say “I say the odds favor a new NH record minimum – put my money there.” I’ve replied (#40) but haven’t heard from him… I must give him an email. [update: I did. He declined -W]

[Update: via Gareth I discover this rather useful picture, which puts the present year more into context. It gives me hope for my bet. So, if you’re still interested… -W]

35 thoughts on “Sea ice minimum kerfuffle”

  1. While I agree with everything you say, I also know that the system has a lot of hysteresis, meaning that starting from a new minimum means that there is a lot less multi-year ice. You are basically saying all other things being equal, which is true of a random walk, and even in that case there is a 50% chance of a new minimum. All other things are not equal.

    [I don’t have the data to had right now, but my recollection is that it doesn’t support the idea of strong auto-correlation. I’m sure there is some. Maybe someone with the right software could calculate the lag-correlations… -W]


  2. Hi William,

    ok, I put £20 that this year sea ice extent minimum will surpass that of the last year. If I loose, you can give the money to NSIDC to do a better research 😉


  3. NSIDC came up with this interesting (and IMHO compelling)angle:

    “To avoid beating the September 2007 record low, more than 50% of this year’s first-year ice would have to survive; this has only happened once in the last 25 years, in 1996.”

    [Yes, I read that. So explain why they haven’t used this novel prediction method to post-dict previous years extent? -W]


  4. The current arctic sea ice area (cryosphere today) is the same as last year as referenced by Gareth, and the extent is greater (NSIDC). This implies that there is the same surface area for ice chunks as at this time last year but that they are more spread out than last year. With more edges it might melt faster.


  5. I’m probably wrong (I often am), but I think a lot hinges on the weather in July and August.

    In 2006 the ice melt was rather more pronounced than 2005 in the earlier months, but a “cool” August saved us from seeing a new record.

    In 2007 the ice extent was somewhat higher than 2006 up until the middle of July. But we had an unusually warm August in the Arctic, and the ice extent just kept on falling.

    I think it’s fair to assume that the melting rate in the next couple of months will probably be higher than we’ve seen recently, based on the thickness of the ice. But whether or not we break the record depends on Arctic temperatures in late summer, and we don’t yet know what they will be.


  6. Looking at the accompanying NSIDC bar graph, it doesn’t look as if the first-year ice survival stats would have been very interesting prior to the very recent large increase in the proportion of first-year ice. Am I missing something?


  7. Hi William,

    It all looks pretty close to my. You figure a less-than-50% probability of a new record this year; CCAR give 59%. I’d have thought that S2 (above) is right in saying that when it’s this close, it’s pretty much weather-dependent… Still: I’ll go a small bet of, say, EUR 40 (settlement via PayPal?) if you like. (You already have details on who I am etc. for our bet on sea ice melting by 2020, I think it was…)



    [OK, taken -W]


  8. If anyone besides me couldn’t make sense of the chart Steve Bloom links above — the explanation for it is on this page, a ways down:

    That page has links to several other pages that, after all read together, made this set of predictions a bit more understandable.

    I think it’s a “crushed ice or ice cubes in your drink?” choice coming along.


  9. And the light blue is one year old sea ice (now two years old, I suppose? That was late 2007)

    from Andy Revkin’s introduction to his recent annotated Polar Bear thread:
    where you can click the link to see

    “An animation of the age of sea ice updated through November 2007 can be obtained in MPG format. The red dots shows the current location of buoys used to estimate the age of sea ice. The areas of older, thicker ice are shown in white, while younger, thinner sea ice is shown as darker shades of blue.”


  10. 1) Since 2002 perennial ice area has shown year-on-year record losses, prior to last year’s melt the area was ~2.6M km^2, down from ~4M km^2 in 2002. And over this winter further perennial has been lost, I cannot find data to put that in context of those areas for March 2002/07. However, qualitatively the preceding trend of record lows in perennial area has once again been sustained this year.

    2) Yet since 2002 we have still seen the autocorrelation behaviour noted by William Connelly, Cecilia Bitz and others (no new record after a preceding year’s record), despite the ongoing precipitous year-on-year decline in perennial area.

    3) The above noted, I still agree with Steve Bloom’s previous statement to William Connelly that people lose money by simple techniques referring to previous behaviour when a major facet of the system has changed. The National Ice Centre data in particular shows that the ice is in a markedly poorer state than previous years NB hi-east arctic (centred on the Pole). As does watching timeseries of QuikScat and the amount of fracturing apparent from HRPT(IR/Vis) satellite images.

    In the limit we’re looking at a seasonally ice free state, a defining feature of which would implicitly be a lack of perennial ice. From my understanding, the closer we get to a virtually perennial ice free state, the more likely that the “autocorrelation behaviour” will be broken. i.e. in the final transition to a seasonally ice free state I expect to see a succession of year on year records in summer minima. The question remains are we now at that point? My only answer to that is “ask me in October 2008”. 😉

    From what I see there’s enough potential for “weather” variability to make this bet little more bounded than betting on pure chance. Which is the only reason I’ve not entered this particular fray to bet against William.

    That said for what it’s worth, my range of expectations is currently anything from 1M km^2 to ~3M Km^2 for the areal minima of 2008.

    However I suspect a role for structural weakness, similar to mechanical buttressing type failures of ice pack under tension (e.g. Beaufort Sea off Banks Island around 9 January, and from Beaufort to Greenland in early to mid April). For that reason I’d even be loathed to bet against an Arctic Ocean full of “crushed” ice by this September.


  11. As of this writing, that Arctic sea ice extent graph is 26 days out of date.

    If you look over at the Cryosphere Today, you can see what is really going on, and you can compare with days from last year or other years.

    Southwestern Greenland just experienced 3 days in a row of record-breaking temperatures (records set in 1998), and the snow cover has pulled back in that region at least 2 weeks earlier (or more) than it did last year.

    The snow and ice cover in the NH is pulling back rapidly and can be said to have pulled back roughly as far as it was at this time last year.

    Planetary moisture flows seem to be going non-stop toward the Arctic, carrying warm air up from the equator. Have a look at GOES WV.

    Also, from NOAA, have a look at an animation of the 500-hPa anomalies for the last 30 days over the Arctic:

    How much did you say you would be willing to bet?


  12. btw, today, I received this e-mail in response to a query I sent to the NSIDC:

    Dear Tenney,

    Thank you for contacting NSIDC. We apologize that the graph has not been updated in a few weeks. We are experiencing problems with the F13 passive microwave satellite, and therefore are unable to use the bad data. We are working to acquire data from a different satellite in order to fix this problem.

    I apologize for the inconvenience.

    Best regards,

    Donna Scott

    NSIDC User Services


  13. NSIDC extent is effectively neck and neck:

    Cryosphere Today, looks to me like area as of 1/6/08 (today) has dropped below that of last year:

    I don’t think that’s happened until now…
    But it’s about the only thing I’ve not kept running copies of.

    However as NSIDC noted on last years coverage:
    “…weather conditions in the Arctic are variable. For example, in July of 2006, we were also on track to set a record minimum, but a cooler and cloudier August slowed the rate of ice loss.”

    Right now I know how Indecisive Dave feels:


  14. Andy Revkin has another blog post about Arctic sea ice:

    Fourteen research teams studying the impacts of warming on the Arctic Ocean have issued independent projections of how the sea ice will behave this summer, and 11 of them foresee an ice retreat at least as extraordinary as last year’s or even more dramatic. The other three groups that issued a numerical estimate see the ice extent heading back toward, but not equaling, the average minimum for summers since satellites began tracking the comings and goings of Arctic sea ice in 1979. Five other groups chose not to issue a numerical estimate.


    [Interesting. But I can’t see where Revkin gets his figures from. Actually looking at the graph shows 6 forecasts above last year, and 8 below. Or as they say, 3 for more, 6 for about-the-same, and 5 for less.

    Anyone for 2-1 odds on it not going below 3.5? -W]


  15. The suspense continues to build. The crowd is hushed, watching, holding their breath, as the line advances toward the next curve ….


  16. >But I can’t see where Revkin gets his figures from.

    A slightly higher minimum from a higher winter maximum can be a “retreat at least as extraordinary as last year’s”. The maximum was more than .2 higher so I don’t see a problem with those 11:3 figures.


  17. Not that it matters, but solely because I have previously been largely siding with William on this: My opinion has changed.

    That’s not so much from new information, more from the hours I’ve recently spent studying time series of HRPT visible images and QuikScat. If there’s any “new information” that’s swayed me, it has been the persistence of the extensive fracturing across the pack that occurred just before 18 May 2008.

    I now think a drop below last year’s areal extent is most likely, although extent might still not drop much below last year.

    Hank, I agree.
    After months this still has me on the edge of my seat. But I don’t pay too much heed to the area/extent plots, it’s the detail that’s awesome.


  18. Cobbly, the fracturing was more or less set in ice, once the pack thinned/disappeared last summer. The most recent Real Climate post makes it clear why this had to happen given the loss of thickness and all the new ice.


  19. Eli

    I know thinner ice will fracture more, but what I am seeing is fracturing right through the heart of the perennial ice mass (as shown by QuikScat), with almost 3 months to go with sufficient insolation to open up these fissures. What will melt faster, 100km^2 more or less solid ice pack, or 10*10km^2 bergs with water in between them? All that’s needed to scatter this in the open waters that are spreading from Beaufort and Chucki are the right winds.

    The issue for me is no longer: Is the “Arctic System on Trajectory to New, Seasonally Ice-Free State.” It is simply: It’s happening, now how fast it will happen?.

    If William wins this bet (and area doesn’t drop substantially more below extent than last year), then we may be going through the sort of blip seen in model runs. Although with the trend of the last 10 years it would need a radical trend reversal for a recovery.

    However given graph 3 of Nghiem 2007 “Rapid reduction of Arctic perennial sea ice” (where perenial ice area is falling off a cliff as I described above), and the overall trend of area/extent reduction. I think we’re looking at the lower (Maslowki) end of the modelling, a very rapid transition. Indeed I think Maslowski’s 2013 may be an overestimate.

    Figure 5 of NSIDC 3 June 2008 page shows, in excruciating detail, the state of the ice cap around 20 May 2008. Note the series of stress cracks arcing westward up from the east of Ellesmere Island from the Lincoln Sea. And the further cracking right perpendicular to the Arcipelago coast through to the Beaufort Sea.

    On the Environment Canada HRPT visible image of 11 May and earlier, whilst not quite billiard ball smooth, there is much less facturing and overall the pack gives a smooth appearance. I had a break from then until 15 May (typical timing!) by which time fracturing had just spread along the archipelago to the Western edge of Ellesmere Island (then I was otherwise engaged – another break! Doh!). On the 18th there was a substantial fissure through to Lincoln Sea (E of Ellesmere) with a mass off fine fracturing out into the Arctic Basin.

    Check out the Env Canada Beaufort Ice-Pack page here:
    Open the NOAA animation and note what happens on 9 January 2008. That was in microcosm what happened just before 20 May 2008. BTW the movement of the ice away from the Archipelago that caused that and the ~18 May event (also early April) happens when the AO is -ve, implying to me it’s wind forced.


  20. More on Arctic ice from the BBC:

    “Data from the US National Snow and Ice Data Center (NSIDC) shows that the year began with ice covering a larger area than at the beginning of 2007.

    But now it is down to levels seen last June, at the beginning of a summer that broke records for sea ice loss.”

    [Looking at last years graph, the end of June should be the interesting time for the collapsists -W]


  21. Check out today’s MODIS:

    It’s not supposed to look like that.

    [I’m not used to looking at MODIS. Whats wrong? You mean the leads? They show up better in the 500m version -W]


  22. Yes, the leads (and yes they are clearer in the 500m but those images are so large that someone not knowing what they are looking for will not see what I mean).
    This ice is shattered.
    That MODIS site is excellent for getting today’s image of more-or-less anywhere.


  23. It’s clear that a lot of those predicting a new minimum for 2008 are in a lot of trouble now. That new ice just hasn’t melted as fast as predicted – simply because the weather/patterns have been a lot cooler up there this summer.


  24. Another US web site lying about Canadian ice!

    Notmany know that some scientists are betting on in which year they wont be ANY ice floating on water in the Arctic. Many bet on 2008, more bet on 2009 and max is 2010.

    The FREAKING US LIARS want the world to see it when it’s too late. Then they will send their war ships and oil tankers and drilling platforms on Canada’s land acting like though we were Irak or any other country.

    We had it with a country that can not even control itself. Zionist drive the USA like a puppet and you Yankees fall asleep in superficial consumerism with images of political correctness.

    You don’t oppose New World Order ? Let’s see what Darwin meant about species extinctions, this time it will be the Americanus automobilus stupidus hypnotesus 🙂


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s