Girding my loins: sea ice

Posted by William M. Connolley on March 12, 2013
(7)
More »

This is not the sea ice post you were looking for. However, it is a placeholder for putting comments, including linking to previous comments.

If I’m feeling energetic I may even make the linkages myself.

The pic shows seaice at “normal” ish; but that means little, as 2012 was also “normal” at this time of year. PIOMASS might be more interesting (thanks CR) but a month of more will make that clearer.

Early update: Oh well, since its there, Open Mind’s Arctic Sea Ice Loss, part 1 is worth a look, esp. figs 2 and then fig 6 (though I don’t think the quadratic fit is meaningful. Unless you can bring yourself to believe that the implied long-term-trend was an increase in the early 80′s).

(7)
More »

Comments

  1. #1 crandles
    2013/03/13

    >”(though I don’t think the quadratic fit is meaningful. Unless you can bring yourself to believe that the implied long-term-trend was an increase in the early 80′s).”

    Tamino has already responded:
    [Response: Depends on your purpose. If you're trying to mimic the detailed changes, perhaps -- but a smoothing method (like Lowess) is better. But if you're testing whether or not there's a change in slope, perhaps not.]

    [Not quite, that's a response to "Wouldn’t it be better to fit only the monotonically decreasing side of the quadratic?" I'm suggesting that the quadratic is *wrong*. And indeed we're sure its wrong, physically -W]

  2. #2 crandles
    2013/03/13

    I agree that the quadratic is wrong.

    However, you appeared to be complaining about Tamino’s article.

    [No, I was recommending it, with the caveat that the quadratic bit was dubious / wrong -W]

    Even without the reply I quoted, I think it is quite clear the purpose of what Tamino did was to test whether there is a change in slope. For this purpose the data looks sufficiently like a quadratic to use a quadratic. Therefore, I see nothing wrong with Tamino’s article.

    The reply I quoted clearly indicates that it “depends on your purpose” which seems a direct response to your complaint that it is not meaningful which was not the purpose Tamino had in mind.

    Anyway while you are talking about the quadratic being wrong physically, what do we want instead to be ‘meaningful’ and physical?

    I think the answer is obviously to model volume decline and maximum volume. With Cryosat2 now supporting PIOMAS that it is worth using it to try to get a physical extrapolation.

    [I still disagree about this. tantalisingly, I haven't got time now to explain why. But I will, one day... -W]

    As volume declines while not hitting minimum of 0 ice, it seems that more ice volume is melting. This is physical due to lower albedo and leads to an exponential function. There could also be a linear decline and/or other more complex terms.

    With a strong freeze negative feedback as you seem to like for your explanation for ice persisting for a long time, it is easier to accept this and look at physical factors affecting equilibrium thickness. I suggest the major terms are GHG levels affecting outgoing radiation (in a somewhat linear fashion) and ocean heat content transported into arctic and perhaps some solar heat absorbed by ocean and mixed downward. Due to thermal inertia these latter terms are going to accelerate upwards.

    I haven’t really got a good argument for the function depending on volume to cause an exponential shape. Nevertheless it seems to look like an exponential. Gompertz can also look like exponential but there hadn’t been much sign of this appearing in the data til we look at volume for day 60 for 2011, 2012 and 2013.

    That would be nice if the max volume isn’t going to decline any more. Unfortunately, I just don’t think that feels physically likely when we had such a record low extent in Sep 2012 leading to more FYI that will absorb more heat. So I think max volume will continue to decline for a bit yet. That means volume melted seems likely to also continue to rise.

    When I do fits with linear and exponential, the resulting fit turns out to end up steeper than just an exponential fit. If the function was more like a gompertz, the fit would end up being shallower than the exponential only fit. Maybe that will change when the April 2013 data is added, but I wouldn’t want to be relying on such a sudden appearance of a gompertz change in direction of the data even if models do seem to predict it will occur.

  3. #3 Eli Rabett
    http://rabett.blogspot.com
    2013/03/13

    Max area is a function of winter temperature south of the Arctic circle, so the changes in max area are telling you something different.

  4. #4 WhiteBeard
    2013/03/13

    Girding the loins, eh. A reaction to last year’s metaphorical chilly willy?

    I find Dirk Notz’s post at Real Climate quite worth remembering as well as the paper cited that he coauthored.

    http://www.realclimate.org/index.php/archives/2011/09/the-unnoticed-melt/

  5. #5 Another Week of GW News, March 17, 2013 – A Few Things Ill Considered
    2013/03/18

    [...] 2013/03/12: Stoat: Girding my loins: sea ice [...]

  6. #6 crandles
    2013/05/04

    Latest PIOMAS data:

    2011 day 90 21.568
    2012 day 90 21.665
    2013 day 90 21.612

    2011 max day 103 21.961
    2012 max day 114 21.923
    2013 max day 107 21.823

    2011 day 120 21.336
    2012 day 120 21.568
    2013 day 120 21.273

    2011 30 day decrease 0.232
    2012 30 day decrease 0.097
    2013 30 day decrease 0.339

    2011 Drop max to day 120 0.625 over 17 days
    2012 Drop max to day 120 0.355 over 6 days
    2013 Drop max to day 120 0.55 over 13 days

    So the maximum is slightly less than last two years and the decline since the peak is pretty rapid.
    Now 0.063 below 2011 and 0.295 below 2012.

    https://sites.google.com/site/arctischepinguin/home/piomas/grf/piomas-trnd3.png

    http://psc.apl.washington.edu/wordpress/wp-content/uploads/schweiger/ice_volume/Bpiomas_plot_daily_heff.2sst.png

    Volume and thickness at record lows but only just.

  7. #7 crandles
    2013/05/04

    Some betting links since you don’t seem to be getting around to providing the links:

    http://scienceblogs.com/stoat/2012/11/26/sea-ice-end-of-season-report/

    points to

    http://scienceblogs.com/stoat/2011/06/26/betting-on-sea-ice-10000/

    which details the $10,000 bet but points to a dead link for where the bets were accepted.
    The link is really
    http://scienceblogs.com/stoat/2011/05/10/sea-ice-part-2/

    [Not my fault guv, that was SB's migration of blogging platform, it broke old links. Fixed -W]

    4 bets with crandles increased to £100 at
    http://scienceblogs.com/stoat/2011/06/26/betting-on-sea-ice-10000/#comment-17275

    Was it a request for a graphic (showing time evolution of snow and ice depths following instantaneous removal) from your paper that started this links qpost? Any chance of emailing me the paper?

    [At the moment, no, I can't - my old disk has just died. I'll need to find my backups... -W]