the Air Vent

Because the world needs another opinion

Sea Ice 2012

Posted by Jeff Id on January 28, 2012

I have reworked my sea ice code to account for leap years and to make it easier to read.  It wasn’t a terribly easy process but it was useful.   Here I will present some plots of sea ice trend as derived from the gridded satellite data.  The purpose of this was to verify their accuracy and lay groundwork for future posts on sea ice.

Unfortunate statistically significant growth of ice during unprecedented death-spiral sea ice doom.

Unfortunate statistically significant loss of ice during unprecedented death-spiral sea ice doom.

All of the graphs here have been taken from the NSIDC gridded data.  Of all the government funded global warming groups I have corresponded with, these people are the best.  The graphs shown are daily data with some linear interpolation for the earliest decade where data was missing.     AR corrected c0nfidence intervals are reasonable but are getting close to 1 so don’t interpret them too closely.  They also assume a normal distribution – which probably isn’t that bad of an assumption.  Bottom line – look at them as an approximation. – – more people should write that.

Now for my signature sea ice plot.  The global sea ice with an offset based on the average area of sea ice.  The purpose of this plot is to explain that there is a heck of a lot of sea ice left on this “little blue marble” and that those who panic are in their own special class.

Ok, to be honest, the other purpose is to drive advocates nuts with the very data they promote.

I will present the code for these plots soon.  It is fairly complex, definitely has a C accent that would drive Mosher crazy, but does the job more accurately than anything I have previously produced.

21 Responses to “Sea Ice 2012”

  1. anon2nz said

    It’s fantastic to see you getting back into the data and analysis. It is where the real action is … and much more productive than the politics.

  2. Gary said

    Jeff, to be specific, this is based on grid cells of what area (in degrees) with 15% (?) or more sea ice?

  3. Jeff Condon said

    Grid cells are 25km and the cutoff was set to 10%. FYI, it was calculated as area not extent. Good question.

  4. steve fitzpatrick said


    Interesting post. There is clearly no reason that the Antarctic would be gaining ice while the Arctic is losing ice… unless there are cyclical processes which are larger in magnitude than any underlying trend due to GHG forcing (which is of course uniform everywhere).

    Still, reason pretty much demands that warming makes the extent, and especially the summer minimum, of sea ice lower. The recent rapid drop in arctic sea ice is probably mostly the result of some kind of north/south seesaw, and that seesaw may well reverse in the next decade or two. That doesn’t mean that the overall (global) trend is not being driven slowly downward by warming, Open water both absorbs more solar energy and loses heat in the autumn more quickly. But since the ice minimum occurs just as the sun sets on the respective pole for the winter, the additional absorption of solar energy is clearly not huge due to the extra open water; both polar oceans ice over shortly after the solar equinox.

    Like much in climate science, Arctic sea ice is probably overblown in importance (and I am sure the polar bears will survive). That doesn’t mean the downward global trend in sea ice is not caused by GHG forcing.

  5. ChrisM said

    Yes the accelerating death spiral isn’t that obvious, is it? I can’t see the Arctic reaching zero in 2013 or 2020 or whenever they said it would.
    And to be the spelling pedant, there is a missing “c” in both Arctic and Antarctic graph titles. It will be the type of trivial thing that people use to rubbish the results. I hope I haven’t done a Muphry in this post.
    One of the interesting cross plots would be to do the anomaly vrs the G!SS arctic temperatures. It appears you have done at least part of that already with your slope corrections. I suspect the correlation wouldn’t be that good, probably only Mannian at best.

  6. HaroldW said

    You might want to correct the units of the slope estimates on your chart. A trend of 141000 deg C/decade (as on the first chart) would definitely be cause for concern!


  7. Brian H said

    Someone observed that when the warm winds blow across the Arctic, they’re still freezing cold, and open up surface water while they pile the ice on the far shores. This maximizes the creation of new ice.

    Just another inconvenient negative feedback. The damn planet is riddled with ’em!

  8. Layman Lurker said

    Here is the Arctic sea ice plotted alongside the AMO. It is a negative correlation, so I flipped the sign of the AMO before plotting. I also standardized the sea ice time series wrt the mean and sd of the AMO. Interestingly, the best fit was obtained by lagging the AMO values by 2 months (I had speculated that sea ice values would lag the AMO). The correlation is r=0.58. While there is a slight trend in the residuals this might be explained by either the highly autocorrelated residuals (AR(1) = 0.7914) or the detrending of North Atlantic sst’s done to derive the AMO series.

  9. Cary boyce said

    Antarctic not antartic.

  10. Jeff Condon said

    Sorry for the spelling and units guys. I spent days making sure the math was good and wasn’t too worried about the plots. I will get it corrected for the next post.

  11. TerryMN said

    Not speaking for Mosher, but – C accent is ok. External C functions, not so much.
    (unless, of course, you’re on the same exact platform as I am so I don’t have to relink 🙂 ).

    Question pulled forward from the ice animations – does the data you’re plotting include the spurious Great Lakes mid-summer ice? Just curious.

  12. Jeff Condon said

    #11, Yes it does. That is actually why I’ve put so much effort into these trends. I want to look at ice outside of the Arctic.

  13. TerryMN said

    Thanks – I’m anxious to see the code, now that I’m using R on a regular basis. Wish I’d have taken a stats class or two with all of my CS classes back in the day…

  14. Carrick said


    Interesting post. There is clearly no reason that the Antarctic would be gaining ice while the Arctic is losing ice

    Well there is a reason of sorts…temperatures in the Antarctica have been generally cooling for the last 30 years or so, and the amount of ice seems to track with that…

    There does seem to be a long-term out-of-phase relationship between Antarctica and the rest of the planet. Prior to circa 1980, it was warming (slightly) will temperatures on the rest of the world were plummeting.

    I’ve assumed it’s just related to the large amount of isolation on the Antarctica continent, but perhaps that’s not realistic.

  15. steve fitzpatrick said


    It makes sense that Antarctica is colder (in part) because it is more isolated from intrusions of warm air by the cold southern ocean. But I don’t see how that could explain why 3 watts per square meter of GHG forcing doesn’t raise the temperature in Antarctica by a fair amount. Evan absent amplification, 3 watts ought to raise the surface temperature by more than 3C, and there is no evidence of anything like that happening. Maybe there is a good explanation for what is offsetting 3 watts of GHG forcing (flat temperatures everywhere except the peninsula), but I haven’t heard it.

  16. Carrick said

    Steve, I agree they don’t have a good explanation for why Antarctica has been cooling since roughly 1980—I was just pointing out that the ice gain is consistent with this. The real anomalous part is the decrease in temperature. Perhaps it’s part of that unmodeled 60-year oscillation.

  17. steve fitzpatrick said

    “Perhaps it’s part of that unmodeled 60-year oscillation”
    Yes, a significant 60 year oscillation with a big “see-saw” north /south component could explain the lack of recent warming in Antarctica. The problem is that models/modelers explain much of the apparent 60-year oscillation as a consequence of volcanoes in the early 1900’s and in the late 1900’s. If that is true, it is hard to see where a N/S oscillation comes from.

    If you know of a plausible explanation, let me know.

  18. Carrick said

    Steve, I don’t think volcanos is a good explanation here. This peak also shows up in 1400+ year long ice core temperature proxies.

    IMO it probably is related to a real atmospheric-ocean coupled oscillation.

  19. Richard said

    I think the most surprising thing is that no one is commenting on the fact that there is more and more ice being produced each year at the Poles. show quite clearly that the winter ice in the Arctic is nearly the same over whole the record period with a small droop over the last few years, The summer ice has shown a vast decrease over the same time which is what everybody is concentrating on. But I fell that the important fact is that the winter ice returns each year to nearly the same figure as previously.

    This makes the amount of ice produced each year now significantly different (and larger) from when the record started.

    11-15.5Km sq range in 1900, 6-15Km sq range in 2011. So 4.5Km sq of ice produced in 1900, 9Km sq produce last year. A 50% increase in ice production!

    So where did all the brine at -1C to -3C from all that extra sea water freezing go? When will it return to the surface from the bottom of the oceans and what will the effect on global temperatures be when it does?

  20. Richard said

    OK. A 100% increase – 1900 was at 50% of this years figure. You figure out which one you want to quote.

  21. Brian H said

    Re: Richard (Jan 31 14:01),
    Fine observation. A graph of ice rebuilt each year, IOW, would show a steady climb.

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 )

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

%d bloggers like this: