People have been posting comments about the sea ice again. We're close to the min, though maybe not there yet, but from http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm, which I think is the best pic available, it looks good for the good guys.
The original in this series is Betting on sea ice?
Yeah, it seems to have leveled out. BTW, do you know how to get the actual data? I found it one time, but my searches since then have been a dismal failure. I want to see what the +/- amounts have been in September so far.
Whoops! Um, thanks for finding me the link to the data that I've been searching for and not finding, even before I ask for it!
I've been noticing these posts for months, but I must have missed the very first one when you define who "the good guys" are, and what is the "good" that I am supposed to see in that graph.
That sea ice extent is pretty similar to last year? (good that we haven't reached climate disaster yet and can now make fun of Algore's doomsday predictions?)
That it is better than last year? (good that climate is improving and we can make fun of Algore because he's so totally wrong about everything?)
That sea ice is at almost the minimum it has ever been? (good that maybe soon we can go drill baby drill up there... and make fun of Algore for being such a sissy?)
Really, what are you trying to say? Speak, man, speak! Or blog, or whatever.
William defines "good guys" as so that he is one of them, of course.
Shows that W might still might have a few percent of a chance of losing. Might want to point out that sea ice area is closer than extent, as well, and other source for sea ice extent give a closer value that JAXA does. Like this one, for example:
W probably does have this bet won. Better not spend it quite yet.
Seaice: looking good
uh, yeah, a good deal lower in extent than the 2005 minimum. I'm sure you're happy to win some bets (congrats), and we're all relieved that it hasn't been as bad as 2007. But if this is 'looking good', we should be alarmed.
[Well you have to interpret words within context, of course -W]
Well you have to interpret words within context, of course
Is that a dragon egg or the remains of Humpty Dumpty you're emerging from?
"On September 12, 2008 sea ice extent dropped to 4.52 million square kilometers (1.74 million square miles). This appears to have been the lowest point of the year, as sea has now begun its annual cycle of growth in response to autumn cooling.
The 2008 minimum is the second-lowest recorded since 1979, and is 2.24 million square kilometers (0.86 million square miles) below the 1979 to 2000 average minimum."
[Ha ha, still I think they are being a trifle premature. I'll wait another week before declaring victory -W]
The metric for this round of bets is Cryosphere Today, and they are a lot closer to last year than the NSIDC. Having said that, the CT area bottomed out at 3.004 a couple of days ago and has increased since. Last year was 2.92. But I agree with WIlliam, still need to wait a couple of weeks...
[Not wishing to start a war, but who said what the metric was? -W]
>"[Not wishing to start a war, but who said what the metric was? -W]"
Um. Take a look at your linked first post. It includes:
"And we use the record as presented by cryosphere today."
So far cryosphere today has done graphs etc on area. I would be surprised if they suddenly switched to extent or volume or something to announce a record low.
[Ah. Jolly good. I'm glad someone is paying attention -W]
Re "they are being a trifle premature"
How did NSIDC arctic sea ice news
on 16 September claim that
"However, we have now seen five days of gains in extent." ?
In the F15 data linked by Clarence (see realclimate comments #478 and #448) which seems to agree to the 4.52 minimum announced, there only appear to be 3 days of increases after minimum on 12 September (only 1 if you count last data which reverses a couple of the gains). If you count increases before the minimum on 12th there are at least 6 daily increases. If you use the 5 day average there is only one increase.
The only way I see of getting an answer of 5 from the F15 data is to count daily increases in the last 11 days which would seem a bit bizare.
JAXA data has 4 consecutive increases before the minimum in 2007 so I doubt that can be all that unusual. So I would have thought you would want at least 6 increases that have not reversed before calling a minimum which was earlier than the middle of September.
Is the pressure to be first to call the minimum that high?
[I think there is an unhealthly pressure to be the first to say these things. Look at the way years/months are declared records before they have even finished! -W]
William, you have been very critical of 5 - 10 year estimates of an ice free summer. Clarence on RealClimate post #491 has posted some volume data. Extracting some info and doing a few simple calculations I get:
Anomolous drop to 14308 6 days later. Of this 2570 drop, 1300 reappeared within 10 days. Rest may take much longer...
Fall in max670
Fall in min148Not clear this even is a fall but would be too low for a trend figure anyway due to anomalous nature of 2007.
Longer term changes may be better than comparing 2008&2007.
IPCC Table4.1 1998 figs??
10y? Fall in max8122
10y? Fall in min14623
Assuming lower of these figures to be more representative:
Different figures, different methodologies etc makes it dubious to assume the trend figures extracted are reasonable. Also don't think a straight line extrapolation is terribly sensible. Many other cautionary note are probably appropriate.
Nevertheless the figures derived do suggest 5-10 years may not be all that unreasonable and does not require any significant volume decline increase.
Does this surprise you at all? Any comments?
[Not perfectly sure what those numbers are. Be aware that volume isn't well measured. Area or extent is, but thickness is guesswork to some extent. Certainly one year trends are a bad idea; 10 year is better. You're calculating a volume trand over 10 years from IPCC AR4 data? -W]
>"volume isn't well measured. Area or extent is, but thickness is guesswork to some extent"
Yes much of the figures at http://seaice.bplaced.net/thickness/
linked by Clarence and labeled "GDAS ice thickness data (work in progress)" is modeled data that only gets corrected when there is no ice present. See post #491 of
There's some ice volume data in IPCC AR4 WG1 4.4, but the time series of sea ice volume in the Arctic Basin is also just model-based (however a better model than the simple GDAS/GFS ice model) and ends in 1998. Table 4.1 says 19â²000 kmÂ³ at the annual minimum and 25â²000 kmÂ³ at the maximum. 4.4.1 states that nearly half of the total ice volume of the Arctic is in ridges.
Some GDAS ice data (every 5 days) is now available at http://seaice.bplaced.net/thickness/ . Note the anomalous values on 2008-04-03. Much ice has been destroyed in the model by data failure. Area and extent have been restored by the ice concentration data of the next day, but volume rebuilds slowly in the model, and some is still missing.
IPCC AR4 WG1 Table 4.1 does say 0.019-0.025 million Km^3 for Arctic sea ice. I am not sure that is for 1998. From what Clarence said it could easily be an average for a period ending in 1998.
So I am trying to use IPCC data for a start point and Clarence's data for an end point. I am not sure the data is consistent though are in Km^3. Clarance's data clearly say they are a work in progress. I am not sure of the start date of the start point. To mention but a few issues. As I said: many other cautionary notes are probably appropriate.
So yes loads of problems with these silly calculations. 5.4 years might need to be doubled if the IPCC data is an average of 1979-1998. OTOH I took the 8122Km^3 fall in max rather than the 14623Km^3 fall in min.
I am not expecting accuracy from this - could be out by a factor of 2. That would still leave 5.4*2 = still close to the 5-10 year range. Could be out by a factor of more than 2 particularly if all the errors conspire to point in one direction. (Is that likely with consistency of answer between 6.5 years and 5.4 years?)
[OK, well I think that "volume" is less likely to be modelled by a straight line that area, even if the numbers were correct. Because as you say, a lot of it is in ridges. They could disappear one year but leave the total area nearly unchanged -W]
Thank you for your answer.
>[OK, well I think that "volume" is less likely to be modelled by a straight line that area, even if the numbers were correct. Because as you say, a lot of it is in ridges. They could disappear one year but leave the total area nearly unchanged -W]
>less likely to be straight line - fair enough.
Last sentence seems a little odd. The ridged ice is difficult to melt so I would have expected you to emphasize that it is unlikely to disappear in one year. If it did all go then it would take some time to build back up (if it ever did - reduced mass causing reduced momentum?). You then might only need a year with the right amount of snow in Autumn to insulate and reduce the ice build up in winter and a lot of area would then melt in summer.
[I meant, you could have big volume changes that make little difference to the area, which is more important -W]
An alternative argument you could have presented for non straight line extrapolation of volume: As the volume decreases faster than area, the thickness is reducing. Reducing thickness = less insulation, so more heat lost in winter, so more ice forms in winter. This reduces the rate of volume loss. I wouldn't be at all surprised if this negative feedback effect was greater than the positive feedback effect of ice albedo when the ice gets thin. This could result in the rate of volume loss reducing to be more similar to the rate of area loss as the thickness finds its equilibrium thickness.
[Certainly, as the ice gets thinner the insulation is less. My recollection is that past 1m ice grows very slowly -W]
Hey William, I'm ready to concede on this year's bet... What's your PayPal email?