UPDATE: 8AM PST 9/13/10 JAXA has updated with their final Sept 12th data, up for the second straight day there’s been a gain:
The latest value : The latest value : 5,005,000 km2 (September 12, 2010 final data)
While the vagaries of wind and weather can still produce an about-face, indications are that the 2010 Arctic sea ice melt season may have turned the corner, earlier than last year.
In the JAXA data, there was a gain of 33,593 km2 in a single day on 9/11/10 and another gain of 18, 594 km2 on 09/12/10 (final data):
09,08,2010,4989375 09,09,2010,4972656 09,10,2010,4952813 09,11,2010,4986406 09,12,2010,5005000
Last year, when I correctly called the turn, it was September 14th:
Arctic sea ice melt appears to have turned the corner for 2009
I wrote:
That is a gain of almost 26,719 km2 from the Sept 13th value of 5, 249, 844 km2 which may very well turn out to be the minimum extent for 2009.
And it is not just the JAXA plot that indicates a turn the corner bump for 2010. The DMI 30% extent graph is showing a very sharp uptick.
Here is the relevant area zoomed and annotated:
ADDENDUM: Last year’s DMI graph about this time had similarly abrupt uptick:
Temperatures at 80°N and above are now dropping quickly, after some delay:
The annotations are mine, the current temperature is approximately -5.5° C. I say approximately, as DMI doesn’t make the data available here, only the graphical output, so I’ve had to draw a line and estimate based on the coarse scale they provide. Seawater freezes at a temperature of -1.9° C (source here) but varies with salinity. Call it -2° C, but clearly now air temperatures are cold enough above 80°N to expect some refreezing.
The NSIDC Arctic extent plot shows the beginning of a flattening, but since their smoothing algorithm adds a reporting delay, we won’t see the turn (if it holds) until about two days from now.
NSIDC Arctic Sea Ice Extent – 15% or greater – click to enlarge
If it is indeed the turn, then Arctic Sea Ice minimum for 2010 will end up at 4,952,813 km2
I may make a follow up post and have a look at all the forecast players mid to late week if the turn is confirmed. Of course my forecast has been proven incorrect already, but then, so have others.
Polar weather forecasts suggest colder weather ahead, and historically, the timing is right for a turn.
One such indicator is the Arctic Oscillation, shown below:
Source, NOAA Climate Prediction Center:
http://www.cpc.noaa.gov/products/precip/CWlink/daily_ao_index/ao.sprd2.gif
The forecast shows a deepening AO in the next few days, which traditionally means colder temperatures and a refreeze.
So, we’ll watch and wait, and I’ll update if the turn is confirmed.
fishnski says:
September 16, 2010 at 2:23 pm
From my amaturist prospect I’m seeing the Real climb starting mon/tue of next week..Are we having fun yet??
_____________________________________________________________
Using a 15-day moving average least squares fit shows 2005 and 2007 as the last to rise out of the well, as it were. Both occured right around the dates you suggest, September 20-21, so unless something highly unusual happens weather-wise, I too would expect to see some major movement upwards starting in that same time frame. Ice area is already on the upswing.
Preliminary JAXA extent shows another brutal loss. New preliminary extent is 4892813 km^2. Not enough to discuss it now though.
-Scott
Preliminary for the 16th = 4892813, a further loss of 55,625.
There has been a brutal High presure system sitting over AK combined with a low over on the Russian/Siberian side that has been pumping in warm air thru the bering straits & off a record heated AK for awhile now & thats been the problem.
Like I posted earlier,That weather set up will ease off by the end of this weekend & is why I thought that we could get back into Ice growing by mon/tue….its gotta stop (crying,bummed out funny face)!!…Oh..the main problem is the fact that Scott Jinxed the Comeback!(angry/Smiling funnyface)
AJB says:
Preliminary for the 16th = 4892813, a further loss of 55,625.
Looks like Bastardi nailed it.
The Arctic is playing late season tricks here and those who called the low earlier this week will need a bit of revision in their outlook. I suppose I should pop over to Steve Goddard’s site to see how he is spinning this…but on second thought, nope, I can pretty well figure that one. It’s all the strange winds, high pressure, etc. etc. etc. Never of course perhaps realizing that pressure system and the differential winds that come form them are just another form of energy here on Earth, and AGW is of course ultimately about an energy imbalance. It doesn’t have to show up just as temperature all the time.
I think this season shows why, at some point in the not too distant future, we could actually see the “summer” low not come in until early October even. More open water equals more residual heat equals a more prolonged melt season.
2008’s low doesn’t currently seem to be in jeopardy of falling to 2010, though this last little bump down has made things a bit interesting. Looks like the folks behind PIOMAS, with the mid-summer prediction of 4.7-4.9 million sq. km. should be congratulated this year…
The metric used here is extent via JAXA, so the losses in JAXA definitely count for the evaluation here.
But if anyone is wondering, CT showed another area gain yesterday (16324 km^2) and even an extent gain (11280 km^2). Anyone know why CT’s extent has shown 6 days in a row of gain while JAXA’s last 3 days are losses?
Just more of the numbers intrigue…I love it, it makes for great discussion!
-Scott
Confirmed JAXA 15% extent for Sept 16th: 4890938. New low.
15-day: http://img375.imageshack.us/img375/3284/15day20100916.png
7-day: http://img530.imageshack.us/img530/7746/7day20100916.png
Lobe at 170E still vulnerable. IMHO, expect further substantial losses yet.
Scott says:
September 17, 2010 at 8:57 am
Just more of the numbers intrigue…I love it, it makes for great discussion!
Chocolate pipe wrenches are best eaten, not used as micrometers 🙂 CT’s animation implies major compaction is taking place, leaving the large lobe at 170E increasingly vulnerable to what I assume are still warmish currents in that region. Anyone know where chocolate SST and current indicators are on offer? Plain chocolate preferred.
I read somewhere that it is better for the Ice to freeze slowly as in opposite of a flash freeze..any truth to that?..+ can all this warmth coming into the arctic caused by the weather set up I mentioned in my last post help or hurt the long term ice health..I mean, is there a burning off of heat right now that will help make it easier for the SST to become User friendly for ice formation down the road?..Inquiring minds want to know..
Scott says : CT showed … even an extent gain (11280 km^2).
Can you tell us where Cryosphere hides its Extent Numbers ? All I get is an Archive that ends before 2010.
DOWN: DMI*, NSIDC, JAXA*, NATICE*, Nansen Area (* 3 new Lows).
Even: Bremen & ROOS-Extent extent even.
UP Cryo Area (4th straight day, yet Nansen down last 3). Both above 2010 Low (by 112, 72K ).
TOPAZ Forecast Page, shows all 4 types: AMSR/SSMI for Area/Extent: http://arctic-roos.org/forecasting-services/topaz/topaz-model-forecast
>> Both Areas Up. … both Extents Down
Total: Extents: 6 DOWN, 2 tie
Total Areas: 3 UP, 1 down.
What we are seeing is the Sea Surface temp fighting the Air temp. And we are talking a 20-to-50 degree fight ! (Fahrenheit).
Cold air is causing 1/10th-Inch ice to form where SOME ice is keeping the surface Cold – – thus converting 20% to 100% which affects Area, but not Extent. Meanwhile, High Winds break thin stuff up, while Behring Strait Water has hit & is melting the “arm” stretching towards East Siberia. And only 3 days of Sun left.
Preliminary JAXA 15% for 17th = 4,832,813, a further loss of 58,125. New low.
As AJB just said, 4832813 km^2 for a new low…rapidly approaching 2008 at this pace. This is getting really interesting. Looking at CT’s animation, the “core” is really getting solid but many of the edges are being compacted while the lobe/peninsula has turned into the island and is melting away.
While I was skeptical of having reached the minimum, I didn’t expect it to go below 4.85e6 km^2 at all…and I really thought the peninsula/lobe was going to pull through. It looks like the few holes in the “main” section of remaining ice have all gotten above 15% concentration and are now added into extent, so the remaining losses at the edges are no longer being counteracted by this. At least that’s my hypothesis. NSIDC called the minimum with the safety net of having a moving average to catch them…but even they may get hammered by this activity. It’s like 2005 all over again (at least in terms of extent).
-Scott
Scott says:
September 17, 2010 at 8:34 pm
As AJB just said, 4832813 km^2 for a new low…rapidly approaching 2008 at this pace. This is getting really interesting. Looking at CT’s animation, the “core” is really getting solid but many of the edges are being compacted while the lobe/peninsula has turned into the island and is melting away.
While I was skeptical of having reached the minimum, I didn’t expect it to go below 4.85e6 km^2 at all…and I really thought the peninsula/lobe was going to pull through. It looks like the few holes in the “main” section of remaining ice have all gotten above 15% concentration and are now added into extent, so the remaining losses at the edges are no longer being counteracted by this. At least that’s my hypothesis. NSIDC called the minimum with the safety net of having a moving average to catch them…but even they may get hammered by this activity. It’s like 2005 all over again (at least in terms of extent).
-Scott
_____________________________________________________________
Scott,
I’ll ask the same question CW asked, where did you get the CT extent number(s) from, all I can find is their daily time series of ice area.
If NSIDC drops by two pixels (daily chart) it will set a new low of ~4.75E6 km^2, makes you wonder if NSIDC will eat crow if this happens and announce a new minimum. As CW states NSIDE droped slightly from the previous day. Also, I’ve been reading their chart daily and plotting the September time series, and as of late, I have to go back through the last 3-4 dailies and check/change them accordingly, and all these have been changing slightly. Kind of makes me wonder about that 5-day moving average, right now it seems more like a 7 or 9 day moving average.
About that lobe, that seems to be shrinking rapidly, it would appear that that area is a proof of concept of ice still melting vs compaction alone. It appears to be shrinking much faster than compacting, which suggests melting as the current dominant process. So maybe the compaction argument alone at this time of year (no further melting) can be shot down as a general argument for explaining extreme minimums.
I agree Scott it is very interesting. I said to Julienne over on Steve’s blog they may have called it too early. If there is another drop tomorrow their average may not help!
However I think the winds have changed in the last 12 hours so I am not convinced we will have too much of a drop tomorrow, perhaps even a rise. I can imagine being surprised again though.
Andy
R. Gates says:
September 17, 2010 at 8:44 am
The Arctic is playing late season tricks here and those who called the low earlier this week will need a bit of revision in their outlook. I suppose I should pop over to Steve Goddard’s site to see how he is spinning this…but on second thought, nope, I can pretty well figure that one.
=======================
Ah, there’s no need to pop over there, R. You can “figure out” how to spin it yourself, as you always do.
A true scientist would just be trying to gather data….not force conclusions.
Chris
Norfolk, VA, USA
OMG OMG OMG….chicken little the sky is falling…
http://ocean.dmi.dk/arctic/meant80n.uk.php
I guess no one knows the answer to my Flash freeze Question..Anyway..The best I can tell is that “the Lobe” is within an area with SST & air temps no higher than 32 with the stem in the Freeze..
Okay, this is embarrassing…
Everyone should disregard everything I’ve said about CT’s extent the last few days. I was looking at the area average value (which is close to extent right now) and got confused and thought it was extent. I certainly didn’t always think that…I used to just ignore that column. Don’t know when I got confused here, but clearly I’ve totally screwed it up the last few days. Shows how little I know what I’m talking about.
On a less embarrassing note, area was up again with yesterday’s value~24000 km^2 again. Assuming I’m looking at the right column of course.
-Scott
fishnski says:
September 17, 2010 at 2:26 pm
fishnski,
I’m no expert (and may have just lost all credibility here with my extent vs avg area blunder), but I think this is more theoretical than anything. In a beaker, what you say is true – slow crystal formation provides superior solid material because impurities are excluded (they’re unfavorable thermodynamically and typically only included to a large extent kinetically). I would assume this would be especially true with salt water.
However, in practice I don’t know if this is a reality with Arctic ice. With winds and ocean currents keeping things well mixed, I would assume the ice is always loaded with impurities. It’s just a matter of scale though…comparing purity on the microscopic level vs macroscopic isn’t always straightforward.
Two examples I can think of where melt and refreeze can actually help the final product are:
1. “Rotten ice”. I’m not discussing here how much of it is up there or if it’s more than in the past, but “holey” ice with lots of air pockets should (a) melt faster with respect to volume and (b) provide more insulation for the refreeze than similar-mass dense ice in the winter.
2. “Dirty ice”. If the ice was loaded with dirt/soot/ash, it’s capable of absorbing much more radiation than normal. A complete melt (vs, say, losing 90% of the thickness in the melt season) might be better here because the contaminant will then sink into the water and the new ice will be a superior product.
Just my thoughts,
-Scott
Scott,
Thanks for clearing up the confusion, but I can’t tell you how hard I tried looking (last night) for the CT/UIUC dataset for daily Arctic sea ice extent.
The best I’ve been able to find is NSIDC’s FTP site for daily sea ice extents for 1972-2007 inclusive, but no dailies for 2008, 2009, or 2010;
ftp://sidads.colorado.edu/pub/DATASETS/seaice/polar-stereo/trends-climatologies/
This is one area where I’d appreciate more transparancy from NSIDC, similar to JAXA extent dailies (although even though shows daily area charts, they do not post a file of daily areas circa 2002-2010).
The same is true of Zhang’s PIOMAS model numbers (anomaly and total ice volumes).
I won’t go into the details, but both area and volume (even model numbers) would greatly aid a forecasting model of Arctic sea ice extent. And I plan on using both even if I have to reverse engineer Zhang’s chart for the JAXA era (right now I’m mixing CT/UIUC areas (y-axis) with JAXA extents (x-axis), with some very interesting results (2002-2006 vs 2007-2009 vs 2010).
Pamela Gray said:
“So Andy, you should have said, “a southWARD direction would mean the wind is heading south and so therefore is a northerly”. According to what I have learned from you and from my own investigation, you cannot have used the word southerly in your sentence. I guess we were both wrong and must remember to suffer the little suffixes.”
———–
Sorry Pam, you’re still muddled.
To go in a southerly direction means to go sort of towards the south. It’s only when the noun qualified by the adjective ‘southerly’ is ‘wind’ that the logic goes south.
Incidentally, the suffix is -erly, not -ly.
Thanks for the reply scott & don’t sweat the minor goof…Your posts are one of the 1st that I look for. I found this below….
“When surface temperature increases, the upper ocean warms and ice growth decreases. This leads to a decrease in salt rejection from new ice. The salinity of the upper ocean falls. Lower salinity and warmer water results in lower water density in the upper ocean. With fresher, less dense upper water, there is now increased stratification of ocean layers which weakens convective overturning. Less ocean heat is transported upwards. This leads to a decrease in ice melting from ocean heat. Hence we observe an increase in net ice production – sea ice increases.
While all that is a bit of a mouthful, it’s actually a simplification of the process as there are various feedbacks along the process. Warming air increases upper ocean temperature which affects air temperature through air-sea interactions. Warming temperature leads to increased precipitation which increases sea ice growth. More sea ice means less atmospheric heat can penetrate waters.”
GO FIGURE…did mr. gates write that?
Does anybody remember putting salt & ice together in the container that surrounded the “cream” in the ice cream maker…the brine mixture was better in drawing heat from the cream making it turn to ice cream faster. Gotta work on sea ice too..maybe a slow cooling helps in the Brine effect (the removal of salt from the water in the ice formation process)…my head hurts..I need a beer!….going to ck my crab pots..see y’all!
NSIDC has a new low as of today’s chart ~4.69E6 km^2 (17th) vs 4.76E6 km^2 (10th).
If that lobe at ~165E vanishes completely (thins in width or seperates completely), expect to see a very late date of minimum and an extent below 2008’s.
Looking at CT/JAXA/NSIDC/Bremen dailies (in sequence, high resolution CT’s saved to HD daily)) that lobe is indeed shrinking as well as other perimeter areas of low concentration (~120E through 225E).
In other words, I would expect to see a few more days of sea ice extent losses.
Meanwhile area continues to grow, primarily in areas that were/are already above 30% concentrations (DMI).
Confirmed JAXA 15% extent for Sept 17th: 4842031. New low.
15-day: http://img412.imageshack.us/img412/676/15day20100917.png
7-day: http://img412.imageshack.us/img412/9127/7day20100917.png
Seems likely to track 2005 but may go deeper yet. Will be interesting to watch the day to day variance from here on, unlike 2005 we may see violent swings.