Sea ice extent update 27: nope, not a fat lady (yet)

I'm regularly writing updates on the current sea ice extent (SIE) as reported by IJIS (a joint effort of the International Arctic Research Center and the Japan Aerospace Exploration Agency) and compare it to the sea ice extents in the period 2006-2009. The IJIS graph is favoured by almost everyone, probably because it looks so nice compared to other graphs (like the one by Arctic ROOS, the University of Bremen and the Danish Meteorological Institute). All the years have a nice colour of their own which makes it easy to eyeball the differences between trends. Most of the betting on minimum SIE is based on the IJIS data. NSIDC has a good explanation of what sea ice extent is in their FAQ.

August 29th 2010

The 2010 melt season keeps stumbling forward. Days of 20-30K decreases (August 23rd, 25th and 26th) are alternating with decent daily decreases of 50K (24th and 28th) and a very decent daily decrease of 78,437 square km (on the 27th).

This means that after having crossed the minimum extent thresholds of all years in the IJIS database prior to 2007, 2010 could be diving under the 2009 minimum extent in two to three days from now, which - with several other (strong) indicators - is making any talk of a continued recovery impossible.

Next up is the 5 million barrier. I don't think 2010 will be taking over 2008, as the latter had a strong run in the past days that has increased its lead over 2010 considerably. But it all depends on the weather and the final date of minimum extent.

The current difference between 2010 and the other years is as follows:

• 2006: -624K (37,697)
• 2007: +618K (57,041)
• 2008: +164K (70,333)
• 2009: -212K (48,654)

The average daily melt for the month of August is between brackets. 2010's average daily melt for August is currently 56,406 square km per day.

If 2010 loses as much sea ice extent as...

• 2006 did after this date it will bottom out at 5.15 million square km.
• 2007 did after this date it will bottom out at 4.87 million square km.
• 2008 did after this date it will bottom out at 4.87 million square km.
• 2009 did after this date it will bottom out at 5.04 million square km.

Here's the IJIS sea ice extent graph:

 

Meanwhile the Cryosphere Today sea ice area anomaly is on the increase. A few decent decreases, combined with the 1979-2008 mean starting to level off significantly, have moved the anomaly number towards 1.5 million square km again. Most of the decreases are showing in the Arctic Basin (that has already dived below last year's minimum for this region). Sea ice area in the East Siberian Sea has been dropping for a few days as well, as has the Canadian Archipelago sea ice area, despite all that multi-year ice from the Arctic Basin that keeps spilling into its straits and channels.

The relatively large drops in sea ice area are also having its effect on our compactness graph that shows the ratio between CT sea ice area and IJIS sea extent. It currently stands at 65.34%:

 

Despite the recent domination of low-pressure areas in the Arctic, a high-pressure area manages to stay put over the Beaufort Sea. In fact, ECMWF is forecasting this high to strengthen in the coming 10 days, combined with some very big cyclones on the other side of the Arctic. If this forecast comes about we will be seeing one hell of a Dipole Anomaly:

1040 mb on one side of the Arctic, and 990 mb on the other... That can't be good for the fractured, ulta-mobile ice.

The first instalment of my End Zone-series on air temperatures will be followed later today by an animated comparison of PIPS ice displacement maps from the final weeks of melt in previous years. I have a feeling that the date of minimum extent is preceded by PIPS ice displacement forecasts with small arrows in them.

I have a hunch the the direction of the arrows isn't as important as their size. Whichever way they're pointing, as long as their big, extent keeps decreasing (fast). We'll see what happened in previous years later today, but for now PIPS is still forecasting big arrows:

TIPS - Other interesting blog posts and news articles concerning the Arctic and its ice:

Chris Mooney has written a fantastic article for NewScientist called Arctic Ice: Less Than Meets the Eye.

Meteorologist Jeff Masters has also written a very good summary of the current situation in the Arctic on the Wunder Blog. Joe Romm quotes this article extensively on ClimateProgress.

Patrick Lockerby is feeling better again and delivers a great piece on Petermann and Jakobshavn Glaciers, and glaciers in general.

Wayne Davidson writes about the apparent contradictions in the Arctic, part of the same process:

The heat of 2010 was so strong it reversed the usual High pressure usually centered North of Alaska, the very reason for the Beaufort Arctic Ocean gyre was nullified by a near constant low pressure in its stead. Even in 2007, equally a warm year, this High pressure stood out as the main reason causing Arctic Ocean all time melt. In principle, a natural High anticyclone exists at the Pole, just like the Earths magnetic poles, it is not situated at the North geographic Pole but North Of Alaska, making perfect sense, usually explained by Hadley cell circulation. However 2010 summer had the reverse Hadley cells at the Pole, or likely a split Ferrel like configuration. Leaving the Sub-Arctic immersed in clear air , causing record heat signatures throughout the Northern Hemisphere, characterized by near static high pressures baking places like Western Russia and the island nation of Japan. The clear air expected due to La-Nina manifested itself there. The outstanding feature of this is the seemingly puzzling flow of warmer air towards colder air Northwards in the Arctic, which is usually not the case, since colder Arctic air is denser, thus is naturally under higher pressure, but the heat generated by static sub-arctic highs maintained descending air, with higher pressure, reversing the usual scenario, causing heat to flow Northwards, naturally causing unstable convection (fueled by high latitude colder upper air) rising at the Pole causing clouds, forced clouds, not the typical Arctic seeded clouds from descending air. In short, quite a remarkable planetary wide role reversal.

Comments

[Artful Dodger]

Top shelf, Neven. I'd also like to share these two observations:

1. 2008 Aug SIE: Recall that NSIDC reports monthly means for SIE, not monthly minimums. So by that metric, Aug 2010 avg SIE to date is 6,035,218 km^2 while Aug 2008 avg SIE for the same days was 6,050,095. This means 2010 is still an average of -14,877 km^2 ahead of 2008.

2. We may now start seeing a systemic undereporting of Sea Ice Area on CT. Recall that the 'data hole' at the North Pole is 300,000 km^2 and is considered to be 100% ice covered (probably true in the past). Now, with the Transpolar Drift carrying areas of known low sea ice concentration across the 'data hole', the Area numbers may be artificially inflated. As a quick guess, 65% of 300,000 means that only ~200 K is ice covered, and the CT number may be up to 100,000 km^2 too high.

[Steve Bloom]

Great work, Neven.

A question: Julienne Stroeve was referring earlier in the year to an area of thick ice in the Beaufort. Looking at the PIPS ice thickness graphic, I see an area of ice no more than 3-meters thick, IIRC left over from the 2007 compaction, located about halfway between the Alaska-Yukon border and the pole. Is that it?

Re the coming extreme dipole, if it really does happen that way it will be a real test of whether this year's ice really is as vulnerable as it's been looking.

Oh, and re the cyclone forecast: Is any of that from Danielle? I don't know how to read those maps, but the timing seems about right.

[S2]

Great stuff as usual, Neven.

This really is a captivating blog - almost addictive. I really don't know how you find the time to post as you do.

Small question - in your table "If 2010 loses as much sea ice extent as..." you have the same expected outcome for the 2007 and 2008 scenarios.
I admit that it's possible, but to three significant figures it seems unlikely that two consecutive years had the same melt in the closing stages of the season.

Is this correct?

[Neven]

Lodger, great observations, especially the one about NSIDC monthly average! BTW, do those cropped IJIS images still show the ice moving towards Fram?

Steve,

A question: Julienne Stroeve was referring earlier in the year to an area of thick ice in the Beaufort. Looking at the PIPS ice thickness graphic, I see an area of ice no more than 3-meters thick, IIRC left over from the 2007 compaction, located about halfway between the Alaska-Yukon border and the pole. Is that it?

I'm not sure if the PIPS ice thickness map is such a great reference (not just because Steve Goddard touts it as the only relevant source of pixel data). For ice displacement PIPS is great, but I believe Topaz is more accurate. Sadly, I don't know how to get up-to-date maps.

What Stroeve was referring to was the multi-year ice that was transported away from the central Arctic Basin, or the coast of Ellesmere Island to be more precise) due to the extreme negative AO this winter. I got this picture from NSIDC. Jon Torrance posted another one:

Oh, and re the cyclone forecast: Is any of that from Danielle? I don't know how to read those maps, but the timing seems about right.

Is Danielle one of those tropical storms? I don't know the answer to that one. Isn't this more of a concern after minimum extent?

S2 wrote:

Small question - in your table "If 2010 loses as much sea ice extent as..." you have the same expected outcome for the 2007 and 2008 scenarios.
I admit that it's possible, but to three significant figures it seems unlikely that two consecutive years had the same melt in the closing stages of the season.

It's a coincidence, I think. According to my spreadsheet 2007 loses 470,313 square km from now to minimum extent (September 24th). 2008 loses 470,468 square km to minimum extent (September 13th). This will change again in the coming days. I think this is a strong indicator - combined with the weather forecast - that 2010 might have a good chance of diving below 5 million square km. Which would be good from a PR point of view. Not that I like PR. In fact I hate it. What I mean: This will kill most of pseudo-skeptic PR.

I really don't know how you find the time to post as you do.

Me neither. :-)

I mostly do it in between work and thus don't sleep much. I'm recuperating during those days when I don't post anything. A few more weeks and the season is done. It's been great so far.

[Phil263]

Great work Neven . Really informative stuff with solid references!

[siili]

Topaz is updated once a week, supposed to be on fridays but they have had many delays lately, but now they have daily forecast maps of temperature, displacement, concentration and thickness untill 2/9, http://topaz.nersc.no/topazVisual/matlab_static_image.php?action=NA_ARC_NWA_Function&file_prefix=ARC&match_date=20100902&depth=0005&variable_name=hice

[Jon Torrance]

Looks like we may, depending on the correction tomorrow morning, have just had our first day of increasing IARC-JAXA extent . Neither the earliest nor the latest such in the record dating back to 2002 and who knows how long it will be between it and the minimum? I was rather wondering whether this, or at least a very small decrease in extent might happen, when I saw the Cryosphere Today images for the 28th and how much the ice on the eastern side of the ice projecting into the East Siberian Sea had spread out/deconcentrated compared to the image from the 27th.

[Phil263]

The first increase in extent came

2006: 24 August (+1,875)
2007: 1 September: +3,907, 2 September was also an increase: +6,093)
2008:3 September : +2,812
2008: 4 September:+8,125

So the first increase came a few days ealy this year.However,in each previous year there was a further decrease between the day of the first increase and the year minimum:

2006: -261,719
2007: -343,282
2008:-219,218
2009:-138,125

[Bfraser]

I notice that the graph doesn't seem to go up on 8/29 and that the total for 8/29 is exactly same as 9/6/2009.

Let's see if the revision shows that this was just a data entry error.

bill

[Phil263]

Bill,

If you look at the dataset, you will find that there are a few extent values that have been measured at different times 5,345,156 was also reported in Sept 2005.
Also a zoom in on the extent curve shows a tiny upward tick at the end,

[Lord Soth]

We always get a small melt; after neven posts saying its not over :)

In early August 2008, we had a day of almost zero melt; after over a 200,000 melt in two days. I wouldn't worry about one day of no melt. I still believe that we will get below 5 million sq km.

[JackTaylor]

Bfraser | August 30, 2010 at 05:53
"Let's see if the revision shows that this was just a data entry error."

A good point Bill.
It's normal for the "Fat Lady" to clear her throat and warm-up the vocal cords.
But since it is so close to the first "up-tick" of previous years, it's within 'range.'
- - - - - - - - - - - - - - - - - - - - - - -
Phil263 | August 30, 2010 at 07:02
"values that have been measured at different times ... Sept 2005."

And, 2005 had TWENTY days of decrease after the first "up-tick" with a daily average
of 15,198 so 20 x 15198 = 303960 also puts the minimum extent within range of being less than 5,000,000.

[Artful Dodger]

Chris Mooney posted a good article on New Scientist Arctic ice: Less than meets the eye

[Neven]

We always get a small melt; after neven posts saying its not over :)

Yes, I'm starting to notice it too. But I said: "IF this forecast comes about..."

CT area just reports a huge drop, BTW: 143K. The anomaly currently stands at 1.550 million square km. Combined with the positive extent number compactness dropped to 62.63%.

This doesn't mean anything. I'm not jinxing, I promise.

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

The daily revision ends up as a modest drop, so no early increase. Given the pixelization of the data it's not surprising that identical values show up at different times, it's expected.

[Artful Dodger]

Uh, "me.yahoo": If the SIE dataset were 'pixelized' as you say, then the Greatest Common Divisor would be greater than 1. However, it is not.

The pixels in the graphs are computed from the data, not the other way around.

[Neven]

I believe the extent number has been revised even further upwards: + 9,844

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

Artful Dodger, the data is compiled from the signal using 6.5x6.5 areas ('pixels'), so the greatest common divisor is 156.25 sq km. Today's value is 34256 pixels of ice area, yesterday's was 34193 pixels, on the 3rd of the month the seemingly improbable 6655000 sq km is 42592 pixels. This means there are only 32 different combinations for the last three digits. HTH

[Bfraser]

Thanks, me.yahoo. I had noticed, then forgotten, that it seemed as though too many values ended in 5000 or 0000. Now I know why....

bill

[Patrice Pustavrh]

me.yahho, I'd be pleased if you can clarify a little bit your post to me to see if I understand you well. If I understood you correctly, satellite measurement uses squares of given size which represent the maximum possible resolution which radar is capable of detecting ?
And given pixel, has this pixel only "black and white" (eg. water/ice) possible values or are there is possibility of measurement of different ice to water ratios within given pixel ?
And, if following is true, is it possible that different signal processing gives different behavior in SIE/SIA graphs due to small variations in processing those reading in very scattered ice like this year is clear example of ?

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

Patrice, basically the pixel has a value between 1 (all ice) and 0 (all water), the extent results use a cut off of 0.15 to define the boundary of the seaice. So the extent of the seaice is the sum of the area of the pixels that have an ice fraction above 0.15, the seaice area is the sum of (area of pixel*ice fraction). There are a couple of papers by Spreen et al. which might be helpful at: http://iup.physik.uni-bremen.de:8084/amsr/amsre.html
The different algorithms can lead to slightly different results, for example IJIS change there algorithm in spring and fall to minimize the effect of melt ponds on the results.

[Patrice Pustavrh]

Thanx, me, you really help me to understand ice measurements better. That definitely clarifies the difference between different ice extent measures and difference in graphs, since you can do some different calculations (and if there are more differences in the processing of a raw signal, this might contribute to some more differences). Now I am more than convinced that we should follow the general trend (which is clear in my opinion) and that any single year variances cannot be sign of ice recovery (and distribution of ice was not even counted in).

[Artful Dodger]

Ah, I see! Your unconventional use of the term 'pixel' (which by convention means picture element) is what threw me. But the math certainly works with a 156.25 sq.km grid element. Thanks for your explanation!

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

Artful Dodger, sorry for the confusion, I was using the shorthand 'pixelization' for the process of dividing the 2-D Microwave intensity field into a rectangular array of grid elements. Thinking about it the correct term is 'rasterization'.

[Peter Ellis]

Note that the common divisor for the extent measurements (156.25 sq km) is not the size of the individual grid elements. Rather, it is the product of the grid size and the granularity of the concentration measurement. Assuming it's measured with a precision of 1%, then each grid element is 15,625 sq km.

[fredt34]

I don't know what it's worth... but today's NSIDC graph seems to be under 2009, as shown on their August 17 post (just my eyechecking). At least this should help stopping the "recovery" claims, as this NSIDC page and graph probably are the most read on this topic. Of course NSIDC computes season's rank from the month's average and doen't really care about records - but most disinformers just don't (want to) know or don't understand the difference.

[Phil263]

An increase of nearly 10,000 sqkm is quite significant for this time of year. Nothing similar was ever recorded in August in Jaxa's records (going back to 2002). I note that Arctic Roos also show a significant uptick. Maybe the fat lady sung early this year?

[Neven]

As this is all my fault, maybe I should jinx it some more?

PIPS arrows are still big, CT area had a century break, 2006 had +21,094 on the 31st of August (and lost almost 200K after that).

Right, that should put an end to this melting season once and for all. :-P

[Neven]

ECMWF agrees: low, low, low.

[JackTaylor]

Neven | August 31, 2010 at 01:11
" 2006 had +21,094 on the 31st of August (and lost almost 200K after that)"
- - - - - - - - - - - - - - - - -
Yes but, keep up the good work.
Also in 2006 on 1st September was an increase over 31st August and then the melt from that HIGH point in September was over 211K to minimum extent, so was there 13 or 14 days of decrease (melt & compaction) in September?

[Lord Soth]

We had another gain of a few hundred km on the 30th. Thats two nights in a row.

However I see the problem. The breakup of the large section of land fast ice in NE Greenland a few days ago have created a huge flow heading south and spreading out.

This is masking the sea ice extent lost in the rest of the Arctic.

This bias should not appear in the sea ice area graphs

I expect several days of minimal sea ice lost, until this ice disperses enough that it drops below the 15% threshold for any given cell in the NE Greenland sea; followed by some sizeable losses as this bias is removed.

[Phil263]

There is also significant amounts of ice spreading north of Svalbard

[FrankD]

@Peter Ellis: Assuming it's measured with a precision of 1%, then each grid element is 15,625 sq km.

I think you'll find its a 25 x 25 km grid (625 sq km - I'm sure I read that somewhere, but I can't see it now). That would makes the significance (156.25 sq km) 1/4 of the grid area. Perhaps this comes from the measurements being averaged over several days. From the IJIS site:
"Averaging period and the update timing of daily data

* In general, sea-ice extent is defined as a temporal average of several days (e.g., five days) in order to eliminate calculation errors due to a lack of data (e.g., for traditional microwave sensors such as SMMR and SSM/I). However, we adopt the average of two days to achieve rapid data release. The wider spatial coverage of AMSR-E enables reducing the data-production period.
* Usually the latest value of daily sea-ice extent is fixed and updated at around 1 p.m. (4 a.m.) JST (UT). Before the value is fixed, we also assign a preliminary value of daily sea-ice extent several times (usually three to four times) as an early report, which is determined without the full two-day observation coverage. (The fixed values of sea-ice extent are determined with the full coverage of observation data.)"

[Peter Ellis]

Actually it's a 12.5km grid, so the granularity really is 156.25 sq km. I forgot that you only need to multiply by the concentration precision when you look at area rather than extent. So, if the concentration is measured to the nearest 1%, then the common divisor of area data values will be 1.5625 sq km.

[FrankD]

*sorted*

:-)

[Artful Dodger]

Good observation, Lord Soth. This is confirmed by the steady rise since mid-August in the Current Greenland Sea Ice Area chart .

Phil is on to the larger Southward displacement which began on Aug 28 with a strong low positioned over the Taymyr Peninsula (135E 75N). Winds pushed the Ice edge ~70 km over the last 2 days along a 940 km front from Spitsbergen to Ushakov Island. This is an increase in SIE of about 65,800 km^2 just from just this area (not including Fram Strait).

Climatology says that expected SIE loss for Aug 28-29 is about -67,000 km^2 (this is the 31-Day Moving Avg for 2006-09. If we use the period 2002-09, then the expected value is -59,200 km^2 for Aug28-29).

IJIS extent reports little net change in SIE for this period in 2010, so the increases in SIE near Franz Josef Land must be offset by near normal losses over the rest of the Arctic.

So, we can conclude that "Melt" continues on pace and when winds abate, SIE decrease likely will return to average values. More to come later on those expected values...

[Kevin McKinney]

Nice discussion--I admit to being taken aback by two days of increase so early (and with the ice looking so vulnerable, concentration-wise.)

On a completely different topic, Oliver and Lancashire had a close encounter (but still only of the "second kind," if I recall that rubric correctly):

I was awoken by Tony saying the water was coming in the tent. He had been woken by something. I looked up and saw the poles of the tent above my feet bulging inwards alarmingly. He now recalls the water sound had been confused in his dream by a low growling - there was a bear pressing on our tent. I grabbed the air horn, Tony grabbed the shotgun and the pressure was relieved on the tent. I poked my head out of the other end and straining to look the other way saw not one but two polar bears not ten yards away. Tony opened the other end of the tent, and poking the shotgun out first, then my camera we watched, hearts pounding as the two bears ambled slowly up the ramp behind the beach, as if in slow motion. Fortunately they seemed to lose interest in red domes that make airhorn noises and proceeded down the beach. I managed one grainy photo in the half light. We watched them disappear and laughed, if not a little nervously.

They've managed to pass Bellot Strait and are about 153 miles from Resolute. (Hey, they're Royal Marines, who better to use Imperial units?)

[Phil263]

CT reports SIA of 3.317 million . 2010 is now in 3rd place for SIA minimum behind 2007 and 2008. As you can see 2008 is not very far below (minimum 3 million).

[Phil263]

CT reports SIA of 3.317 million . 2010 is now in 3rd place for SIA minimum behind 2007 and 2008. As you can see 2008 is not very far below (minimum 3 million).

[David Klein]

After an early posting when I discovered this site, I've been content to just read and learn more about the complexity of arctic ice. I came here via Patrick Lockerby's site and I can only express admiration for his (and Neven)'s informed determination and lateral thinking. I thought I knew a lot, only to discover how little it really was.
My take on the present situation is an ongoing melt masked by spreading. This may well be a recurring phenomenon because as the ice is thinning and more exposed to wave action, the late melt season's lower daily melt figure is easily overtaken by the spreading of thinner and more fragmented ice. This is somewhat in line with quite a few comments here. It will follow the freezing season with interest to see if the maximum will stay on trend. Thanks for a great site, much appreciated.

[Bfraser]

Thinking back upon my suspicion a few days ago about the uptick (which I now am convinced about). The other piece of evidence (in addition to the coincidental number, which is now exaplained), was the lack of an uptick in the graph. It seems as though the JAXA graph is a day behind the data. Notice that the uptick is now clearly visible, but today's very slight decrease is not.

Also, have you seen the low pressure region near the pole?

http://www.uni-koeln.de/math-nat-fak/geomet/meteo/winfos/arcisoTTPPWW.gif

That's some cyclone!

bill

[Artful Dodger]

The low shown at 35 W 87 N is probably an anomaly in the data, it happens some times. Look again for the 18 hrs GMT map.

[murmeli]

In other news, Steve Goddard and WUWT have gone their separate ways.

http://stevengoddard.wordpress.com/

[Christoffer Ladstein]

Finally stumbling into this site, i feel like a miner from Chile finally getting out in the True light, not any more suffocating, walking around blindless!

Though I've been an addicted "fan" to both IJIS and NSIDC (+ Mauna Loa, just for the fun of "how fast will the CO2 concentration grow this year/month"), visiting daily in "peak season", Neven`s site and blog combined really help bring the personal education level to another "dimension". Again folks, thanks a lot, I knew other had my interests, but sure, there ain`t many out there on a daily basis.
In the Holy Grail of Internet on the other hand...

For those not already aware, check the blog of the Polar Explorer Børge Ousland (www.ousland.no) , who is attempting to cross through both the NW and NE Passages in one season in a 31 ft. Trimaran, that is truly some adventure, only to arise thanks to the lessen icecover on the Arctic pole!

[Artful Dodger]

Welcome, Christoffer. Neven has blogged about our Norse heroes Sailing through the ice. You may also wish to pay a daily visit to Neven's Arctic sea ice graphs page.

Now back to some interesting Stats for avid ASI watchers: In 2006, the annual minimum 'Compactness ratio' (Sea ice area / Sea ice extent) occurred on Sep 25 (very late indeed!). Here are the minimums for the last 6 years, with the Date they occurred:

Year:  minA/E  Date:
2005:  70.16%  Aug 19, 2005
2006:  67.20%  Sep 25, 2006
2007:  60.56%  Aug 20, 2007
2008:  60.16%  Aug 14, 2008
2009:  64.41%  Aug 17, 2009
2010:  60.16%  Aug 14, 2008

[Artful Dodger]

Pardon my 'typo' error above: The 2010 Year-to-Date minimum 'Compactness' ratio of 61.98% occurred 2 days ago, on Aug 29, 2010.

[L. Hamilton]

Lodger, are your "compactness ratios" CT area/IJIS extent?

[Artful Dodger]

Yes, Larry.

[Kevin McKinney]

So Goddard has gone solo. . . maybe he and Watts had creative differences?

[Christoffer Ladstein]

A. Dodger, thanks for the "summary" page of Neven!
The most intriguing found there, also pointed out in Goddard's blog, is the "ironic" fact that July/August have proved really cold above the 80'th degrees; and still we are about to witness another seaice-minimum on the very low side of modern registrations. Just imagine then what we would have witnessed during a HOT northern summer (last year also proved cold, though!).
Interesting soon will be what percentage of 1-2-or multiyear-ice will exist in the end of September?!

[Christoffer Ladstein]

Last year proved a sea-temperature-anomaly of 5-7 C above normal in the Baffin bay, late summer/early autumn.
Do we have a chart covering temperature in the sea?

I suppose an anomaly on the upper side will inflict more to a late minimum ice cover than air-temperature. At least this may explain why the minimum in 2010 is competing for a "top 3" position, regardless cold summer (air) temperatures?
Of course i am aware of other explanations, as an even before fragile & young icecover, never able to recover, due to the situation the last 4-5 years...

[Artful Dodger]

Christoffer asked: "Do we have a chart covering temperature in the sea?"

Visit Neven's Arctic sea ice graphs page (see my link posted above at 22:20).
Click the images to visit the source page.

[Jon Torrance]

Trying my hand at very short range prediction, looking at the Institute of Oceanography, University of Hamburg's latest near real time concentration map compared to the one from a day ago, I expect the IARC-JAXA extent is going to go up slightly again for August 31st. My impression is there's more ice spreading out into lower concentration but not below 15% along the ice edge from the Beaufort to the East Siberian Sea than there is retreat of the ice pack edge anywhere. We'll see how may eyeballs and the processor behind them are doing in just over 40 minutes.

[L. Hamilton]

Jon, I admire the spirit of your short range prediction! You had me convinced but apparently not IARC-JAXA, at least until tomorrow.

[Jon Torrance]

Barring implausibly large adjustments tomorrow, so it would seem. I can't say I'm unhappy - I do, after all, have money riding at Intrade on just more than another 70,000 square kms of decline from the new value - but it will now be interesting to try and figure out where I missed ice disappearing.

[Kevin McKinney]

Yes, a decrease of around 28K. We'll see how it holds up, of course.

Cristoffer, my take on it is that the "cool summer" has been much exaggerated. You realize that the DMI data is for north of the 80th parallel--but most of the melting happens south of that parallel. Neither the Canadian Archipelago, Greenland, nor Siberia has had a cool summer--for the most part, it's been quite the contrary.

I've watched temps in the Canadian Arctic pretty closely this summer, and they have only occasionally descended to normal seasonal levels. There even been have been significant stretches where several of the stations were reporting minima well in excess of the normal *maxima!* And if you look at NCDC data, or the ECMSF website, you'll find that warmth reflected in the maps and summaries they've put out.

By contrast, the DMI anomaly is relatively small (and has reversed late in the season, for what that's worth.)

[toby]

"So Goddard has gone solo. . . maybe he and Watts had creative differences?"

Yes, one was creative, the other was different. Oi!

[FrankD]

Lodger,

Those are some interesting numbers on "compactness". The comparison figures using my read of IJIS area graph (as opposed to CT data):
Year: minA/E Date:
2005: 81.76% Aug 15, 2005
2006: 81.87% Aug 3, 2006 *
2007: 76.52% Aug 21, 2007
2008: 78.62% Aug 15, 2008
2009: 79.67% Aug 17, 2009
2010: 78.60% Aug 30, 2010 *

Mostly they follow the same pattern (although a higher ratio) and reach their nadir on about the same date. The departures:
- 2006, I note a dip in late September also, but not below the August nadir.
- 2010, the big difference there is that the ratio is *still falling*

I've loaded the latest version (to 30 Aug) to here: http://img62.imageshack.us/img62/9341/areaoverextent300810.png

The "compactness" is now the lowest of all for this time of year (which eyeball says is right) and almost the lowest of any year.

As I've observed in connection with this graph previously, when the ratio is falling, the pack is becoming more diffuse, when it is rising, it is becoming more compact. Now if you look at 2007 which lost a lot of extent in September, the ratio rises sharply, meaning that a lot of that loss of extent was the pack "compactifying". But for this year, the losses over August were against the background of an increasingly diffuse pack, meaning any extent loss is melt. In otherwords, real loss of ice, not just rearrangement.

This increases the surface area making it more vulnerable to go on melting. It also makes it easier to export the loose free flowing pack ice. If the weather favours it, there is also a lot of scope in the current pack for "compactification". This triple whammy means we could still see some big loss of extent, IMHO.

This is the only year for which we ahve this sort of data where the pack is still becoming more diffuse at the end of August. Mind is racing with what that portends, but for now I'll settle for "Weird, huh?"

[FrankD]

SEARCH have put out their August ensemble September Sea Ice Extent predictions: http://www.arcus.org/search/seaiceoutlook/2010/august

At first glance, not too many surprises:
- Any changes have been slightly upwards, except Wu and Grumbine, who buck the trend with a downward revision.
- The predictions have converged towards the centre (average slightly higher than July, Std Deviation much lower).
- Wilson has jacked up his prediction from 1 million to 2.5, which is still ...errr... ambitious,
- (amusingly) they've included Goddard, except they've quoted him at 5.1, so he must have revised that earlier (and already passed) 5.5 prediction.
- todays extent of 5.3 means that reality has bypassed five of the teams (range 5.4 - 5.6), with Goddard next in line.

[Phil263]

Even though SEARCH released these predictions today, they must have been made a a couple of weeks ago before the SIE losses reduced considerably. IMHO, on the basis of previous years patterns, it now seems unlikely that SIE will drop below 5 million, which would mean a further drop of 300 k over the next two weeks (?). Something between 5 and 5.2 milionlooks more plausible., and this assumes that melt hasn't already stalled out!

[Phil263]

I have been looking for information about the DATES when the arctic sea ice minimum was reached over the last 30 years. Could anybody point me in the right direction?

[Kevin McKinney]

FrankD, thanks for that comment (and maintaining this useful metric.) It really seems to me to capture the character of this year's melt season: diffusion and fragmentation of the pack. You comment that your mind is "racing" around what this portends; I think that's another sign that your mind is a bit quicker than mine! ;-) But I'm wondering what it portends, too.

I don't think its likely that the melt has/will "stall out" immediately, as Phil puts it. (And, of course, FWIW.) Lodger seems to me to be correct in identifying the relatively more diffuse character of the pack as a vulnerability; we have continued positive SST anomalies over most of the Arctic Ocean; and we have (relatively) warmer temps in the high Arctic.

(On the other hand, the eastern portion of the Archipelago is forecast to come in close to climatic normals this week--and of course, the climatic normals for the first week of September are considerable lower than they are for the heart of the melt season!)

SIE is always tough to call, though. We'll see--and as I write that, I can't help but think how much shorter the timeframe for that has become. So, "we'll soon see."

[Artful Dodger]

Good analysis, Frank. I think the reason our A/E methods produce different results is the way CT vs. IJIS handle the 300,000 km^2 'data hole' at the North Pole. I also agree that loss of actual ice volume (as opposed to compaction) is occurring now. This is no doubt helped by the thinness and wide distribution of the remaining sea ice.

As I've said previously, during 'Bottom Melt' season it is SST's that matter. There also appears to be a significant feedback from water vapor. Near the Healy (N. Beaufort Sea) humidity levels have consistently been at or near 100%. Consequently night time temps remain near their daytime values (water vapor is the most potent green house gas, and is limited by, but can preserve, air temperature).

With sea ice 'compactness' at a seasonal (and near all-time) low, humidity across the Arctic more resembles a Maritime climate. This extra water vapor will have the effect of extending the melt season by delaying the formation of frazile ice and allowing bottom melt to continue. I predicted back in July that this year's SIE minimum would be reached 7-10 days late (Sep 21-24), and offered circumstances where it could be as late as Oct 4.

Since July, I've spent more time analyzing SIE data from the AMSR-E era (June 2002 - present) to extract some pretty startling results for Climatic norms. I've forwarded some of this analysis to Neven, and will post more soon.

[FrankD]

Lodger,

I missed your July call, but independently have been saying the same thing for several weeks now - that we can expect a late minimum; I also picked "after 24/9" for the date.

SST's look high in the Beaufort and E.Siberian Seas, and should do some more damage yet. I posted over at Patricks as well - N.Atlantic SSTs are generally high, and I wonder if some of that will flow into the Greenland and Barents Sea as well. If you look at the regional charts for icea area on CT, you will see that a lot hit their maximum negative anomaly in October, or even November, as the sea surface refused to freeze until six weeks or more later than climate. I think we'll see the same this year (mind you the late freeze didn't stop a high extent by April).

[FrankD]

Phil,

Minimum extent from IJIS since 2002 occurred on:
2002 9-Sep
2003 18-Sep
2004 11-Sep
2005 22-Sep
2006 14-Sep
2007 24-Sep
2008 9-Sep
2009 13-Sep

If you want thirty years, you have to go for CT area. The data is here: http://arctic.atmos.uiuc.edu/cryosphere/sea.ice.minimum.2008.html

But its self-contradictory.
1) They have archived pix of the minimum for each year, and list the following dates:
5-Sep-79, 28-Aug-80, 6-Sep-81, 9-Sep-82, 4-Sep-83, 10-Sep-84, 7-Sep-85, 8-Sep-86, 27-Aug-87, 10-Sep-88, 12-Sep-89, 3-Sep-90, 16-Sep-91, 1-Sep-92, 4-Sep-93, 5-Sep-94, 24-Sep-95, 8-Sep-96, 29-Aug-97, 11-Sep-98, 11-Sep-99, 9-Sep-00, 15-Sep-01, 2-Sep-02, 16-Sep-03, 2-Sep-04, 6-Sep-05.

I note that the extent minimum is after the area minimum (on the four years for which there is an overlap) - I guess that means there's a few days where interstitial patches of water are freezing up in the centre while there is still a bit of fringe melt of the loose stuff. The trend is +.17 days/year (ie minimum is getting later by about 1 day every six years on average)

2) Further down the page, they have a graph of the date of the day minimum is reached. It bears no resemblence to the above data, and the trend is virtually flat (perhaps +.05 days/year at most).

I don't know which is correct. Maybe you should email them at the addy on their front page.

[fredt34]

Hi all,

I've created an animated gif from http://www.aari.ru/odata/_d0015.php?lang=1&mod=0&yy=2010.

It's on http://img521.imageshack.us/i/aaru.gif/

(please feel free to give me any advice about anigifs; Neven, of course you can reuse this image if you find it interesting).

Enjoy !

[Lord Soth]

The Canadian Ice Service has come out with its 30 day forcast, and they expect above normal tempertures. Combined with warmer ocean waters; they don't expect the onset of freeze up until the last week of September; and that is mostly sheltered bays.

I image the remainder of the arctic is in the same scenario. I'm expecting a late minimun this year; maybe even after the 24th of September.

[Jon Torrance]

Lodger,

Regarding " I think the reason our A/E methods produce different results is the way CT vs. IJIS handle the 300,000 km^2 'data hole' at the North Pole.", you clearly aren't saying the difference between CT area and IJIS area is explained by differing treatment of the 300,000 km^2 'data hole' at the North Pole since that difference is much larger than 300,000 km^2. I assume you're saying you think differences in how they vary from day to day are due to differing ways of handling the data hole but I'm not sure why you want to attribute them to that rather than to other differences, including the ones, whatever they are, that explain why CT area is so much smaller than IJIS area. Could your elaborate as to your reasoning?

[Christoffer Ladstein]

Kevin: I have also noticed the special warm spring/summer across Canada/Greenland, but based upon DMI.dk and met.no/retro, the anomaly for the stations in the very north of Greenland and Svalbard, have lingered between 1,5 - 2,0 C, which is quite moderate compared to places further south. Kangerlussaq, in the West of Greenland, have had strong positive anomaly in ALL months since December 2009. Also Jan Mayen, Svalbard, Bjørnøya and Hopen (small islands south and southeast of Svalbard), have for several years now experienced considerate positive anomaly in 11 out of 12 month. Strong evidence for a warmer Arctic?

[Artful Dodger]

Jon: CT's area figures are consistently 300 K higher than IJIS because IJIS does not add in the data hole. Strangely, IJIS does add 300 K to it's extent figures to compensate for the data hole. I can not speak to any other reasons why CT Area / IJIS extent differs from Frank's method because I have not seen his data.

[Peter Ellis]

Jon: CT's area figures are consistently 300 K higher than IJIS because IJIS does not add in the data hole.

This can't be correct. CT currently gives the area as 3.3 million sq km, whereas IJIS chart gives it as a little over 4 million sq km. Not only is the difference a lot more than 300K, it's also the opposite direction to your suggestion.

[Phil263]

FrankD 1/9 @ 17.21
Thanks for the information. The graph at the bottom is plotting the dates of SIA minimum while I assume the dates on the images are for minimum SIE.
It would be interesting to find out while dates for minimum SIE are trending upwards while dates for minimum SIA appear to follow a flat trend!

[Phil263]

Big upward revision of the SIE figure by JAXA (+9,275 sqkm).I also note an upward movement on the SIE graph by Uni Bremen . I wouldn't be surprised if we got another SIE gain for 1/09.

[Charles Wilson]

Sea Ice Updates were handed in Aug 16.
Earlier were: May 30, June 30 (me) July 14 (everyone else).
... They did finally printed my July 14 Update as an Appendix. It was mostly about Wayne Davidson & how the Clouds WOULD depart ... I laid out Several Possibles.
>> By August, I had a partial winner: the "CTI". a La Nina Index. High Pressure returned on schedule (6 week lull), Dipole Drift began a week later.
>>> But the CLOUDS ... are still here ! High Pressure did NOT mean Clear Skies !

Now I don't believe I did this, but even though I added 3 OTHER ways of using the El Nino rating (compared to 2007, 1.8/1.1 or 1.64), which was the HEART of all my calculations (which made my own little "ensemble" of 4 in my August Outlook - - the 2.5 was thier average) ... I failed to multiply the CLOUDS times 1.64

The Whole Point of Wayne's Theory is: that La Nina DRIES OUT THE CLOUDS.
>> And the El Nino making them peaked 1.64 times higher.

>> 2007 had a melting "lull" in Most of May to short of Mid-June. Thus 5-to-6 weeks x 1.64 = SUNNY SKIES somewhere around Sept 5 by what I will call Method 1 - - or Spt 12 or so by Method 2m -- Method 2m will adjust for Modoki El Ninos - - which last 2 months longer - - by adding 33%:
Method #1: 42 days x El Nino Rating/1.1 = ? (as 2010 = 1.8, thus = 69 days from 29 June = ~ Sept 5)
Method #2m: 35 days x El Nino rating + 33%/1.1 = 78 days = Spt 12

Obviously I am a bit uncertain judging Cloudiness in 2007 by eyeballing its Melt - - thus the use of 5 weeks in 1 & 6 in the other - - but the whole year has been guessing at an algorithm to figure out the WINDS, & then testing it.

I know it all has been seemingly Unsuccessful. But we kept finding something NEW to add which may look like EVERY thing is going "wrong". But SCIENTIFICALLY, Nature was just teaching her all her Tricks - - well, i hope all -- and Each was a Unique & precious GAIN.
>> Yes, I think we can - - Now - -Actually PREDICT just WHEN the Dipole drift, High Pressure, and even the CLOUDS, will occur & when they will stop. And even WHY (as sketched by Wayne Davidson).

Once we see which fits better, method #1 or Method #2

(Of course future researchers will doubtless adjust for how strong the following La Nina is; Method #1 essentially assumes it matches the preceding El Nino's total "area", as it were = intensity TIMES duration - - this may even be a way to PREDICT LA NINA INTENSITY as the coming one is showing signs of being even MORE extreme than the recent El Nino was - - WUWT is suggesting a 3 !
... PS: an El Nino "Modoki" (japanese for "strange") - - arises in the mid or West Pacific where the Classic Nina & Ninos start off Peru & spread Westwards). It takes months longer to fade than the standard kind.
PPS ... if the Clouds stay away later than Spt 12, maybe 6 weeks times Method #2, etc.

[JackTaylor]

Added some August images to my onLine AGW folder.

1. Is a static panel of cropped IARC-JAXA sea ice concentrations and with some red arrows pointing toward changes I see in the left panel and replaced with questions marks in the right panel for waiting to see what happens.
http://www.polk-nc.com/agw/ten0831panOpti16.png

2. Is an animated "Flash" from July 30 with every fifth (5th) day of August to end then daily. It magnifies (if your browser is 'maximized') and shows some very interesting dynamics of the ice pack (to me anyway).
http://www.polk-nc.com/agw/framExp.swf

A lot of action still going on up there. I might get to see some export through the Olga Strait between Svalbard and Franz Josef Land yet.

And look at the instability in the East Siberian Sea over to the Beaufort Sea.

Until it settles down, it will not allow the Fat Lady "On Stage."

[Lord Soth]

It flat lined again tonight with a insignificant gain.

Its beginning to look like 2005, where it also flat lined for a week and the NSIDC declared the season over. They had to retract as 2005 went ahead and lost over 300,000 km with the minimun on Sept 22.

[Phil263]

are the conditions in 2010 similar to 2005? Isn'it possible that we already passed the SIE minimum on August 31 when considering the weather forecast for the next 10 days?
If so 2010 would be the fith lowest minimum behind 2007, 2008, 2009 and 2005 which " beats" this year by a mere 14,000 sqkm!

[Phil263]

Jack Taylor

And look at the instability in the East Siberian Sea over to the Beaufort Sea.

Until it settles down, it will not allow the Fat Lady "On Stage."
If you compare the map for 31st August with the map for 1st September , you can see that there is little melt happening in the Beaufort and the ESS. OTOH, on can see that the ice front is advancing to the East of Svalbard and to the North East of Greenland. This seems to be somewhat compensated by some melting happening in the Laptev Sea. But overall the last 4 days have seen a "stalling" in the melting.
Being a non-specialist, i cannot say for sure what this portends, But there is a definite possibility that we would not see any further sizeable SIE decrease that would bring the SIE minimum further down. A minimum extent of 5,329,275 would still be considerably lower than the long term average, but surely not as low as what it was expected to be two months ago (see Patrick's piece for a dicsussion of 2010 melt.
In this context, the fact that 2010 may only rank as the fifth lowest SIE extent does not invalidate the hypothesis that arctic sea ice is not recovering. OTOH I am a bit puzzled that the minmium ( if it turns out to be that way!) would have been reached much earlier than in previous years. You have to go back to 1997 to find an August minimum (29 August).

[FrankD]

@Phil263
The graph at the bottom is plotting the dates of SIA minimum while I assume the dates on the images are for minimum SIE.
<. facepalm > note to self: RTFM! Thanks for clearing that up.

I'm sure there's a tale in the differences, but I don't know what it is.

[FrankD]

Lodger - my data is simply derived by reading values as best I can from the IJIS area graph: http://www.ijis.iarc.uaf.edu/seaice/extent/AMSRE_Sea_Ice_Area.png

Crude as hell, but I don't have access to the source data. Neven asked them about their data and they responded that the area numbers were not very solid and they didn't want represent that they were so they weren't planning on making the source data available. I'm down with that. While having authoratative data is far preferable to my method of interpolation, I feel that by approaching it like this I am at least comparing like to like, since I'm drawing on the same source. But its a flawed process, and I don't claim otherwise.

OTOH, I find the CT numbers a little confusing. I presume that the last column is the average against which the anomaly is calculated, but that seems to shift from year to year. Is it a rolling average? If so, that seems a pretty weird way to calculate an anomaly from a baseline to me...

But I agree with Peter that the "hole" doesn't account for the difference Just compare our minima differences - in 2008 you're at 60% and I'm nearly 80% (one day apart). On an extent of 6,000,000 for 14th or 15th August 2008, you have an area of ~3.7 (?) and mine is 4.7ish. So even allowing for a systematic 300K there is still 700,000 sq kms ~discrepancy unaccounted for....I can only presume that their counting methods are different, or perhaps CT stops counting at a higher latitude...??

[Kevin McKinney]

Phil, anything's possible, I suppose--but why would this season, with considerable warmth available in the system and the ice appearing to be vulnerable, "stall out" this early? It would be the earliest minimum in years, and I see no particular physical justification for it.

[me.yahoo.com/a/nSjChi4X3vr8X3DRw93GkY1.cerja.8nvWk-]

It may be stalling out in extent but the CT area continues to drop, -72,000 yesterday, the chance of passing the 2008 and 2007 values seems good given the areas of low concentration still around. It would be surprising if the extent doesn't follow to some extent.

[JackTaylor]

Phil263 | September 02, 2010 at 06:10
"If you compare the map for 31st August with the map for 1st September , you can see that there is little melt happening in the Beaufort and the ESS."
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
It's difficult for me to get an ("exacting enough") direct comparison of 31 AUG and
01 SEP because of the *.pdf vesus *.png you refer scale and some rotation. In their archive, which I prefer,
http://www.iup.physik.uni-bremen.de:8084/amsredata/asi_daygrid_swath/l1a/n6250/2010/aug/asi180-n6250-20100831_nic.png
versus
http://www.iup.physik.uni-bremen.de:8084/amsredata/asi_daygrid_swath/l1a/n6250/2010/sep/asi180-n6250-20100901_nic.png

I think the Bremen images are great, but, for me with limited experience & skill to process to get undistorted exact overlays, I'd be fooling myself.

BTW, if as I see it on the Bremen, ice patches are actually growing in Beaufort & ESS.

I like to think of instability - melt as two different things.

As I get time, I'll do a five (5) - seven (7) day difference animation, as a one or two day plot sometimes does not show any difference.

[Kevin McKinney]

"me," I noticed that as well. It seems that there's spreading of the ice going on. Supporting that notion, there are extent upticks (CT data) in the Greenland Sea, the Archipelago, and Baffin Bay (meaning, probably, Nares Strait) even as the value for the Arctic Basin continues to drop. It would appear that ice transport south continues.

[Jon Torrance]

I'm finding it quite interesting looking at the Cryosphere Today ice concentration animation of the last several days at fairly high speed. It's far easier to appreciate just how much the boundaries of the ice pack are moving than by comparing static images from successive days. It also gives one a better feeling for just how hard it is to judge by eye which of advance or retreat has the upper hand.

Regarding stalling in extent decrease, there's definitely a strong temptation to read too much into it. In 2006, things hadn't merely stalled at this point, there was actually an extent increase of 39,844 from August 30th to September 1st but the extent still went on to drop a further 211,719. Almost anything may be possible given the unpredictability of the weather but past experience says the extent is likely to drop at least somewhat more. In the JAXA record, 108,906 is the smallest drop from September 1st to minimum extent.

[toby]

Well I am confused - blame my inexperience.

Both the CT and UofB maps are showing a hell of a lot blue - areas where ice concentration is below 50%. There is as much now as there ever was. That ice "peninsula" in the East Siberian Sea looks like it could completely disintegrate.

I notice also that at http://www.meteo.uni-koeln.de/meteo.php?show=En_We_We there are dominating high pressure areas and the Arctic temperatures are warmer than average for this time of year. But low pressure and high winds might have a bigger impact.

What does it all presage? On the figures, it seems that the fat lady has cleared her throat at least, but the maps make me surmise there is still some way to go in the plot before she bursts into song.

Apologies for the extended metaphor at the end (blame Neven for that!).

[Jon Torrance]

Toby,

I can't offer any confident answer to your big question. I just thought I'd comment that the low colour version on the CT concentration maps - http://arctic.atmos.uiuc.edu/cryosphere/NEWIMAGES/arctic.seaice.bandw.000.png - is perhaps even better than the version whose large areas of blue you mention at highlighting things such as the apparent fragility of the ice peninsula sticking into the East Siberian Sea, since the colour scheme running smoothly from white for 100% concentration to black for open water graphically shows how much of the surface in low concentration areas is water rather than ice.

[toby]

Jon,

Wow!

[Jon Torrance]

Here's another good illustration of the state of the ice pack compared to other years, also from Cryosphere Today - http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=08&fd=31&fy=2009&sm=08&sd=31&sy=2010 - just look at home much larger the red and green areas (about 40 to 60% concentration) were on August 31st of this year compared to a year before.

You can compare to other years too, of course. The comparison to 2006 ought to be particularly instructive for anyone who ever entertained the possibility that Steven Goddard might be onto something with all his writing about how much like the ice pack of 2006 the ice pack of 2010 has been - see, for example, his contribution to the August Sea Ice Outlook.

[Jon Torrance]

"home" when I meant "how" - one of the stranger typos I've ever produced. I wonder what was going on in my brain to make that happen.

[Gas Glo]

NSIDC August average extent and areas have just been put up:
2000 8 Goddard N 7.21 4.71
2001 8 Goddard N 7.47 4.87
2002 8 Goddard N 6.53 4.23
2003 8 Goddard N 6.85 4.44
2004 8 Goddard N 6.83 4.57
2005 8 Goddard N 6.30 4.14
2006 8 Goddard N 6.52 4.24
2007 8 Goddard N 5.36 3.11
2008 8 PRELIM N 6.07 3.43
2009 8 NRTSI-G N 6.26 3.79
2010 8 NRTSI-G N 5.98 3.54

2nd lowest extent, 3rd lowest area

[Jon Torrance]

Something I just noticed - the University of Hamburg's latest near real time concentration map (currently ftp://ftp-projects.zmaw.de/seaice/NEAR_REAL_TIME/Arc_latest_large.png though it will of course point to the newer images each in turn as they come along) now shows a decent-sized lake (technically, I suppose it would be better to call it a very largy polynya) in the middle of the ice pack at about 165W/86.5N. At least, it looks like exactly the same shade of blue they're using for open ocean to my eyes.

[JackTaylor]

Gas Glo | September 03, 2010 at 00:47
"NSIDC August average extent and areas have just been put up ... 2nd lowest extent, 3rd lowest area"

That doesn't seem to support the Arctic Ice is recovering

Jon Torrance | September 03, 2010 at 03:58
" it looks like exactly the same shade of blue they're using for open ocean to my eyes"
Hopefully both of us are not color-blind, looks open water blue to me and
also shows on the Bremen University map, as in Frame 09 02 of this animation
(magnified) http://www.polk-nc.com/agw/asiBre0902.swf
original scale at http://www.polk-nc.com/agw/asiBre0902.html

More action to come?

[Phil263]

JAXA reports SIE of 5,301,406 (-30,938) (preliminary). Unless there is a huge correction tomorrow, 2010 is now in fourth place behind 2007,2008, and 2009 with only 51,562 before overtaking 2009 to third place.

[Kevin McKinney]

31K extent decrease, more or less, as a prelim. Knocking on the door to get under 5.3--we'll see what the revision brings.

[L. Hamilton]

Updated graph of mean ice concentration (adjusted NSIDC area as % of extent) through August 2010:
http://img.photobucket.com/albums/v224/Chiloe/Climate/area_vs_extent2.png

As in the past 4 years, mean concentration reached its low point in August. 2010 was 3rd-lowest on record, below 2009 but above 2007-8..

[Neven]

Phew, gone for just 2 days and it took me many hours today to read all the (excellent) comments here and look at all the maps and graphs. Compactness graph is very interesting!

Next SIE update will be out tomorrow.

[Christoffer Ladstein]

24,9 C in Nuuk, Greenland, yesterday, a brave new record for the whole of Greenland in September. And they say the average temperature in the era of Vikings were lingering 1 C above present temperatures... I figure even them would feel hot in their furs these days. Wonder how much ice will have turned into water from the great icesheets of Greenland this HOT summer? Enough to raise sealevel another mm...

[Steve Bloom]

Jack, with winds apparently having gotten the ice into that condition in recent days, as long as they hold up it seems like warm water should be able to finish much of the job on it even if air temps and sun have ceased to help. These are exciting times indeed (as these things go).

[JackTaylor]

Steve @ 01:44
Winds and warm water it seems to me also, but do not know.
Air temps I wondering about the data, as Jon Torrance | September 03, 2010 at 03:58 and the polynya (lake) or open water near 165W/86.5N which to me is high-arctic and shouldn't it be freezing over at that high a latitude, see:
http://www.polk-nc.com/agw/arc2x100903.png

In that image I'm curious if a separation may occur along the red or yellow line, freeing about one-third ( 1/3) of the ice-pack to be worked over by the Transpolar Drift, then again, maybe it's as Artful "Lodger" Dodger has pointed out it's warm water from the Bering Strait - Chukchi Sea having more of a free path to the Nares Strait - Greenland Sea. Oh, well warm water from an El nino has to dissipate somewhere or a La nina would be blocked.

All in All, I don't know, but it's fun watching and speculating. LOL