« Nares Express is ready to leave | Main | Ice pack in full »


Feed You can follow this conversation by subscribing to the comment feed for this post.


I'll stay with 2.2

The rational is more not changing horses at halftime than any intuition or modeling.

However, if I was doing it again, I would say 3.2 since I think there are large parts of the ice that have little damage and won't be touched now, much more than I thought earlier.

Take which ever prediction seems more in line with what the survey requires i guess.


4.0 M Km^2

Based on intuition. I think this year's melt will be significant, but it won't reach the level of 2012.



Thats what I expect; Volume about last year; extent and area (yet) above last year.

4.0 Thats what I hope; given the steady volume.

3.0 Thats what I fear; given the fractured ice visible, and the cyclones.

Rob Dekker

4.38, with a standard deviation of 227 k km^2.

Details in this post, and my latest comments therein :

I did not change my prediction from last month, neither up nor down, since ice losses over the past month increased consistent with this prediction.

Frankly speaking, I don't think that weather plays a big role any more after June. With clear weather, direct melt prevails due to increased SW radiation, but with more clouds and storms more heat is brought in from the south, and IR losses to space are reduced.

So either way, the ice is going to be beaten badly. But the statistics suggest that not enough energy was absorbed early on (until beginning of July) to validate a prediction much below 4.3 M km^2.

So the question is if 2013 will beat 2007, and statistics of snow cover, ice extent and area suggest there is a 50 % chance it will.

We'll see what September brings...



Fairly subjective, but with detail. I've been in and out of Alaska (including the arctic, and parts of arctic Canada) since 1973, and have experienced the weather and climate in that part of the world each year since, either directly or vicariously through friends and contacts.

This year is beyond all living memory (not just my own). Mid-June South Central Alaska hit into the 90's for more than a few days, with 80's common and frequently repeated (including yesterday). The sun is out so much this year, the tourists think clear days are the norm (NOT).

Meanwhile, some of that extra solar energy from the past decade or two, that's been hiding in the deep oceans, is mixing with a pretty blended mess of thin and slushy ice above 70 degrees N. And the statistics Rob D refers to above are built from past experience, which is pretty much old news, the old days, another time.

I'd like to be hopeful that the thin blue line for extent over at NSIDC is in fact leveling out along the bottom of the gray range, but am afraid of August.

I'll be happy to see more than 4.


2.5 +1.0 -0

Same as last time.

Things may be happening slower, but when they happen, they happen very quickly. I believe there is a lot of momentum building which will play out in August and September.

Also I've noticed the comments over the years that the melt season is moving further into September. which means that if the melt season runs to the end of the 3rd week in September, the average will be lower.

Just my take.

r w Langford

Others have probably noted that the record melt last year came about because of a lengthened linear decline in extent. The linear decline continued for three to four weeks longer than other years. This is easily seen in the extent graphs for the period from August 1st through into early September. In other years the melt slowed and the graphs started turning several weeks earlier. No doubt this is a result of more heat in the system and thus a longer melt season. There is no reason to think that this has changed this year. Weather still has a major effect on melt rates but transferred heat from the tropics via currents and air currents will continue to extend the melt season by several weeks both spring and fall and thus result in frequent record melts. This year, the length of the melt season as determined by stored heat in the system will determine whether we have a new record or not.

Kevin McKinney

Interestingly, Canadian MSM has this on North Pole 'melt cam':



It also attracts the birds


Kevin McKinney

Birds, or one, at least!--wonder what it (they) is/are doing up there? What's the ecology of that? Or just passing through, maybe?

Every bird watcher--even a rather poor one like me--knows that random wanderings happen every now and then. "Accidental" is the usual term, I think.


4.18 The chances of a big ice grinder 960-965 mb showing up in Mid-August 14-16 and then sticking around for 2-3 weeks is looking more and more favorable. If no ice grinder then lack of spring melt will raise the sept. avg. to at least 4.46-4.62. There just is not enough warm water available to provide bottom melt.


Great catch Derek! I'm sure someone can tell if that's a Kittiwake or Fulmar, or am I completely wrong?

james cobban

2.8 I'm sticking to my original figure out of sheer bloody-mindedness. Commenters on this blog make powerful arguments for both higher and lower figures, making it difficult for me to decide who's more likely to be right, and having no real knowledge of my own to draw upon, I am left as adrift as a bouy at the North Pole. Admittedly, most of the weight of those arguments is in the direction of higher estimates, and if I were betting money I would raise my estimate substantially. But my gut feeling is that the ice is vulnerable this summer, and will flash melt sometime in August, and then continue to melt until around the end of the third week in September, giving a low monthly average.


Sticking with 2.9
Probably mid range 3-4, but here is my thinking.
based on weather, in the 1980's there would be a major uptick. This means the ice is changing and that there are other 'heat' factors at play that we may still not understand or know.
Also, I am convinced that MYI, which based on what happened last year is really no more than 3-4 years old at best, is so porous and fragile that it is no better then a 20ft thick slushy. So in terms of ice we could be starting to think in terms of mass. How to do that I have no clue.
Based on that thinking, all it would take is a week or 2 of the next 8 weeks of good melt conditions to flash melt a large % of the current extent.


4.28 km^2
Strictly a guess and up from my last guess. From observing MODIS, the ice seems spread out and fragmented, so an August weather event or two could certainly have a dramatic effect, but right now I think we may end up with higher extent of very weak thin and fractured ice.


4.28M km^2 - forgot the M in the last post

Kevin McKinney

"I am left as adrift as a bouy at the North Pole."

Nice one, James!

Joshua McCurry

4.9 MKm^2 +-0.4 Km2

The ice seems to be holding its own this year. Volume (at-least out to July)is slightly higher than 2010, and judging from Bremen and Eosdis Worldview there is significantly less whispy, deteriorated ice than last year. Specifically, the CAA and Beaufort are incomparably better off than last year, with break up near the archipelago occurring nearly two weeks later. Export also appears to have been curtailed by the lack of a dipole and relatively late removal of the Nares bridge. The next few days will probably not see much in terms of melt, as anomalous cold moves into the ESS/Laptev and the Siberian/Atlantic side of the CAB as per the GFS. The relative warmth forecast for the Beaufort/CAA will most likely be offset by wind driven intrusions of MYI. Barring a massive cyclone, the remaining 3-4 weeks of melt time will not be enough to drop the ice below 2011 levels even if favorable conditions return. OTOH, if a weak low persists in the CAB, the low temps could keep the ice close to 2009 extent. Note: I could well be wrong, since I'm far from an expert at this

Fairfax Climate Watch

area: 0.0 (zero)
reasons: (1) Based on comparison to Navy CICE model output since 2010, the amount of ice that is highly vulnerable to complete melt (based on thickness, concentration, and location) by the end of August is approximately 70% of current ice area, far above 2012. This much ice loss before September may enable feedbacks that eliminate the remaining ice. (2) Sea surface temperatures in the North Atlantic are high, like last year, indicating similar northward heat transfer as last year (3) The area between 85 N and 90 N has substantially less ice volume and concentration than seen since 2010 in the NAVY CICE models. Because this area is the first to lose summer sun, it is the most prone to refreeze first. Yet, if it loses the majority of its ice cover by mid-August, vertical mixing and atmospheric heat transfer could melt the rest of that area (and all other areas) by September 1. (4) If the Arctic is ice-free by September 1, feedbacks could keep the water ice free all month.

Ned Ward

Joshua McCurry writes:

"4.9 MKm^2 +-0.4 Km2" [... followed by a nice paragraph of explanation ...] "...Note: I could well be wrong, since I'm far from an expert at this"

That's a very reasonable choice, and your explanation makes a lot of sense. Earlier this month I stupidly announced that I was sticking with my earlier prediction (4.4) but if I hadn't done that I'd join you at 4.9.

As of today, NSIDC daily SIE is in 6th place, at 7.275 (behind 2007, 2009, 2010, 2011, and 2012). If the recent slowdown continues for another week, SIE will fall behind 2005 as well. On the other hand, if it speeds up a bit, it would be pretty easy to pass 2009 and 2010 and move into 4th place.

If you start at today's daily SIE and use statistics to predict the Sept monthly min based on how much has melted after this date in other years, you get 4.8 +- 0.9 (if "other years" means 2007-2012) or 5.0 +- 0.8 (if "other years" means 2002-2012).

So your prediction of 4.9 is nicely in the middle of that.

Paul Klemencic

I expect a SIE of around 4.9 as well. Last year's melt would've come in close to 2007 without the GAC of 2012, or about 4.3. But last year there was very large amounts of solar absorbed in the open areas of the Beaufort, Chukchi, E. Siberian, and Laptev, that set the table for the cyclone. The ice pack this year in those regions isn't going to come anywhere near last year's melt.

The only wild card would be another GAC, and the forecast doesn't show anything close to last year's storm. For a storm to have maximum impact, it has to hit in the right spot (centered around 80N, somewhere between 105W to 165E longitude), and hit in early August. We don't see anything like that on the horizon.

So the best forecast would extrapolate current SIE to end of season based on normal year's melt. This would give a forecast SIE of around 5.0, but allowing for some thin ice in some regions, I get 4.9. We get a one year reprieve in bad news.


1.3 million square km,

and here is a cleaned up version of the messy numbers I provided earlier:

According to PIOMAS data from Wipneus: (https://sites.google.com/site/arctischepinguin/home/piomas ), average volume drop in Jul, Aug, Sep & Oct from 2000-2012 (13 years) has been 8.000 km3/13y = 600 km3 per month on average (- 2600 /+1200 km3). To use this number at face value, would constitute the “conservative bias”, as discussed by Neven here: (http://neven1.typepad.com/blog/2013/07/are-scientists-conservative-about-sea-ice.html ).

Maximum volume drops in Sep year on year this century occurred in 2010 and 2007 (>2500 km3) and the maximum year on year volume gain in Sep was in 2001 (1200 km3). It is noticeable that – similar to this year - both 2001, 2007 and 2010 had Arctic mean temperatures below normal throughout the entire summer (see http://ocean.dmi.dk/arctic/meant80n.uk.php ). Both in 2001 and 2007 we lost 5500 km3 between Jul and Sep, whereas in 2010 “only” 5100 km3 was lost between Jul and Sep indicating, that we may still lose > 5000 km3 before the end of Sep this year. This will eventually lead to a year on year drop of > 2500 km3, which brings down the remaining ice volume to 1100 km3. If you spread it out a bit and use an average thickness of 1.1. m, we should end up with a mean Sep sea ice extent close to 1.3 million square km. This extent is - by the way - similar to the hypothetical melt scenario provided by A-team here: http://i1340.photobucket.com/albums/o728/OlTom67/meltScenario_zps17c002cf.gif

Again referring to Wipneus’ site, both 2007 and 2010 had some of the highest Fram Strait volume exports this century – around 200 km3/month in both Aug & Sep - so a new record ice loss this year requires high export numbers through the Fram Strait, which again requires favorable wind conditions to push the remaining MYI out of the Arctic basin. With the current weather forecast, we should expect to see a strong W-E pressure gradient north of Greenland after the weekend, which could take care of that.

However, Fram Strait ice export alone is not enough to get rid of the remaining ice. Other mechanism will have to kick in to help out.

In principle, four mechanisms may play together, assuming that the remaining piles of ice rubble will be clogging the straits surrounding the Arctic:

1) Solar heating requiring cloud free conditions and thus a confined central PAC.
2) Advection of latent heat from lower latitudes would be a continuation of the current Jet Stream pattern
3) Advection of dry warm air aloft would require a bunch of tropical cyclones in Aug and Sep, which would also hurt the GrIS.
4) Upwelling due to katabatic winds would require an extended summer season well into Oct.

Cheers P

Doug Lofland

I am staying with 1.8 Mkm^2

Just uploaded another melt video through July 30 with a side by side at the end comparing to the same date in 2012.


As Kevin says the melt pattern is very different, but the really significant change this year is the highly broken up core north of 85 degrees. In 2012 everything that was dark purple stayed in September, but this year there is very little of that same dark purple anywhere. Instead the textured (if you will) purple seems to correspond with broken sheets with significant open water in between, as seen when the clouds allow a visual view. The Sea Ice Extent daily reports are not counting all the open water and it ranges up to gaps of 10 kilometers.

So to me its like the solid sheet of ice, thanks to all the cyclones, got busted up into a millions of pieces and spread out over a larger area. It also moves around and changes much more than in previous years. NSIDC draws a line around the ice and them measures the area, and in past years that worked. There were never internal areas of significant open ocean. This year is so different.

One of the big changes to look for is with the mosaic of all open sea at the pole, will be the continued jet stream weirdness. If its super hot in Siberia, its likely cold in Kansas, as the cold air that should have stayed north is now driven south.

I fear that the extremes will get much worse very soon.

Chris Biscan

I am going with 4.65 million.

Even that will take either the largest drop on the NSIDC extent charts from this point on or close to it.

The Weather models show compaction picking up a bit soon but are still really cold for most of the arctic. The Beaufort gets torched good.

jaxa is sitting at 6.98 mil while NSIDC single day is at 7.15 million.

I can't see anyway over the next 5-6 weeks the rate of extent drop per day will be near 100,000K to reach the blogs consensus.

L. Hamilton

I'll keep the polls open through Friday if anyone has a late guess or revision.

I'll summarize our results in a post early next month. Although I'm focusing initially on the numbers, rationales are getting archived too, and will be part of a more detailed analysis this fall. They've been noticed by at least a few scientists.


3.75... up from 3.0!

I still think that we will see another cliff soon!

James Lovejoy

4.35 MKm^2 I've edged up a bit. It'll be interesting to see not only how this year's melt turn out, but what the next few years bring.

michael sweet

I guess 4.0 Mkm2. It has been so cloudy and cold that I have to raise my guess. I think there will be a lot of melt in August, but not enough to catch 2012.

Tor Bejnar

My guess for NSIDC Arctic sea ice extent September 2013 average is increased from 3.6 (previous guess) to 3.75 million km2. (Curiously, this keeps me in the same bin on the ASIF vote opportunity that ended 10 or so days ago. I didn't peek before my calculations, honest!)

This projection is largely based on three HYCOM thickness maps: July 29, 2012 & 2013 and September 15, 2012. I compared the 2012 maps to see what melted (or compacted) between late July and mid-September and projected similar changes onto the 2013 map. The Beaufort-Chukchi-East Siberian side of the Central Arctic in 2012 had much “2” to “3.5” meter thick ice melting (or moving toward Greenland); I do not expect a repeat. As I have no electronic graphic manipulation software or skill, I cannot accurately determine the extent I’ve drawn on a printout of the late-July 2013 map (talk about crude science!) I took my HYCOM projection (~4% increase) and increased the NSIDC 2012 September extent average by the same factor. [The quoted HYCOM thicknesses are because that is what the graphic indicated, not what was really there at the time.]

I still expect ASI volume and maybe area to have new minimum records this year, but I expect icy rubble will remain dense enough to cause NSIDC extent to not enter record territory. As many a fellow ASI watcher has written, the weather will be significant in determining if new records are set or not.

Even as I remain pessimistic about Arctic ice survival, I realize more clearly this year that my understanding of ice dynamics and influences are significantly deficient to qualify me to say much of anything in regards to Arctic sea ice. That said, I grow increasingly appreciative of posters here with significant graphics expertise, and I participate as I can.


4.6 Mkm2
Up from 3.4 the last 2 rounds.

Not based on any scientific logic, I've increased it just because of the cooler season & the limitations of the means of measurement here, understanding full well the ice is in poor shape.
I wish it was because the ice was MUCH thicker & sturdier, was 'recovering', but I'd be dreaming!

John Christensen

4.8 Mkm2
Up from 4.25 in last round.

Ice is spreading, keep lows forming due to high temps on surrounding continents, will in turn keep Arctic temps down and prevent SST from rising much.

james cobban

Larry, I would like to change my entry upthread from 2.8 to 4.8 Mkm2. I no longer think there can be a huge flash melt in August given the moribund state of the melt in the last week and the -5C air temperatures.


4.5 Mkm2. Up from my 3.8 Mkm2 prediction in July.

I think the cold and cloudy weather we have been seeing throughout the summer is finally starting to show on the ice graphs. At best we may challenge 2007 for second place but I don't see the 2012 record getting broken this summer even if the weather becomes warmer and sunny. We just have too much ground to make up.


Yes, the extend is up, spreading out the mushy ice pack. But I still maintain it will be a very low year. I realise the sun is leaving and the clouds have been helpful but I'm looking at the import of air over the next few weeks - it will include heatwave air from China, smoke/warm air from Russia and heatwave air from southern Europe. All these sources feed into the arctic more or less of a period of a week - some quicker, some slower.

Certainly though by mid Aug we'll know if the spreading was helpful or not. I'm waiting till then.


I'll stay at 3.2, though this has the feeling of being anchored by my earlier guess. But as rational as I can make it: I still think the ice is porely (ha ha) and I think Stormy August will take a big hit out of what is left.

Climate Changes

2.92 MKm2 and NP ice free.

In spite of the unexpected melt freeze up, I'm only going up 10k from July. Main reason state of the ice and the still to come cyclone to match GAC2013 that I've expected all summer but so far has failed to materialize :/



Looks like the trend is heading for the upper end of the region defined by the previous three years + 2007.

The Russian side of the Arctic could melt out suddenly and pull the rug from the high-end predictions, but it seems less likely right now.

And 4.8m km^2 is still not very much ice historically, even though so many here have gone for much lower figures that "high-end" seems like an appropriate label.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)