« ASI 2016 update 4: breaking point | Main | PIOMAS August 2016 »


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


Nice to read that top 3 is in the projection mindset, there was never any doubt about this in my mind. However Dr Slater's forecast is still not as good as the Met office :) and there is a lot of work to be done in judging melts for instance:


What does extent numbers mean when there is huge melting and hardly a rate to prove it? My latest bit of presentation demonstrates there is a continuous resupply of sea ice North of Beaufort Sea, even though nearly all of it melts within days. Leaving the extent numbers practically unconvincing while there is quite a lot of action going on.


quote>>>The SIO is organized by the Sea Ice Prediction Network (as part of the Arctic research program 'Study of Environmental Arctic Change', or SEARCH), and is a compilation of projections for the September 2015 Arctic sea ice extent, based on NSIDC monthly extent values.<<<
Neven is the date in this correct?


It was correct last year! ;-)

Thanks, Philip. Fixed now.


I see Watts and his merry band have remained silent..

Have they finally gone away from SEARCH or do we have to put up with them popping up some time in the future predicting that the climate will turn itself on it's heat through "faith"...


WUWT disciples wont know the difference between a real event and a made up one, because they spend so much time confusing people. As a result, they likely delude themselves a great deal more.

Yesterday's 86K JAXA drop was strong despite Nature trying to elude our perceptions as well. The sea ice North of Wrangel Island is particularly vulnerable and truly slush like, I don't give any chance for it to subsist very much longer. It is a massive extent of slush ice without a melt zone, which luckily give clues about how fast ice melts:


Robert S

Given the current state of the ice, and the weather trends, it's interesting to theorize what conditions might lead to higher or lower extent minima. It seems likely that a worst case scenario is a continuation of current trends (lots of dispersed low quality ice, broadly low average alebedo, consistent heat import due to mild persistent lows, moderate cloud cover), followed by a major cyclone event in Mid September causing rapid compaction.

The best case scenario is probably the development of a mild persistent high over the central arctic, maintaining dispersal and reducing heat import.

The current ice condition is the key indicator leading to these scenarios, rather than the current extent, which has become deceptive, due to the dispersion/fragmentation of the pack. My feeling is that daily drops in extent during the next month are of relatively low predictive value - there's just so much potential compaction in the current ice conditions that massive drops are possible at any time... and the later they occur, the worse they are likely to be. At this point, therefore, any predictive model appears to be a crap shoot on guessing the weather over the next 2 months... and the usual model inputs likely have less predictive value than usual.


Robert S

"there's just so much potential compaction in the current ice conditions that massive drops are possible at any time... "

not only compaction potential, but melt potential due to the dispersion of sea ice rendered vulnerable by sea water encirclements:


Extent numbers must be qualified, there is no valid comparison possible without identifying ice pack densities. Extent numbers are still very useful, especially very telling holistically. Unfortunately we are not practicing qualitative descriptions of sea ice melt seasons enough.

Last day of July, 96,228 km2 sea ice extent loss was mainly due to numerous ice islands vanishing, according to JAXA map, especially North of Beaufort Sea. I would expect more plummeting CAB numbers going along with substantial overall decreases. We are now at the stage when we can say with confidence 2016 will be less than 4.5 million with recent lesser melt on record rates, the possibility for 2016 being #1 is still very much possible.

Rob Dekker

I tend to agree with wayne, "expect more plummeting CAB numbers going along with substantial overall decreases".

The thing is, due to the low sea ice "area" over the entire summer melting season, and the record low snow cover, there is a stupendous amount of energy in the system that has not quite materialized due to sustained "lows" over June and July.

Over the melting season, weather tends to average-out, so I expect a decent August melting month and short-term, expect some very solid declines that will put 2016 clearly at the ranking it deserves : 2nd or 3rd, either way close to 2007 despite the "lows" that have slowed down ice melt over the past two months.

Rob Dekker

Thank you Neven for featuring Dr.Slater's Probabilistic Ice Extent (SPIE) projection.
Please note that you explained his "Persistence" model, which Dr. Slater himself notes has no true skill.
His SPIE model DOES have skill, and indeed projects just shy of 4.4 M km^2 for the Sept average minimum in SIPN.
Personally I think that estimate is still too high, and we are heading lower (to 4.1 or even lower) but we will see in September who is right.


Indeed, Rob, we could be getting an exciting ending to the melting season after all. I'm less convinced now that it will be difficult for this melting season to end in the top 3. Despite a lack of classic melting weather patterns for almost the entire melting season, there seems to be so much heat in the system that extent/area drops slowed down, but never really stalled. Fascinating stuff, but it doesn't make it easier to analyse/quantify things (at least for me).

More on that in the next ASI update, of course.

In the meantime, on the Forum the August poll for predicting the 2016 September monthly average is up. It's the last poll of the year, so be sure to vote, everyone!

You can also vote on the JAXA daily minimum here.

John Christensen

Thank you very much for the update Neven!

With regards to the Sept avg. minimum I am inclined to consider the average poll forecast being too pessimistic given the recurring areas below freezing at the DMI 60N temp chart, the persistent lows and the moderate volume decline (if the DMI volume model holds, that is). We will see in a few days if PIOMAS is in agreement with DMI on volume, which will be important.

If the volume models are in reasonable agreement, then the Sept avg. extent should be in the 4.5-5.0 range, and then it will be interesting if it remains above the 4.63 mark..

However, given 2015 being the warmest year on record, 2015/16 being the warmest Arctic winter, 2016 so far being considerably warmer than 2015, and Arctic waters showing some of the highest warm anomalies globally, I cannot say I am surprised that extent numbers are still going down at a considerable pace. The question again will therefore be if the volume in the CAB has been shielded, and if this will cause extent drops to slow down from now on and until end of the melting season.


Everywhere we look we see vulnerable sea ice tossed around rapidly at the whims of geophysics, what matters is to find if melting is occurring simultaneously, this is observed by sea ice "goodbye waves" now common , even Northern Nares Strait:



It looks like the coastal route of the Siberian seaway will open in the next few days.

The northwest passage is rapidly breaking up as well and should open in the next several weeks.



@Neven: I'm confused about these two polls. There are a few reasons - to do with titles, Attention notices, acronyms and definitions - but, as explaining all of that is wordy and somewhat confusing itself, I'd be grateful if you could clear up the core issue for me first. :-)

As I understand it, the September minimum is the lowest daily extent in September. This is a "Highlander" value - there can be only one. The September average would be the average daily extent for all of September. Nothing to do with minimumity. I've got no idea what a September average minimum is. What set of minima gets divided by what length?

John Christensen

Hi IceBunny,

The outlook is for the Sept sea ice extent average.

'Minimum' just tends to sneak in there as we are looking at minima in September, compared to the maxima of March.

I hope this helps.


Minimum implies 'lowest'. A daily minimum is the lowest number reached on a given day. A monthly minimum is the lowest monthly average of the year. There is no contradiction here.

There are a couple of organisations that share their data in near-real time. In the past, we amateurs preferred to predict a daily minimum because we look at the Arctic on a daily basis, and who wants to wait a couple of weeks until September is over?

But at the same time the pros look at monthly averages, like for the Sea Ice Outlook, for various reasons. And so that's what the media reports on.

This is why we have two polls on the Forum. One for the standard NSIDC September average/minimum sea ice extent. The other for the JAXA daily minimum sea ice extent.

We could have picked Cryosphere Today sea ice area if reports would be correct (still not using the new satellite data). Or perhaps MASIE if it were consistent.

JAXA simply is very popular because it has been posting daily data for a long time now.

If you go to the Forum and read the opening posts for the polls, you'll see the minimums of previous years, and this tells you what the range is in which to vote. If you're keeping an eye on the Arctic, that is, and more or less compare it to previous years.

Rob Dekker

As beautifully shown by an animation from A-team on the forum :

the North West passage (the southern Amundsen route) is OPEN as per July 31 this year.

Jim Hunt

Rob - For an alternative interpretation see for example the (admittedly slightly out of date) Barrow webcam:


Another day, another large melt for 2016 season, 116K despite lack of compaction, despite no favourable weather for accelerating the process as with other summers. And now ECMWF announces a coming mini dipole lasting less than a week, but will offer some glimpses to further observe the action.

I keep on looking for "goodbye waves" , finding them pretty much wherever there is loose sea ice:



Nice image Jim, and many of us were expecting this coast free of ice by or around first of June.

Rob Dekker

Thanks Jim,
You are right. In the North-West passage, some remnants of ice remain. You pointed out Point Barrow (which really is open, except for the ice strip close to the coast) and Bellot Strait, which is still iffie.
But the ice in Larsen Sound that disappeared on AMSR2 on July 31 also re-appeared on the Aug 1, and is clearly visible on Modis Aug 2 :

This is a band of ice that is clearly at the end of its proverbial rope, but I agree that to have a true "open" North-West passage (as in : a ship can move through without touching ice), we need to wait a few more days.


August 2 120,801 km2 JAXA sea ice extent drop is more like it should be, there is great potential for more. However, there is an apparent solid ice bridge similar to 2007 which was not the creation of a great Arctic Dipole, but rather a lack of snow:


Rob Dekker

wayne, your note that " Areas with more open water can be traced where there was more snow," does not make any physical sense.

Also, areas with the most ice right now (the Laptev) did show the MOST amount of snow cover until deep into the melting season :

John Christensen

Sam, you said;

"It looks like the coastal route of the Siberian seaway will open in the next few days."

To the point made above by Rob, there seems to be too much ice in Laptev - and around the Sewernaja Semlja islands - for ships to pass safely, unless they are accompanied by heavy ice breakers.

John Christensen


I would partly agree with you: Since the snow cover acts as insulator, it may have reduced sea ice melting in Laptev.
However, due to the same insulation effect, it really depends on the timing of the snow fall: Heavy fall/early winter snow on the ice reduces sea ice creation, while late spring/early summer snow should reduce ice melting for some time.

Bill Fothergill


The sailing ship Northabout is currently attempting a West >> East crossing of the Northern Sea Route and then the North West Passage. You may be interested in the articles that Jim Hunt has been posting on his Great White Con blog on this topic.

Northabout is currently anchored in a bay on the southern flank of Ostrov Pilota Makhotkina in the Nordenskiöld Archipelago. This is about ~300kms from the pinch point at Cape Chelyuskin in the Vilkitsky Strait - i.e. the entry into the Laptev.


Bill Fothergill


Sorry, in my earlier comment, I thought I had included this link to the ship's blog as well...


John Christensen

Thanks for sharing Bill!


Hi Rob,

" Areas with more open water can be traced where there was more snow," does not make any physical sense."

It does make a lot of physical sense, and it is a well known fact from multiple observations wherever lake or sea ice forms. Thicker snow cover reduces accretion during winter. The June Rutgers map does show a great deal of surviving snow in the Canadian Greenland sector, some spots in the Laptev region, nothing like on the opposite side of the Pole. The Greenland/Canadian sector greater snow depth was gained during the long night, deeper snow is a proxy for sea ice, say there is 100 cm of it on top of new or old ice. It acts as if sea ice was much thicker, thicker sea ice does not accrete as much as new.

Kevin McKinney

Re the snow on ice thing, it connects with a point denialati tend to miss, which is that ice mass balance (if I can borrow a term from the glaciologists) isn't just a product of the melting season, but of the accretion season. Hence I've been told by such folk that Arctic temps can't possibly be affecting sea ice, because they haven't changed much during the summer.

Of course, there has been a highly marked *winter* warming, and of course that reduces ice growth rates during the freezing season, just as can deeper snow.


86.5N 120W weather buoy 48560 surface temperature is +2 C,


sst in the same sector scanned at -1, but there are 2 other arctic weather buoy surface temperature measurements of +1 C near the Pole in the 180 Longitude area. It is very warm up there.


correction :

"but there are 2 other arctic sst measurements of +1 C near the Pole in the 180 Longitude area. It is very warm up there."

John Christensen


Surface temperature of +2C seem to be perfectly normal, as north of 80N average is about +1.5C at this time of year.

Bill Fothergill

@ Kevin McK
"... a point denialati tend to miss ... Hence I've been told by such folk that Arctic temps can't possibly be affecting sea ice, because they haven't changed much during the summer ..."

That is indeed true Kevin. However, as I'm certain you're aware, such people have a propensity for not actually checking to see if their assertions are backed up by reality.

For example, using the "North Pole" zonal temperature anomalies from the UAH v6 beta5 dataset reveals the following trends (in degC/decade) for the 1979-2016 period...

Jan 0.30
Feb 0.22
Mar 0.26
Apr 0.40
May 0.34
Jun 0.26
Jul 0.17
Aug 0.15
Sep 0.17
Oct 0.15
Nov 0.21
Dec 0.19

Although the Jul-Oct trends are lower than the other months, they still manage to exceed the global annual trend (which is 0.12 degC/decade on that database).

As regards the effect of snow cover on ice formation, The NSIDC has this to say...

"Snow cover is one factor that dramatically alters the actual sea ice thickness calculated from the above formula. Snow is an effective insulator, slowing the transfer of heat from the ocean, through the ice, and to the atmosphere. Snow essentially slows the growth of ice."

The above quote is taken from...


Hi John

You forget, that is 80 Northwards average, so 80 latitude is naturally warmer than 87 for instance, especially when there is a good chunk open water at 80 n Fram area.


I agree with Kevin McK that less accretion in winter is very important in the decline of ASI. A look at Jim Pettits volume graph at the ASI Graphs page in which he depicts Annual Maximum and Loss, and Ice Remaining at Minimum demonstrates very clearly that the overwhelming majority of sea ice decline stems from less accretion in winter.
The trend lines in this graph also make clear that we will practically for sure have a blue ocean event in less then five years. It really takes not too wild an extrapolation to state this.

It sure is fun to theorize and hypothesize about the different factors and their effects during melt season. But summer ice melt has just marginally increased in the last decades. The real damage comes from less cold in winter.

That's why i also agree with Wayne's repeated statements that this years minimum will at least be top 3, if not record low. Even a very regular melt season suffices when you start out with such low volume numbers in april. Just look at Jim's graph.


Hey Neven, I wanted to put a vote in for Sept extent at ASIF, but no activation email has been sent when I tried to register at the forum. Clicked to get another email sent. Nothing comes after half an hour, not even in the junkmail section.

If it's ok with you, I'd like to register my vote for the 4.25 - 4.5 range.


Activation emails came through an hour after submission. I've registered my vote at ASEF. I'll try not to pester the board with too many questions.


Another day another 116,632 km2 extent drop. Its no longer a question of top 3 minima, but top 2 and beyond. As well as North Pole just about to get looser pack ice, the Trans Polar Drift seems robust as well:


Rob Dekker

wayne, I'm sorry. I thought you were talking about land-snow, which held out longer along the Laptev coast this year.
Regarding the snow-on-ice graph that you present, it is indeed interesting that it matches the ice-melt pattern this season, at least when comparing the ESS and the Laptev.

But if snow cover on ice really would make a substantial difference for ice thickness, should we not see that reflected in the ice-thickness graphs from observations like Cryosat ?

Here is that Cryosat graph for March'16 :

I do not see much of a difference between the ice thickness in the Laptev and the ESS.

If anything, it seems the Laptev (which received the LEAST amount of snow in you graph) has THINNER ice than the ESS.

May I suggest that the earlier melt of the ESS (than the Laptev) this year was caused by some quite significant "torch" events in June and July ? I think you even reported on that as well.

Long story short, I don't see that "snow cover" on ice has affected the melting season in any significant way this year yet.


People are,.. and have been for a long time, reacting in Australia.

I dare say Malcolm Turnbulls new government- who only just scraped in by the skin of their teeth owing to what must be the safest psuedo-conservative seat in the world (That being Julie Bishops in Western Australia,... bet they don't even get fibre internet either, lol)- have as of today been shamed into reinvigorating their previously poorly considered stance on the importance of climate change no doubt because of this: http://www.greenpeace.org/seasia/ph/PageFiles/735232/Climate_Change_and_Human_Rights_Petition.pdf

It's well known those of the relevant Government Departments around the world are very concerned that the multi-year sea ice of the Arctic could disappear this year. Maybe I should have said 'were' because this was the water cooler conversation six months ago.

The previous water cooler conversation was of course that the Turnbull/Abbott/Turnbull government was woefully undercooked and it would seem the situation has not changed.

Who remembers the mining companies who took down Heavy Kevvy? This global democracy is owned by carbon incorporated: it's called Jevons Paradox and the elite are not legally obliged to share.. (I ask you: exactly who's standing armies support all this fascist behaviour?)

** Now THAT is intergenerational theft!!

Rob Dekker

Just to be clear, I've gone on record to declare that :

There is a 92% chance that 2016 will end up in the top-3 (only 2 out of 24 years showed a larger difference between 2011 and this 2016 projection).

There is 66% chance that 2016 will be second place (after 2007 but shy of 2012).

There is an 8 percent chance of 2016 beating the 2012 record in September.




"Here is that Cryosat graph for March'16"

I've noticed that, there seems to be clear hints that there was a thicker area of sea ice where the Komsomolets ice bridge is, but not as prominent as showing now, areas of more snow had also generally thinner ice in except for the ridging and shore zones. Note the extra ice thickness North of Wrangel Island. Quite the interesting study to see if Cryosat data was accurate, or picked up the extra ice but not completely.

"There is 66% chance that 2016 will be second place (after 2007 but shy of 2012).

There is an 8 percent chance of 2016 beating the 2012 record in September."

74% probability for at least #2 sounds good for now, but these odds increase day by day. Was still quite a swell projection then.

John Christensen

Based on DMI ice volume and ROOS sea ice area, it would seem that one of the reasons for the current low extent numbers, is that the pack is undergoing compaction, in addition to melting.

John Christensen

And a sure sign we are now in August; surface temps are below zero in most of the HP area, while above zero in the LP area:


Rob Dekker

Hi John,
NOAA presents a temperature profile over the Arctic which puts the low temps over the remaining ice rather than the LP/HP areas from DMI's graph :
Which one is more accurate ?

John Christensen

Hi Rob,

The temperature and air pressure on DMI 60N is a model forecast for the present time, updated multiple times a day, while NOAA shows a one-day average temp with some delay, which could explain the differences.

From DMI:

Surface Pressure and 2 meter air temperature over sea ice and ocean, North of 60°N.
Data are based on analysis data from the global operational weather forecasting model at ECMWFs. The model analysis can be considered to be a 0-hours forecast, and is the best estimate of the current state of the atmospheric system and it is based on available observations and the model physics.

John Christensen

However, that is just small talk: It's PIOMAS time..

Andy Lee Robinson

Indeed it is PIOMAS time!

Average July volume
2010 10.153
2011 9,457
2012 9,177
2013 10,446
2014 11,880
2015 11,550
2016 10,172

Slipped to a close 4th place, only 995 km³ behind 2012


John Christensen

I would not have used the word 'slipped', but let's wait for the thread to be posted to discuss volume.

Rob Dekker

Report from the Forum, by BFTV :

A drop of 188k takes the daily NSIDC extent down to 2nd lowest on record.

It also puts the 7 day loss up to 942k, and means that we only have to average a 138k drop over the next 2 days to achieve a mega melt week (>1,000,000km2).

That kind of sums up the current developments in Arctic sea ice.

Bill Fothergill

RE: The comment from BFTV on the forum just mentioned by Rob.

When I read BFTV's comment yesterday, I had a quick look at the NSIDC's "Charctic" 5-day figure. That 188k drop moved 2016 slightly lower than 2011 and slightly higher than 2007 for the equivalent period.

If my maths is correct, a 75k drop today in NSIDC's one-day figure would result in the 5-day figure overtaking (undertaking???) 2007. However, as most people here realise, it is not uncommon for a large drop in the NSIDC daily number to be followed by a small uptick.

Yesterday's 117k drop in the ADS/IJIS figure was followed this morning by a less "impressive" 47k drop. As ADS/IJIS use a 2-day smoothing technique, this implies that the difference between the most recent Day N value, and the Day N-2 value that it just replaced in the 2 day smoothing was "only" about 94k.

John Christensen

From the Forum:

"It also puts the 7 day loss up to 942k, and means that we only have to average a 138k drop over the next 2 days to achieve a mega melt week (>1,000,000km2)."

This comment blurs the difference between sea ice extent changes and sea ice melting, two different aspects of assessing sea ice changes.

Firstly, the margins of the pack have very low ice concentration, so very little actual melting can result in significant sea ice extent reductions.
Secondly, the sea ice moves around, which also is not melting.

A 'mega melt week' would be a week of excessive sea ice volume loss, very simple.
Looking at volume changes, it seems the past days have been about average.


Whatever happened to the Navy gif?



Rob Dekker

If 1 Million km^2 disappears from the NSIDC daily extent picture, there is no shame in calling that a "mega melt week".

The 942 k drop that we just witnessed came very close to that, despite other graphs and other data out there.

Nuf said.

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.)