2016 Mega-Dipole

As reported in the last ASI update, the forecast Dipole has now set up. Yesterday's last analysis chart provided by Environment Canada, at 18Z, showed the high pressure moving in via the Bering Strait at 1037 hPa, while a rapidly weakening storm that had moved in from Siberia (lowest central pressure was 967 hPa, two days ago) was still at 974 hPa on the other side of the Arctic. This then evolved to a massive 1041 hPa on one side of the Dipole, and 977 hPa on the other:

That's a humongous pressure gradient of 63-64 hPa! You don't see that often around this time in the Arctic, a not so nice story for the grandkids. Right now, the high is still at 1041 hPa, but the low has weakened further to 980 hPa. That's still a pressure gradient of 61 hPa.

It means very strong winds are blowing over the ice pack, from Siberia to Canada, compacting the ice pack and further pushing open water towards the pole, as can be seen on this animation of Uni Bremen sea ice concentration maps:

If one looks carefully at the ice in the so-called Wrangel Arm, you can clearly see the ice flash in and out of existence. There comes a point where the ice no longer flashes back, but in the meantime it makes for wild swings on the CT sea ice area graph, as calculated by Wipneus:

The 2016 trend line has already dipped well below the 2007 minimum. It looks highly likely that 2016 will end up second on record for this data set.

Swings are less wild on the JAXA sea ice extent graph, but there have been some big drops there as well in the past couple of days:

2016 is now second there too, but still has to drop 339K to go lower than the 2007 minimum (it already is 5th lowest at the moment). Of course, the minimum is reached later for extent, and in the two weeks that are left most of the ice in the Wrangel Arm will have to melt out, and the ice pack needs to be compacted further for the second place to come in sight.

The ECMWF weather forecast model looked much worse the last few days, but has improved somewhat. Here's what's in store for the coming 6 days (click for a larger version, and go to the ASIG Forecasts page for daily forecasts):

The Dipole will persist for another 2-3 days, but the high gets stretched out in the process and is then pushed aside by a weak low. So, in a few days the game of guessing when the minimum is reached will start. We'll have to see how much lower extent and area can get due to compaction and the melting out of the Wrangel Arm.

Here are two more images I'd like to share, both coming from the Danish Meteorological Institute. The image on the left clearly shows the size and intensity of the Dipole. The image on the right shows how for a brief time temperatures above 80N went above zero again, as the Dipole sucked in warmer air from lower latitudes:

And finally another illustrative animation by ASIF commenter Thawing Thunder showing the anatomy of the Wrangel Arm:

Comments

[NickWhalenMP]

If there is a silver lining, the ice pack on the Canadian side should be nice and thick following 4 days of winds. The DMI daily temperature chart is very telling. That graph could be unrecognizable in a summer with no sea ice.

[NickWhalenMP]

Thanks, Neven, for doing all the work telling this sad story.

[navegante]

Hello Neven, thanks as always. Small typo, there is a "941 hPa" that should read "1041 hPa"

[Jim Hunt]

Neven - The Northern Sea Route actually opened a while ago, in so far as the yacht Northabout recently managed to complete it without assistance from an icebreaker:

http://GreatWhiteCon.info/2016/08/northabout-races-for-the-date-line/

The Vilkitsky Strait wasn't exactly "plain sailing" though!

[Neven]

I agree, Jim. I meant it is open as far as Uni Bremen AMSR2 data is concerned.

Thanks, navegante. Fixed the typo. Imagine that, 941 hPa. Maybe one for the future? ;-)

[Cato Uticensis]

Agree 100% with your post Neven. PS: the pics of the Wrangel arm are just fantastic :)))

[wayne]

Neven

I called it an ice bridge or pan handle, but arm is ok, and it has showed interesting features:

http://eh2r.blogspot.ca/2016/08/wrangel-island-ice-bridge-no-more-its.html

Also High's of the 1050 mb's or more are common in winter at times.
It is a mega dipole if it lasts long, but the damage was already done.

[navegante]

That final frame with all the annotations is so cool. This season has been really special, nothing like any other before. You'll know better.
Wiyhin all the annotations I only miss the "cookie monster" over the Atlantic front :-)

[Bill Fothergill]

@ Neven & A-Team

A-Team had obviously meant to post a comment on this thread, but it has inadvertently gone on the "Update 5: Big Cyclone" thread instead.

Perhaps one of you could either shift it or re-post it?

[wayne]

101K drop yesterday, quite the feat considering the lateness of the season. The most loss ever from now until minima was 2012 at 446K, and I seems quite certain that this number will be matched or exceeded, meaning 2016 is comfortable 2nd for now , at about 3.7 million come mid September.

http://eh2r.blogspot.ca/2016/08/atlantic-front-retreating-likely.html

ode to the JAXA staff and their software deciphering Goodbye Waves from real sea ice. It seems not an easy task. The Atlantic Front is also retreating Northwards.

[viddaloo]

I think the main reason Neven and others got this melt season so wrong, was that he ignored annual average extent/volume ("at his peril" I might say, with wayne) and looked only at what made previous low years stand out, like 2007 & 2012. 2016 found its own way to go record low, and while the exact way wasn't obvious to anyone, it WAS rather obvious we would end up in the Top 3.

Now, 2016 needs only drops of 2k on average down to the 2015 (3rd lowest) minimum in order to make the Top 3 for annual minimum.

Big: http://i.imgur.com/S4nKtvw.png

Of course, 2016 *is* already lowest and has been every single day so far. But that's just semantics: 2016 is a calendar year. As such it was lowest from start to finish.

[NickWhalenMP]

How much more of this wind http://ocean.dmi.dk/arctic/icedrift_anim/plots/icedrift.arc.d-00.png

Before we are talking about worse than 3rd lowest extent. Very glad this didn't happen in June!

[navegante]

Viddaloo, so Neven got it so wrong spending the time learning and analyzing fundamentals, while you can foresee the Arctic sea ice future correctly by consulting a couple of charts. That is called tarot. In economics that is called gambling, easily leading to fraud.

[Neven]

Perhaps one of you could either shift it or re-post it?

Bill, the Typepad software doesn't allow me to move comments like that.

As for my good friend, absolute genius, and judge of all, Vidileo Vidilei:

If your annual average oracle has been lowest all year, then why doesn't this year's minimum crush the 2012 record? Or maybe I shouldn't ask about physical details when it comes to black-and-white sweeping stats?

I will readily admit that I personally didn't expect the pre-preconditioning momentum would take things so far, given the lack of melt ponding during June and July. And of course, I couldn't foresee with any certitude the spectacular end to the melting season, due to the GAC (or close to it) and the current Mega-Dipole.

But I never ruled out anything either. You can read back the introductions and conclusions of all ASI updates this year. I always explain the range of possibilities, even if they seem far-fetched at that moment.

In fact, on May 27th (2016 ASI update 1) I wrote:

Considering what has happened so far, a new record minimum is a distinct possibility though. If a lot of melt pond formation occurs this month and the next, it will take some extremely cloudy and cold weather during July and August to prevent records from being broken. And even then the minimum will most probably be among the lowest on record.

I don't understand why you're so intent on smearing my reputation? Am I not alarmist enough for you? Why not go to the Sam Carana blog then? It's much more your style. Especially your extrapolated zero ice on August 1st 2023 will be popular there.

[D]

This summer, because of the stormy Arctic weather, a little more heat went into the subarctic seas and a little less went into the Arctic waters, relatively speaking. The strong thermal gradient has intensified the storminess as summer transitions to fall. Perhaps a well constructed long-term forecast model could have picked this up. NOAA's CFS v2, however, has been unable to forecast the tropical Atlantic and the southern ocean so it cannot be trusted.

However, we can always make simplistic forecasts based on our pet theories then say stupid stuff about Neven. I chose not to because I value my reputation. Clearly, there are others here who don't care about their reputations.

I am hoping that the professional researchers who have invested their lives into understanding the Arctic can build better models that can more accurately predict the behavior of the polar climate systems. Until then I humbly observe what's happening with amazement.

Thanks, Neven, for the excellent graphics and persistent reports.

-Fish

[viddaloo]

Neven,

I'm truly sorry if I may come across as wanting to smear your reputation. That was never my intention, and I hope you'll accept my apologies.

I agree to what you say that your ice updates have been carefully worded to account for all (or most) possibilities, and that it's merely in the tendencies of those posts that you may have aimed for an extent minimum too high in relation to observed reality.

I think we all of us will learn tremendously from this 2016 ice melt season, and I do agree with everyone who's said this year was truly exceptional and extremely exciting.

PS: Only the 2016 year–to–date average extent has been lowest ever every single day from January 1st through August 30th. The 365–day annual average extent has just recently gone South, in the 2nd half of this year.

PPS: I'm also very thankful for your many blog posts and updates, which truly make the ice–melt year interesting and spectacular.

[www.facebook.com/profile.php?id=100000035525134]

" I always explain the range of possibilities, even if they seem far-fetched at that moment."

de lurking..

That captures EXACTLY why I read you every day. It's what I expect from a good analyst and you never fail to deliver..

re lurking

[Artful Dodger]

Hi folks,

For the exotourists among you, the Northwest Passage cruise ship Crystal Serenity is about to shoot the Bellot Strait, a 2 km wide 25 km long strait connecting the Gulf of Boothia on the east to Peel Sound on the west.

Here is the current view from the bridge (click the image to see it full screen):

Also take a look at the scenery to the port (left) and starboard (right) as the ship makes the passage, which should take at least 2 hours. Enjoy!

Cheers,
Lodger

[L. Hamilton]

At a Polar Prediction Workshop held at Lamont-Doherty last May, well ahead of the melt season, I ran an informal poll asking 35 attendees for their guess about the September mean extent. The average among those guesses, 4.14, turned out to be strikingly close to a simple quadratic projection summarizing the downward trend to date, 4.15.

This result became a "heuristic" contribution to the SIPN Sea Ice Outlook this year. That graph seems worth reposting as food for thought, now at the start of September.

http://img.photobucket.com/albums/v224/Chiloe/Climate/Workshop_prediction_2016_2.png

[Neven]

I hope you'll accept my apologies.

Okay, fair enough, Vid. And I see your annual average stuff as a useful source of information, a piece of the puzzle. That's the reason why I've always given you some leeway.

As for the influence of the pre-preconditioning (early opening up in the Beaufort, rapid snow cover retreat, etc) on how this melting season will end up, I'm still not sure how much of it was also influenced by ocean heat flux, an element that is impossible to get a handle on.

I mean, if I look at SST anomalies, it's as bad or slightly worse even than in 2012, but things were rather cloudy when the Sun was high up in June and July. So where did all that heat come from?

Mind you, it's ocean heat flux that got Maslowski's model to predict a 80% drop from the 1979–2000 summer volume baseline somewhere between 2013 and 2019.

[Jim Williams]

Eventually the Atlantic and the Pacific (mostly the Atlantic) are going to snuff the Arctic, and when it happens it will come in the form of Dew Point. The only question is when, and how to tell....you can tell when the ice minimum extent runs a lot late, not with the minimum extent itself.

I think it is this year, but who knows? (My guess for this year is based upon how warm last Winter was.)

[Andy Lee Robinson]

Lodger, thanks for the image - not enough ice there to make even one G&T!

These last few days are quite extraordinary. As the ice thins, it really can be a case of "here today, gone tomorrow".
As the ocean surface becomes exposed and can evaporate, it could help fuel cyclonic activity, with the extra water vapour helping to keep the temperatures up and delay refreezing.

2016 is too soon for a new record, but the forcing is ramping up, cumulatively and inexorably.

It seems to me that the possibility of an ice-free Arctic is approaching, and dependent on what dice the weather throws.

As the climate probability distribution shifts to the right, then what was previously improbable becomes more probable, and what was previously *impossible* becomes possible.

I'm looking forward to making more animations when the data are published!

[viddaloo]

Interestingly, every single year since (at least) 2005 has been a Top 3 year at the time for JAXA minimum extent, with only two exceptions: 2013 & 2014. This makes it easier to pinpoint the exception to the downward trend, and predict that a year will NOT make the Top 3.

For 2014 it's easy: March 1 is 6th lowest in annual average extent (AAE), and the 1st of the next months are 5th, 5th, 5th, 5th, 4th & 5th lowest. Result: 7th lowest at minimum.

2013 was harder to predict: March 1st AAE was lowest ever, along with the 1st of every following month including September. The only purely AAE clue to its result at 6th lowest at minimum was the persistent shift from lowest ever and falling (AAE–wise) in March to May, and over to lowest ever and rising in June to September.

2016 didn't look like any of these exceptional years: Its AAE was falling on the 1st of every month March–September, starting out 3rd lowest in March via 2nd in April–July to lowest ever AND falling in August–September.

That's why the Top 3 status looked like a slam–dunk. Exact positions — lowest or 2nd @ minimum — is much harder to tell, and anyone's guess, really. I personally think it will be a new record low in October.

[wayne]

Integrating every bit of knowledge and observation is essential into achieving a good projection. What is still missing is the holistic approach, which dictates 2013 as being a very big melt despite extent numbers saying otherwise. This explains 2016 in some ways.

[Artful Dodger]

Szívesen, Andy Lee!

Did you see the GINORMOUS iceburg off the port side of the C.S. on the far-left of the 14:45 PM port-side image? Must be 10 m in the air, so wot 100m keel to sail? I'd like to know which glacier calved that!

I couldn't find a Hungarian version of the Stan Rodger's classic "Northwest Passage", but I hope you enjoy this Polish rendition I posted earlier: (you must click on "Kelvin" in the linked image to go to the song on Youtube).

Cheers,
Lodger

P.S. This MUST be a new RECORD:
Seven (7) surface lows in the Arctic Basin/Greenland/Nordic seas simultaneously! Hurray in, this image in this link will be over-written after 18z hrs tomorrow!

[Artful Dodger]

Pardon, Dingos ate my Baby...

18Z EnvCda N.pole

Cheers,
Lodger

[Rob Dekker]

Larry Hamilton said :

At a Polar Prediction Workshop held at Lamont-Doherty last May, well ahead of the melt season, I ran an informal poll asking 35 attendees for their guess about the September mean extent. The average among those guesses, 4.14, turned out to be strikingly close to a simple quadratic projection summarizing the downward trend to date, 4.15.

Thanks Larry for running that poll.
It would be interesting to see how close the final number ends up to that 4.14 from the poll, and the 4.15 from the super-simple quadratic projection. I would like to add that I feel in good company, since my own projection (based on June land snow cover, ice concentration and ice area) also projects 4.1 this year.

I noticed before that simple (linear or quadratic or Gompertz-curve) projections tend to be no less accurate (measured in standard deviation) than the most complex GCM based projections, and I'm simply not sure what to make of that.

But either way, after September is over, and the final numbers are in, it would be really interesting to see which projections (from SIPN and also the various polls in the forum) were spot-on and which were way off, and which methods were used for each.

After all, as Feynman once stated about models :
""It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong."

[AbbottisGone]

Rob said:

"I noticed before that simple (linear or quadratic or Gompertz-curve) projections tend to be no less accurate (measured in standard deviation) than the most complex GCM based projections, and I'm simply not sure what to make of that."
<<

It means, presumably, that nothing has gone into unpredictable tipping point territory yet. If you talk to old engineers then they say, "..oh, it's all a hoax and everything obeys the predator-prey yadda yadda yadda.."

Reading these forums has scared me in the exact point that people seem to be saying PIOMASS is directly heading toward nothing... are we infact concluding that that is an exaggeration?

** The point about experiments is curiously missing from all this 'climate science' is it not, btw? (I'm not being skeptical of course!!)

[wayne]

Well done Rob

but is this 4.1 from JAXA or another metric? I only follow JAXA from habit and respect for accurate work extremely well done at the limits of our potential. If it is JAXA, 4.1 is too high!!. Sorry...

[L. Hamilton]

Rob Dekker:
"I noticed before that simple (linear or quadratic or Gompertz-curve) projections tend to be no less accurate (measured in standard deviation) than the most complex GCM based projections, and I'm simply not sure what to make of that."

Julienne Stroeve and I have a paper recently accepted for Polar Geography, which could hit the street later this fall. The paper, with catchy title "400 Predictions," includes a comparison between median SIPN Sea Ice Outlook predictions over 2008-2015, and three naive models: linear extrapolation, quadratic extrapolation, and "persistence" (guessing that this year will be same as last). The median SIO predictions outperform all three naive methods, although sometimes not by huge amounts.

This and other analyses in the new paper confirm what we concluded a few years before: "Sea prediction has easy and difficult years." For an un-paywalled & public-friendly version of the argument, including unique additional data from a contest to win ice cream, see:
https://www.academia.edu/7504534/Sea_ice_prediction_has_easy_and_difficult_years

An abstract for our more formal Geophysical Research Letters paper is here:
http://onlinelibrary.wiley.com/doi/10.1002/2014GL059388/abstract

What all three papers report is that years where September ice extent is near its long-term downward trend (climate), many different prediction methods look good. Years with abrupt excursions above or below that trend (weather), most prediction methods look bad.

And, looking good (or bad) in one particular year does not necessarily forecast how well your method will perform in the next.

[L. Hamilton]

Our forthcoming paper concludes:

Thinning ice that is sensitive to summer weather, complicating prediction, reflects our transitional era between a past Arctic cool enough to retain much thick, resistant multiyear ice; and a warmed future Arctic where little ice remains at summer’s end.

The 2016 season, which was not part of our analysis, seems to show that dramatically.

[wayne]

Thank You Mr Hamilton

Do most sea ice prediction models depend on PIOMAS volume? If so does PIOMAS regard 15% threshold for sea ice extent or area as acceptable for 100% sea ice coverage? The summer weather bit is not as important anymore compared to warming GT's, SST's and especially over all thinner sea ice. Lesson learned especially this melt season.

[L. Hamilton]

Wayne, the 400-odd prediction use a wide range of different input data and methods, including some from blog contributors here. Our paper classifies these into 5 broad types:
- heuristic
- statistical
- mixed statistical-modeling
- ice-ocean modeling
- coupled ice-ocean-atmosphere modeling

The statistical and coupled models performed somewhat better overall, but group differences are not large.

Basically our conclusion echoes something Don Perovic remarked last year at a conference (I'm paraphrasing):
"Ice used to be easy to predict, there was always a lot. In the future it will be easy again, there won't be very much. Right now we're in between and it's hard."

[Robertsekeris]

Maybe Cryosat 2 (see them largely being ignored) will provide alternate ice-volume info, since it's scheduled to resume the measurements in September (suspended because it can't handle melt-ponds).

Go: http://www.cpom.ucl.ac.uk/csopr/index.html and http://www.cpom.ucl.ac.uk/csopr/seaice.html

[wayne]

Thank You Mr Hamilton

Mr Perovic was right on, it is difficult, perhaps the hardest thing to predict. Therefore all the more fun. A greatest challenge would be to interpret or classify the melt states, which vary season to season. For instance statistics fail to 'see'. Integrating the quality with numerical quantity is key.

Best of luck with your work and Mrs Stroeve hope to see you in the wide spectrum of medias more often than the latest gossip about nothing news...

[viddaloo]

Xmas came early this year to Hardanger.

Listen to last night's book launch interview with Dr Wadhams here: https://extinctionradio.files.wordpress.com/2016/09/57-wadhams-09-01-16.mp3

[navegante]

The mega-dipole is gradually rotating in direction. It may bring warmth from the Eurasian continent and more wind stress on the ice, but also dispersion of the recently compacted pack and coldness to the Pacific sector of the Arctic.
What could we expect out of all this?

[Robert S]

I think that as long as we continue to get significant wind, wave action will continue to degrade the vast areas of vulnerable ice out there. Right now the best case scenario would be little or no wind, which would reduce degradation and water mixing, and potentially allow ice to begin reforming as temperatures drop. But the forecast continues to show wide areas of 20+knot winds for the next several days, and across the available fetch wave action will be considerable.

[Rob Dekker]

Abbott said :

The point about experiments is curiously missing from all this 'climate science' is it not, btw?

While it is true that humanity is running one single 'experiment' (there is no Planet B), the issue of experiments depends on the question you are asking. For example, when we are asking to predict September sea ice using data from earlier months, every year is a new experiment.
Since we have consistent data since at least 1979, we have run 36 experiments already.

Any model projecting the outcome of the next 'experiment' (Sept 2016) can and should be tested for past performance. And if it can't predict the past, it should be discarded.
That is the nature of Feynman's statement.

[Cato Uticensis]

I think that, once more, this melting season has provided a very interesting lesson. Talking about "worst case scenarios", I think this year has put together an impressive series of negative factors that in the end by far offset the lack of pre-conditioning, namely:

1) very early melting leading to inevitable sea water warming and associated huge positive anomalies for seawater temperature

2) Persistence of LP systems leading to ice dispersion

3) Very windy conditions leading to water-ice mixing, higher exposure of ice to warm water, bottom melting etc.

4) Heat surplus in ocean waters at the beginning of the melting season after the huge el nino event

5) Worst possible dipole configuration at the end of melting season with strong and warm winds from Pacific ocean, very long fetch, huge compaction and Fram export.

The only favourable event was represented by cloudy conditions which limited radiation, but overall in my humble opinion this has by far been offset by all the other factors. Overall, in my view, one of the most unfavourable combination of events throughout the melting season. One more lesson learned I guess.

[Cato Uticensis]

PS: IMHO we are very close to the minimum. The synoptic configuration in the next few days is leading to ice dispersion and a relevant cooling over the Arctic and less transportation through the Fram.

See what happens. I guess there's not much to decide, regarding the outcome of the competition with 2016 gaining the second place of the podium, after 2012.

[viddaloo]

From yesterday's interview with Peter Wadhams (mp3 link in my previous comment): Peter said IPCC climate modellers predicting slow climate change up to 2100 simply didn't want to look at the very real things creating rapid and abrupt temperature changes. This is a very common theme when polluticians (pardon my bad spelling skills) order a science review for some atrocity they are planning to do to Nature: The science experts who get the assignment normally KNOW what sort of answer the polluticians want. And what direction that answer ought to point in. "Street–wise" experts adjust to these wishes from the buyer, and deliver their "scientific results" in a manner that they'll get more assignments from that same buyer later.

In our case: Scientists know that the polluticians want to hear climate change will be slow, and manageable by 2100. Fast permafrost & seabed feedbacks are then routinely kept out of their estimates.

[Cato Uticensis]

Wadhams has been predicting disappearance of arctic ice for the last few years, every year. All the newspapers of the world have been flooded with his forecast of ice disappearance in September 2016. He is not even respected by his own colleagues.

His tendency to overindulge in announcements of imminent disasters cast discredits over all the people who take climate change seriously, and professionally.

[wayne]

I knew Peter a while ago, he sounds the same, glad to hear he is doing fine. Predictors usually have a wide range of failures before they get it right. this is a normal scientific process. However, it is very hard to conceive ice free Arctic Ocean until winters get a lot milder/cloudier/snowier. This said, this year is becoming the new Arctic look in September. All solid sea ice left is against the Canadian Arctic Archipelago. Tide driven every day of the year no matter what weather can do. SST's would have to warm up a lot and somehow be driven under this area ridged sea ice.

Speaking of this year, today's JAXA 4.16 million, the number close to # 2 minima, looks a whole lot better than the actual AMSR2 map. There is vast icefields with very little sea ice per grid. The remainder dense ice field is the new look, everywhere else will be gone regularly each year.

[D]

Several points are respond to below:

Slowly but surely the Arctic ocean is becoming a vapor source for NH precipitation. An excess of deuterium to Oxygen 18 has been increasing in New England as the Arctic water vapor source grows.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4789600/

The warming of the waters in the Barents sea is affecting global atmospheric circulation patterns according to other papers. Ultimately, there is a water vapor feedback which is a major contributor to "Arctic amplification" and sea ice loss.

I suspect the water vapor feedback effects are one reason this year has seen such a dramatic loss in sea ice despite the cloudy weather.

---

A coupled model that makes bad forecasts but can be improved is worth working on because it may ultimately perform better than other models. A heuristic or statistical model which is not subject to improvement should be simply disregarded when it falls behind other models. Some statistical models may be subject to improvement but how do you improve an heuristic model that fails?

---

There will always be a range of scientific predictions. Wadhams is now at the extreme end of the range predicting catastrophic climate change. However, there are a few scientists on the opposite extreme who think CO2 is good for plants and that a warming climate is an improving climate. The three percent on the top and bottom ends of the distribution do not reflect on the quality of the science of the 94% of scientists who fall in the mainstream.

In fact, Wadhams has a far better case than the bottom 3% that think CO2 is beneficial plant food. One of these years he is going to be right. We are just arguing over the rate of change, not the ultimate result, zero ice.

-Fish

[Rob Dekker]

I'm sorry that I came across quite harsh with my Feynman quote and statement that climate models need to be "discarded" if they can't even predict the past.

What triggered this is my discontent with many IPCC GCM models and their inability to predict Arctic Sea Ice development.
Since 2007 it was already suspected and by 2012 is was absolutely clear that we should discard a good number of these models since they simply do NOT predict even the past few decades with even the slightest degree of competence :
http://neven1.typepad.com/.a/6a0133f03a1e37970b017744cf5360970d-pi

And it was not just ice extent, but also ice volume is heavily underestimated :
http://www.realclimate.org/images/for_rc_fig3_ts.png

These are the models that appear in the IPCC reports and, as prof. Wadhams again clarifies in his interview (above), they create the false impression that Arctic sea ice in summer will disappear somewhere at the end of this century, while in fact the trend of volume decline suggest that summer sea ice in the Arctic will be gone in the next 5 years or so. In fact, the past volume trend line hits virtual-ice-free around 2016, and we may have just been saved by two outliers (2013 and 2014) and perhaps saved by the bell (cold June and July) in 2016 as well.

Yes, these models can be improved. And it is not even unclear where they need to improve. For starters, they should get (the trend) in loss of land snow cover right, since that impacts sea ice cover very significantly (as the statistical models suggest). Also, some of these models don't even have melting ponds.

I suggest that the GCM IPCC models get a frank review, and that we discard the ones that cannot even predict the past couple of decades.
THAT is what I meant with Feynman's comment, and I believe that we do mankind a great favor is explaining how much more serious the developments in the Arctic are than the currently used GCMs suggest.

[Rob Dekker]

Cato said

He is not even respected by his own colleagues.

Where does this remark come from ?

[www.facebook.com/profile.php?id=100000035525134]

"Yes, these models can be improved. And it is not even unclear where they need to improve. "

well, you its hard to get Ice right as long as
absolute temperatures are off..

Here is the spread in models, not in ANOMALY, but in absolute temperatures.

http://www.climate-lab-book.ac.uk/files/2015/12/hawkins_sutton_bams_fig1.jpg

this is for GCMs..

[michael sweet]

At RealClimate they have an arctic sea ice link at the start of the new monthly thread that links to this blog!

I rarely post but I read all the posts on the blog. I especially appreciate the OP's by Neven. They are more even handed than many of the posts by others. It is remarkable how Neven has informed himself about the science. As Neven often points out, the Arctic sea ice has many surprises that we will not anticipate. Last year was low melt but this year is going low at the end. How much longer will it go lower?

[Kris]

Michael Sweet asked:

How much longer will it go lower?

Quite a bit, alas. Do Check Reanalyzer --> Arctic --> Precipitation and clouds. You'll see from the 9th september on there will be a severe attack from the Atlantic side. Very strong warm winds and huge rainfall from Fram strait till deep into the Arctic, winds even reaching Wrangle Island ...

[viddaloo]

#3 for low volume so far.

Big: http://i.imgur.com/Z5TOKJu.png

[L. Hamilton]

D:
"Some statistical models may be subject to improvement but how do you improve an heuristic model that fails?"

Heuristic models are a catch-all category of course, but in principle some of them can learn and improve.

An interesting example of a heuristic prediction is the poll of 35 scientists at a polar prediction workshop that I mentioned upthread. The "collective wisdom" of workshop participants presumably integrated, informally, their experiences from past years. The mean of their informal predictions (in May) was 4.14, which still looks reasonable while some more formal predictions have been passed.

Of course, replication is the real test. I'll be trying this again as opportunities arise, perhaps as soon as this month (looking toward 2017).

[L. Hamilton]

Regarding Peter Wadhams, the 2015 Sea Ice Outlook (where he offered the same low prediction in June, July and August cycles) illustrates his standing as an outlier among sea ice researchers.

http://img.photobucket.com/albums/v224/Chiloe/12_Climate/SIO_2015.png

Of course, being an outlier does not mean one is wrong, but in that year Dr. Wadham's prediction (0.98 million km2) was far off the mark.

[Bill Fothergill]

@ D/Fish

Well George, that paper about isotopic ratios certainly forced the old grey matter out of neutral!

There are two sentences in the opening paragraph of the Abstract which state that ... "We found a significant reduction in δ Oxygen18 and δ Deuterium values over the 43-year record, coupled with a significant increase in d-excess values. This gradual reduction in δ Oxygen18 and δ Deuterium values unexpectedly occurred during a period of regional warming."

That stopped me dead in my tracks until I managed to think it through cursively. (That does not mean that a lot of swearing was involved!)

In a warmer world, we would expect to see more Oxygen 18 and deuterium present in precipitation as more energy becomes available to aid evaporation. However, regional variations will most certainly occur.

In this instance, the loss of sea ice has created a larger ocean area in which evaporation can occur. As the saturation vapour pressure is higher over water than over ice, additional evaporation is going to take place.

However, and this was the stumbling block for me, this recently exposed ocean surface is cooler than most of the rest of the nearby ocean. That means that fractionation is going to work in the opposite direction, in that the lighter isotopes will be preferentially evaporated and then precipitated.

A molecule of "normal" water (good old H20) effectively has a mass of 18 Atomic Mass Units (1+1+16). However, a molecule with a deuterium atom will be 19 AMU and one with a heavy oxygen will be at 20 AMU. These mass differences would therefore explain the observation that there was a reduction in both isotopes, but the heavy oxygen was even more affected than the heavy hydrogen.

It would have been interesting to see the results had they also checked the before/after Oxygen 17 ratios.

[wayne]

Ha the models, such a nebulous simplification of a word for algorythms, all different, all applying their own interpretation of how sea ice should behave. Yet we are in the dark, I don't know which model integrates PIOMAS data, I don't know if PIOMAS model uses the 15% inane rule as for extent or area. Imagine if it does, say one grid has 16% sea ice considered as 100%, how much of a miscalculation of volume is that?

As far as predictions are concerned, simply look at the hollistic approach which I used to make many predictions in April all but 2 came through:

http://eh2r.blogspot.ca/2016/04/2016-annual-spring-projection-made-by.html

The return of hurricane Hermine, after 10 years without some over North America is a very recent marking point, tornadoes less numerous, another. All this by studying the Arctic. We are linked together, from one end of Earth to the other. Despite distances . a dramatic change in sea ice has huge impacts all over the world.

Now the last 2 unfufilled predictions, less sea ice extent than 2012, and water at the North Pole, may appear illusive, or unpredictable. But consider the looks, of the vast fields of broken up sea ice with extent grids barely above 15% (counting as 100%) and how close there is wide open water at the North Pole. I have achieved what mega computers can't because I observe
and integrate these observations like no other computer can. Perhaps if we knew more about how they work, under which algorythm application, especially if we are informed in simple over all terms, we can make these super computers ultra precise, much better than a single person.

[Kevin O'Neill]

Wayne, a full description of the model can be found here (free access): Modeling Global Sea Ice with a Thickness and Enthalpy Distribution Model in Generalized Curvilinear Coordinates, Jinlun Zhang and D. A. Rothrock

"The model biases are within 8% in Arctic ice motion, within 9% in Arctic ice thickness, and within 14% in ice extent in both hemispheres. The model captures 56% of the variance of ice thickness along the 1993 submarine track in the Arctic. The simulated ridged ice has various thicknesses, up to 20 m in the Arctic and 16 m in the Southern Ocean.

[Jim Hunt]

Wayne - Only slightly tongue in cheek, I ponder the prospect of "wide open water at the North Pole" later this month:

"Could Northabout Sail to the North Pole?"

Here's the current state of play:

[wayne]

Hi Jim

yes, looks like water indeed, the answer would be yes, timing is everything though.

[wayne]

Hi Kevin,

Right, but I want to see how they work along with results. So many models but no idea about their main working points.

[wayne]

Kevin

http://neven1.typepad.com/.a/6a0133f03a1e37970b01b7c8897d5e970b-pi

How is TED doing for this year? Can't find its prediction.

[Rob Dekker]

Jim, ""Could Northabout Sail to the North Pole?""
In the forecasts :
https://sites.google.com/site/arcticseaicegraphs/forecasts
it looks like Northabout will have mostly tail wind (smooth sailing) across the the remainder of the NW passage once they have their repairs done in Tuk.

As for the theoretical trip to the NP, your AMSR2 image

suggest that beyond 86N there is a fairly low concentration ice path all the way to the NP, which is quite unusual by itself (so Serreze is right in my opinion).

But again looking at the forcasts link above, that strech of low concentration ice may become more "open" water in the week to come.

Temp forecast (for the NP area) is well above normal, which means there will not be much ocean water freezing yet. There won't be much melting either though, since the sun is quickly loosing strength now.
But the winds over that low concentration ice will be mostly towards Wrangel Island, which means (Coriolis force) that ice movement will be towards the Laptev, which will further lower the ice concentration in that band.

How much dispersion, and will it be enough for a clean open path to the NP ? That is hard to tell, but we should know in about a week and before the real refreezing starts in the NP area.