Support the Arctic Sea Ice Forum and Blog

Author Topic: The 2015 melting season  (Read 1875724 times)

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3800 on: August 30, 2015, 08:56:32 AM »


Dots put on that manually, to show the similarity with other recent years and what seems to be an emerging cyclic behaviour. It will be interesting to see if we have large losses next year in August. If there is a cycle, and not just random behaviour, this 'predicts' large August losses in 2016.

(Had to edit the graphic as I originally put 'four cycles'... too early in the morning)

jdallen

  • Young ice
  • Posts: 3262
    • View Profile
  • Liked: 536
  • Likes Given: 209
Re: The 2015 melting season
« Reply #3801 on: August 30, 2015, 09:53:32 AM »


Dots put on that manually, to show the similarity with other recent years and what seems to be an emerging cyclic behaviour. It will be interesting to see if we have large losses next year in August. If there is a cycle, and not just random behaviour, this 'predicts' large August losses in 2016.

(Had to edit the graphic as I originally put 'four cycles'... too early in the morning)
Interesting... your graph implies a 3+ million KM2 ice loss in August of 2016.
This space for Rent.

helorime

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 4
  • Likes Given: 17
Re: The 2015 melting season
« Reply #3802 on: August 30, 2015, 10:22:25 AM »
Ice thickness for the end of August comparison for the last 4 years.  All August 28 except for 2012 which was missing it.  I remember assuming that some old fast ice would always remain in the CAA.  This year is so different on that front. OTOH every year is unique.
Any sufficiently advanced technology is indistinguishable from magic.

seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3803 on: August 30, 2015, 10:47:38 AM »
Continuing drama from the Beaufort Sea storm as it continues to smash the sea ice in that region and now further inwards towards the North Pole.

Observe the new low concentration region between about 150E and 170E and at 7 to 9 degrees from the Pole! Green patches there correspond to below 50% concentration.

Worldview is only showing cloud cover there at the moment. What will we see when the cloud lifts?!

At least Climate Reanalyser is now forecasting (GFS code) the winds dying down now, both from the storm and around the Arctic Basin in general. Is this the consensus of the forecasts?

Click on gif to crossfade between yesterday's and today's U.Bremen ice concentration maps...

One would think clouds are fooling the measurement/algorithm, but, it shows some patchy structure with openings. Winds may have broken and dispersed the ice pack a bit. We'll see. Warm air will reach over there, so if broken, may talke time to heal.


Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3804 on: August 30, 2015, 11:12:31 AM »
O-Buoy 12 seems to have ceased to function during the recent cyclone. IMB buoy 2014G survived however, and set two consecutive daily distance records:

http://GreatWhiteCon.info/2015/08/barrow-battered-by-big-waves/#Aug30

Quote
There is now a large potential fetch across the East Siberian, Chukchi and Beaufort Seas should any further cyclones occur in the area over the next few weeks. Hence the conditions now exist for an even more damaging swell to be generated before the 2015 minimum Arctic sea ice extent is reached.
Reality is merely an illusion, albeit a very persistent one - Albert Einstein

Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3805 on: August 30, 2015, 12:16:53 PM »
The clouds are clearing after the recent cyclone:

http://GreatWhiteCon.info/resources/arctic-sea-ice-images/summer-2015-images/#CAB

This is the "Beaufort Bite" north of 80 degrees. Feel free to click through to Worldview, switch to Terra, and examine what's left below 80 degrees also.
Reality is merely an illusion, albeit a very persistent one - Albert Einstein

slow wing

  • Grease ice
  • Posts: 819
    • View Profile
  • Liked: 155
  • Likes Given: 513
Re: The 2015 melting season
« Reply #3806 on: August 30, 2015, 12:32:00 PM »
Jim, was just about to post a similar image of roughly the same region. (And jdallen has already posted one as well.) Strangely, the cloud cover is in different places, so will post it anyway.


Neven

  • Administrator
  • First-year ice
  • Posts: 7821
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1139
  • Likes Given: 546
Re: The 2015 melting season
« Reply #3807 on: August 30, 2015, 01:20:32 PM »
Wow, that looks like an explosion...
Il faut comparer, comparer, comparer, et cultiver notre jardin

Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3808 on: August 30, 2015, 01:29:24 PM »
Strangely, the cloud cover is in different places, so will post it anyway.

Try switching between Aqua and Terra, and back and forth one day. In that neck of the woods the images are recorded at significantly different times of day.
Reality is merely an illusion, albeit a very persistent one - Albert Einstein

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3809 on: August 30, 2015, 01:32:56 PM »
Interesting... your graph implies a 3+ million KM2 ice loss in August of 2016.

I wouldn't draw any such conclusion myself, it's just a three point series with the last point being strongly impacted by weather (2012).

But I do expect high losses next August.

greatdying2

  • Grease ice
  • Posts: 574
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3810 on: August 30, 2015, 03:28:29 PM »
Wow, that looks like an explosion...
It must reflect to some serious volume loss (in reality at least).
The Permian–Triassic extinction event, a.k.a. the Great Dying, occurred about 250 million years ago and is the most severe known extinction event. Up to 96% of all marine species and 70% of terrestrial vertebrate species became extinct; it is also the only known mass extinction of insects.

Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3811 on: August 30, 2015, 03:30:49 PM »
A post cyclone closeup of the JAXA AMSR2 RGB false colour and thickness/melt maps:

http://GreatWhiteCon.info/resources/arctic-sea-ice-graphs/#JAXA-AMSR2

« Last Edit: August 30, 2015, 03:36:09 PM by Jim Hunt »
Reality is merely an illusion, albeit a very persistent one - Albert Einstein

Tor Bejnar

  • Young ice
  • Posts: 3850
    • View Profile
  • Liked: 671
  • Likes Given: 505
Re: The 2015 melting season
« Reply #3812 on: August 30, 2015, 03:33:01 PM »
Before seeing Chris's graph, I've 'expected' large August sea ice losses every year.  The graph shows that I get what I expect about twice per 4 years (yellow and red dots are large, historically, if not when compared with 2012).  I do not believe this is a 4-year climatological cycle, however.  Per Chris's graph, I should expect a 50-50 chance of >1.75 August sea ice loss and a 1-in-4 chance of >2.0 August loss. This goes for 'any' August though, not just 2016's.
Arctic ice is healthy for children and other living things.

Tor Bejnar

  • Young ice
  • Posts: 3850
    • View Profile
  • Liked: 671
  • Likes Given: 505
Re: The 2015 melting season
« Reply #3813 on: August 30, 2015, 03:47:58 PM »
Wow, that looks like an explosion...
It must reflect to some serious volume loss (in reality at least).
I wonder if the scattered remaining floes (in the North-of-Beaufort Sea [Should I trademark this name? 8)]) are mostly thick 4-5-year ice whereas all the 'white scat' (and open water) is (was) the thin 1st year ice.  The consequence for volume could be less disastrous than what's apparent.  Hypothetically, the 95% (per area) of 'white scat' could go from 0.3m to 0.1m while the 5% of 6m ice goes down to 5.8m.  Visually it looks horrible, but not all that much volume was lost.
Arctic ice is healthy for children and other living things.

greatdying2

  • Grease ice
  • Posts: 574
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3814 on: August 30, 2015, 03:56:33 PM »
Wow, that looks like an explosion...
It must reflect to some serious volume loss (in reality at least).
I wonder if the scattered remaining floes (in the North-of-Beaufort Sea [Should I trademark this name? 8)]) are mostly thick 4-5-year ice whereas all the 'white scat' (and open water) is (was) the thin 1st year ice.  The consequence for volume could be less disastrous than what's apparent.  Hypothetically, the 95% (per area) of 'white scat' could go from 0.3m to 0.1m while the 5% of 6m ice goes down to 5.8m.  Visually it looks horrible, but not all that much volume was lost.
Yeah, quite possibly. However, does this not also suggest a potential continued increased rate of bottom melt over coming weeks? If so, those large floes might continue to suffer in volume if not visibly on the surface. Do I recall correctly that there a small cyclone that passed over the area 'inside' the arm earlier in the season, not long before it began to melt out?
The Permian–Triassic extinction event, a.k.a. the Great Dying, occurred about 250 million years ago and is the most severe known extinction event. Up to 96% of all marine species and 70% of terrestrial vertebrate species became extinct; it is also the only known mass extinction of insects.

seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3815 on: August 30, 2015, 04:36:27 PM »
Wow, that looks like an explosion...
It must reflect to some serious volume loss (in reality at least).
I wonder if the scattered remaining floes (in the North-of-Beaufort Sea [Should I trademark this name? 8)]) are mostly thick 4-5-year ice whereas all the 'white scat' (and open water) is (was) the thin 1st year ice.  The consequence for volume could be less disastrous than what's apparent.  Hypothetically, the 95% (per area) of 'white scat' could go from 0.3m to 0.1m while the 5% of 6m ice goes down to 5.8m.  Visually it looks horrible, but not all that much volume was lost.
Yeah, quite possibly. However, does this not also suggest a potential continued increased rate of bottom melt over coming weeks? If so, those large floes might continue to suffer in volume if not visibly on the surface. Do I recall correctly that there a small cyclone that passed over the area 'inside' the arm earlier in the season, not long before it began to melt out?

Yes it does. Bottom melt wont cease until October as other years, this one with more heat available than 2013 or 2014.  It will make volume stall or grow very slowly.


seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3816 on: August 30, 2015, 04:56:33 PM »
The clouds are clearing after the recent cyclone:

http://GreatWhiteCon.info/resources/arctic-sea-ice-images/summer-2015-images/#CAB

This is the "Beaufort Bite" north of 80 degrees. Feel free to click through to Worldview, switch to Terra, and examine what's left below 80 degrees also.

That is way North, the location of the yellow area in Bremen map, isn't it? We'll see this open space at least for one week more, I would say two, without signs of refreezing. Unless it gets compacted.
It is amazing how weak the ice is in this area behind its presumed solidity. All July month and much of August it underwent constant surface melting. And, the new buoy 2015F is toward ESS from this area, in a much more solid area, still shows some bottom melt.

Also, the openings are not a-la thick MYI (this would look more like the leads of the protruding part toward ESS), but rather as thin weak ice.

AbruptSLR

  • Multi-year ice
  • Posts: 19225
    • View Profile
  • Liked: 2134
  • Likes Given: 264
Re: The 2015 melting season
« Reply #3817 on: August 30, 2015, 05:04:15 PM »
Interesting... your graph implies a 3+ million KM2 ice loss in August of 2016.

I wouldn't draw any such conclusion myself, it's just a three point series with the last point being strongly impacted by weather (2012).

But I do expect high losses next August.

Historically strong El Nino years do not result in large Arctic Sea Ice losses; however, strong La Nina events normally follow immediately after a strong El Nino event; and historically strong La Nina events do induce conditions that support large Arctic Sea Ice losses.  Thus as our current strong El Nino event should likely dissipate by March 2016, expecting a high La Nina assisted Arctic Sea Ice loss by the end of August 2016 seems like a good bet.
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

solartim27

  • Grease ice
  • Posts: 569
    • View Profile
  • Liked: 31
  • Likes Given: 38
Re: The 2015 melting season
« Reply #3818 on: August 30, 2015, 05:06:06 PM »
This is the forecast from http://cci-reanalyzer.org/Forecasts/ for next Saturday.  I know the latest cyclone forecast fizzled out, but if this comes to pass it will be a mess.
FNORD

seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3819 on: August 30, 2015, 05:30:43 PM »
Interesting... your graph implies a 3+ million KM2 ice loss in August of 2016.

I wouldn't draw any such conclusion myself, it's just a three point series with the last point being strongly impacted by weather (2012).

But I do expect high losses next August.

I very was intrigued by this statistics, and thought it was maybe because of the way ice pack configuration evolved, maybe setting the stage in Fall 2003, 2007, or 2011, or some resonance in the FYI creation/MYI destruction process due to higher mobility of the ice pack, or something.

What I found is that July 2004 and 2008 were badly lagging in ice melting in Barents and Kara (and Baffin), which then melt in August on top of the melting in the Arctic ocean itself. For 2012 there was vigorous melting and also the GAC.

So I cannot see explanation to this apart from the weather, which is random too. By Occam's Razor I take the easy way, which is, the probability of next August melting be higher than the post-2007 average is slightly higher than 50% (50% plus the added chance due to global warming).

Gonzo

  • Frazil ice
  • Posts: 123
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3820 on: August 30, 2015, 09:11:17 PM »
I wonder if this on the Atlantic side will just break off, and half of it just head straight out the Fram?

Buddy

  • Young ice
  • Posts: 3379
  • Go DUCKS!!
    • View Profile
  • Liked: 49
  • Likes Given: 34
Re: The 2015 melting season
« Reply #3821 on: August 30, 2015, 09:47:03 PM »
You know....Gonzo brings up something in my mind about the next 10 or so years.  We will be seeing things that are so "off the charts"....and haven't been seen before.

1) The Arctic melting out....earlier and earlier.
2)  Large "chunks" of Antarctica and Greenland heading out to see
3)  Large "precipitation events" occurring...

Humans aren't very well adapted to adjusting to things they haven't seen before.  We're much better at things "happening again" that have happened before in HUMAN HISTORY.

A couple more weeks of this melting season....and on to a likely MORE exciting season in 2016.

FOX (RT) News....."The Trump Channel.....where truth and journalism are dead."

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3822 on: August 30, 2015, 10:04:47 PM »
Before seeing Chris's graph, I've 'expected' large August sea ice losses every year.  The graph shows that I get what I expect about twice per 4 years (yellow and red dots are large, historically, if not when compared with 2012).  I do not believe this is a 4-year climatological cycle, however.  Per Chris's graph, I should expect a 50-50 chance of >1.75 August sea ice loss and a 1-in-4 chance of >2.0 August loss. This goes for 'any' August though, not just 2016's.

& SeaIceSailor,

I don't think it's weather related (climatological) I suspect that the cycling time for muti-year ice in the Beaufort flywheel is being revealed as the ice has thinned (winter peak thicknes) and open water increased in late summer. The increased in losses seems to be due to the Chukchi and ESS regions as shown below. MYI export into those regions seems to suppress August melt, lack of MYI enhances it.



The current blog post on my blog has a link to a previous study. If 2016's August losses bear out a third cycle I may revisit the matter.

Nick_Naylor

  • Frazil ice
  • Posts: 291
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3823 on: August 31, 2015, 12:58:14 AM »
The shattering on the Atlantic caught me by surprise. I really hadn't thought it was solid enough to break up like this.

Lou

  • New ice
  • Posts: 13
    • View Profile
  • Liked: 12
  • Likes Given: 54
Re: The 2015 melting season
« Reply #3824 on: August 31, 2015, 05:30:00 AM »
You know....Gonzo brings up something in my mind about the next 10 or so years.  We will be seeing things that are so "off the charts"....and haven't been seen before.

1) The Arctic melting out....earlier and earlier.
2)  Large "chunks" of Antarctica and Greenland heading out to see
3)  Large "precipitation events" occurring...

Humans aren't very well adapted to adjusting to things they haven't seen before.  We're much better at things "happening again" that have happened before in HUMAN HISTORY.

A couple more weeks of this melting season....and on to a likely MORE exciting season in 2016.

This is the premise of Bill McKibben's book "Eaarth", namely that via our greenhouse gas emissions we've transformed the only planet and climate our civilization has known into something else, something no human being has ever seen.

This is why I so seldom say anything on this site.  While I learn a great deal from the contributions of others here, and I find it extremely fascinating -- for which I thank you all -- I think it's clear that a lot of the details of daily, weekly, monthly, and even yearly happenings in the Arctic will be washed away once we start having Blue Arctic Events (nearly) every year.  And barring a massive and immediate geoengineering push, we can't escape taking that major step even further from our original Earth.

slow wing

  • Grease ice
  • Posts: 819
    • View Profile
  • Liked: 155
  • Likes Given: 513
Re: The 2015 melting season
« Reply #3825 on: August 31, 2015, 05:30:31 AM »
Strangely, the cloud cover is in different places, so will post it anyway.

Try switching between Aqua and Terra, and back and forth one day. In that neck of the woods the images are recorded at significantly different times of day.
Thanks, great tip!


In the end, the day-to-day change in the U. Bremen map wasn't quite as dramatic as I expected. The concentrations are still relatively high in parts of the Beaufort Sea (and North of it - how far North does it go?) where the ice pack has been reduced to rubble. And some of the peek-a-boo ice even returned, close to the CAA.


Click on gif for the comparison with yesterday...


AySz88

  • New ice
  • Posts: 25
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3826 on: August 31, 2015, 08:16:20 AM »


Dots put on that manually, to show the similarity with other recent years and what seems to be an emerging cyclic behaviour. It will be interesting to see if we have large losses next year in August. If there is a cycle, and not just random behaviour, this 'predicts' large August losses in 2016.

(Had to edit the graphic as I originally put 'four cycles'... too early in the morning)

Some math, just to get a concrete handle on how close the "cycle" is to statistical significance (or whether it's close at all).

p-value = Chance of a random ordered set of 12 years, where the top 3 years are denoted set A, next 3 set B, bottom 6 set C, will seem to show any 4-year cycle (any specific ordering of one A one B two Cs, repeated 3 times):

Code: [Select]
12 possible orderings in the 4-year cycle (A B C C, A C B C, etc.)
* 6 orderings within set A
* 6 orderings within set B
* 6! orderings within set C
/ 12!

= 1/1540 ... but then since I don't know how the hypothesis was generated I need to correct for the potential for multiple hypotheses:

Code: [Select]
1/1540
* 6 for the choice of month (conservatively high estimate)
* 4 for the choice of cycle length and number of years to use (estimate)
* 2 for choice of extent vs. area
* 2 for maybe choice of data source?

= [b]24/385 (0.062)[/b]

So the p-value is roughly on the order of 0.062 ...arguably pretty close.

(Apologies in advance if there's an error - this is rather late-night math....)


(edit:formatting)

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3827 on: August 31, 2015, 08:46:08 AM »
Aysz88,

Thanks for that. Isuspect that the likelihood that it's random might be higher because there are various oscillation processes within the Arctic, some of which may possibly have similar behaviour in the frequency domain.

Killian

  • Frazil ice
  • Posts: 299
    • View Profile
  • Liked: 62
  • Likes Given: 8
Re: The 2015 melting season
« Reply #3828 on: August 31, 2015, 09:06:11 AM »
Quote
Historically strong El Nino years do not result in large Arctic Sea Ice losses; however, strong La Nina events normally follow immediately after a strong El Nino event; and historically strong La Nina events do induce conditions that support large Arctic Sea Ice losses.  Thus as our current strong El Nino event should likely dissipate by March 2016, expecting a high La Nina assisted Arctic Sea Ice loss by the end of August 2016 seems like a good bet.

Beg to differ, somewhat. I recently posted this at RealClimate:

Further to the discussion on EL/LN (ENSO) and correlation with ASI Extent lows, my original observation noted rough correlations between EN’s followed 1 or 2 years later by new lows. This observation was based on eyeballing graphs and included no other conditions affecting the ice. However, it is clear the ’07 low was strongly affected by winds and weather generally. Likewise, the lack of winds and relative cold correlated well with the last two years of rise in ASIE and this year almost certainly not hitting new lows. In fact, I had not called new lows, or even 2nd or 3rd new records in extent because I think all the open water makes high variation in extent *less* likely. The weather has cooperated to make that still true.

I think, in fact, it makes little sense to look for new lows in extent as the primary metric when it is so easy for ice to spread in open seas. Area and volume both tell us more about the *amount* of ice, anyways, which I have focused on since ’10. Unfortunately, we don’t have good records on those so rely on the continuity of the extent record to check for this correlation.

I went further back, eyeballing from an extent graph through 2010 or so with poor detail (what I could find) and a list of ENSO years and intensities. This is the rough. If anyone has more detailed resources and can nail this down better, please do.

Here is what I found going all the way back to the beginning of ASIE decline @ 1953-ish.

EN ’51 – ’54 = inception of ASI Extent decline.
EN ’57 – ’59 = Near New Low/New Low
EN ’65 – ’66 = Near New Low/New Low
EN ’68 – ’70 = New Low
EN ’72 – ’73 = possible correlation, some delay
EN ’76 – ’78 = New Low
EN ’79 – ’80 = New Low
EN ’82 – ’83 = New Low
EN ’86 – ’88 = New Low (’89,’90)
EN ’94 – ’95 = New Low
EN ’97 – ’98 = Drop from Previous (?)
EN ’04 – ’05 = Near New Low/New Low
EN ’04 – ’05/’06 – ’07 = New Low
EN ’09 – ’10 = New Low (’10, ’12)
EN ’15 – ’16 = New Low ’16,’17?

What about La Nina? First, I don’t care if the actual cause is LN or EN, the correlation with ENSO is the key. However, we have a problem, Houston. I didn’t check the full record of LN because I found this from recent years:

La Nina
’07-’08 After
’10-’11 After
’10-’11, ’11-’12 = New Low

If LN is following big melts, it can’t be contributing to them, so the correlation is already weaker than with EN. Also, the ’10-’11 EN could be said to be correlated with the ’12 low given the hypothesis of near new lows/new lows following EN’s by 1 or 2 years. Feel free to check the LN record, tho.

What I think is going on is the EN’s put a lot of heat into the ocean surface *and* the air, yes? I propose these waters and air temps propagate via multiple routes (directly, storm tracks, etc.) into the Arctic enhancing melt. We know melt from the bottom due to water temps is responsible for up to 2/3 of the ice melt, and also that it takes much longer for energy to propagate through the oceans than the atmosphere. It makes sense an EN correlation might be delayed.

It would be great to correlate wind and storm patterns and temps with all this. Let’s be clear: I am not saying ENSO *is the primary cause of melt*, only that there is what appears to be a strong correlation.

http://www.realclimate.org/?comments_popup=18643#comment-635199


Pi26

  • New ice
  • Posts: 41
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3829 on: August 31, 2015, 09:06:32 AM »
Code: [Select]
[quote author=AySz88 link=topic=1149.msg62366#msg62366 date=1441001780]
[quote author=ChrisReynolds link=topic=1149.msg62301#msg62301 date=1440917792]
[img width=483 height=409]https://farm1.staticflickr.com/774/20987222552_200984cb30_o.png[/img]


p-value = Chance of a random ordered set of 12 years...

Clearly when ice melts there is fast cooling as it consumes warmth immediatly, but also slow cooling until the ice water heats up to world wide average temperature.

Killian

  • Frazil ice
  • Posts: 299
    • View Profile
  • Liked: 62
  • Likes Given: 8
Re: The 2015 melting season
« Reply #3830 on: August 31, 2015, 09:15:39 AM »


Dots put on that manually, to show the similarity with other recent years and what seems to be an emerging cyclic behaviour. It will be interesting to see if we have large losses next year in August. If there is a cycle, and not just random behaviour, this 'predicts' large August losses in 2016.

(Had to edit the graphic as I originally put 'four cycles'... too early in the morning)

Some math, just to get a concrete handle on how close the "cycle" is to statistical significance (or whether it's close at all).

p-value = Chance of a random ordered set of 12 years, where the top 3 years are denoted set A, next 3 set B, bottom 6 set C, will seem to show any 4-year cycle (any specific ordering of one A one B two Cs, repeated 3 times):

Code: [Select]
12 possible orderings in the 4-year cycle (A B C C, A C B C, etc.)
* 6 orderings within set A
* 6 orderings within set B
* 6! orderings within set C
/ 12!

= 1/1540 ... but then since I don't know how the hypothesis was generated I need to correct for the potential for multiple hypotheses:

Code: [Select]
1/1540
* 6 for the choice of month (conservatively high estimate)
* 4 for the choice of cycle length and number of years to use (estimate)
* 2 for choice of extent vs. area
* 2 for maybe choice of data source?

= [b]24/385 (0.062)[/b]

So the p-value is roughly on the order of 0.062 ...arguably pretty close.

(Apologies in advance if there's an error - this is rather late-night math....)


(edit:formatting)

That looks an awful lot like the wobbles approaching a phase change in a non-linear and/or chaotic system.

6roucho

  • Frazil ice
  • Posts: 296
  • Finance geek
    • View Profile
  • Liked: 0
  • Likes Given: 1
Re: The 2015 melting season
« Reply #3831 on: August 31, 2015, 11:09:57 AM »
That looks an awful lot like the wobbles approaching a phase change in a non-linear and/or chaotic system.
It does. The classic three-spikes-and-you're-out rubric of statistical analysis.

seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3832 on: August 31, 2015, 11:36:14 AM »
Both GFS and ECWMF predict warm conditions over the Pacific side for the next five days, after the amazing pull of warm air from Canada.. So flow of warm air from America and Pacific, with compacting winds.

After this range, the GFS predicts a nice central cyclone shutting everything down, while the Euro keeps the lights on in the Pacific side for longer.

While unreliable, I attach the nice GFS forecast :--)

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3833 on: August 31, 2015, 11:46:48 AM »
Killian, 6roucho,

Critical slowing is a decreasing recovery from perturbations and it is considered to be a useful indicator of approaching a 'phase transition' or 'bifurcation'. The problem here is that the idea of tipping points has become a common meme which is cast about with little understanding and little regards for physical reality, thus the casting around of the loose idea is likely to lead to incorrect expectation.

Could this oscillation be conceivably interpreted as critical slowing? Yes it could, it could be viewed as a reduction of ability to recover from perturbations (interannual variation of extent or thickness) leading to an oscillation in August extent loss. However would it be expected to have a four year period? I suspect not, I would expect a random periodicity, which with a sufficently large dataset (e.g. parallel runs in a model), might have similar frequency domain charateristics but not a regular cycle (e.g. 4 years in every run of the model).

To invoke the idea of critical slowing and to present this as an early warning of an abrupt transition one would need to establish the causal mechanism. I have already suggested a mechanism, one that does not lead to the conclusion of an imminent crash. If you view this as critical slowing in prelude to a rapid transition, what is the mechanism that you propose?

Pi26

  • New ice
  • Posts: 41
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3834 on: August 31, 2015, 12:43:11 PM »
Killian, 6roucho,

....However would it be expected to have a four year period? I suspect not, I would expect a random periodicity, which with a sufficently large dataset (e.g. parallel runs in a model), might have similar frequency domain charateristics but not a regular cycle (e.g. 4 years in every run of the model).


I think melting water from sea ice puts in its slow moving cooling capacity.

1. A lot of ice melts and its melting water cools the local water - helps to (nearby) refreeze
2. All melts again but expands wider and refreezes
3. All melts again but expands too wide
4. It has expanded far too wide

The duration may vary with the melting volume
« Last Edit: August 31, 2015, 05:23:06 PM by Pi26 »

iceman

  • Frazil ice
  • Posts: 285
    • View Profile
  • Liked: 7
  • Likes Given: 19
Re: The 2015 melting season
« Reply #3835 on: August 31, 2015, 12:44:25 PM »
Before seeing Chris's graph, I've 'expected' large August sea ice losses every year.  The graph shows that I get what I expect about twice per 4 years (yellow and red dots are large, historically, if not when compared with 2012).  I do not believe this is a 4-year climatological cycle, however.  Per Chris's graph, I should expect a 50-50 chance of >1.75 August sea ice loss and a 1-in-4 chance of >2.0 August loss. This goes for 'any' August though, not just 2016's.

& SeaIceSailor,

I don't think it's weather related (climatological) I suspect that the cycling time for muti-year ice in the Beaufort flywheel is being revealed as the ice has thinned (winter peak thickness) and open water increased in late summer. The increased in losses seems to be due to the Chukchi and ESS regions as shown below. MYI export into those regions seems to suppress August melt, lack of MYI enhances it.
  ....

That's the first hypothesis I've heard for an intermediate-period cycle (slower than 1- or 2-year autocorrelation, faster than ENSO-driven - which is not really periodic anyway).  Concur that other Arctic ice influences could produce a similar pattern, so we'll have to wait and see.  It does appear, though, that MYI in Beaufort will be unusually low during the 2016 melt season.

ktonine

  • Frazil ice
  • Posts: 363
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3836 on: August 31, 2015, 01:34:38 PM »
I don't think it's weather related (climatological) I suspect that the cycling time for muti-year ice in the Beaufort flywheel is being revealed as the ice has thinned (winter peak thicknes) and open water increased in late summer. The increased in losses seems to be due to the Chukchi and ESS regions as shown below. MYI export into those regions seems to suppress August melt, lack of MYI enhances it.


Chris, if we assume this explanation is correct, then it just pushes the question down one level; what causes a cyclic variance in the amount of ice being developed by the Beaufort Gyre and deposited into the ESS/Chukchi?




seaicesailor

  • Guest
Re: The 2015 melting season
« Reply #3837 on: August 31, 2015, 01:48:15 PM »
Before seeing Chris's graph, I've 'expected' large August sea ice losses every year.  The graph shows that I get what I expect about twice per 4 years (yellow and red dots are large, historically, if not when compared with 2012).  I do not believe this is a 4-year climatological cycle, however.  Per Chris's graph, I should expect a 50-50 chance of >1.75 August sea ice loss and a 1-in-4 chance of >2.0 August loss. This goes for 'any' August though, not just 2016's.

& SeaIceSailor,

I don't think it's weather related (climatological) I suspect that the cycling time for muti-year ice in the Beaufort flywheel is being revealed as the ice has thinned (winter peak thickness) and open water increased in late summer. The increased in losses seems to be due to the Chukchi and ESS regions as shown below. MYI export into those regions seems to suppress August melt, lack of MYI enhances it.
  ....

That's the first hypothesis I've heard for an intermediate-period cycle (slower than 1- or 2-year autocorrelation, faster than ENSO-driven - which is not really periodic anyway).  Concur that other Arctic ice influences could produce a similar pattern, so we'll have to wait and see.  It does appear, though, that MYI in Beaufort will be unusually low during the 2016 melt season.

I remember this one too. But I don't remember the author giving or searching for an explanation of it, apart from the obvious 2007-2012 time span of five years.

https://forum.arctic-sea-ice.net/index.php/topic,994.msg37259.html#msg37259

Before seeing Chris's graph, I've 'expected' large August sea ice losses every year.  The graph shows that I get what I expect about twice per 4 years (yellow and red dots are large, historically, if not when compared with 2012).  I do not believe this is a 4-year climatological cycle, however.  Per Chris's graph, I should expect a 50-50 chance of >1.75 August sea ice loss and a 1-in-4 chance of >2.0 August loss. This goes for 'any' August though, not just 2016's.

& SeaIceSailor,

I don't think it's weather related (climatological) I suspect that the cycling time for muti-year ice in the Beaufort flywheel is being revealed as the ice has thinned (winter peak thicknes) and open water increased in late summer. The increased in losses seems to be due to the Chukchi and ESS regions as shown below. MYI export into those regions seems to suppress August melt, lack of MYI enhances it.

The current blog post on my blog has a link to a previous study. If 2016's August losses bear out a third cycle I may revisit the matter.

Chris,
Do you probably expect a faster melt in Chukchi and ESS next year because this year's MYI "shipment" will never make it there? That makes a lot of sense to me.
Perhaps a clearer signal is emerging, the problem I had by looking at the ice-ageing maps is that I couldn't see a clear periodicity, but as you are explaining, there may not be.

For instance: for August 2016 fast melt in Chukchi and ESS, an additional condition apart from lacking MYI is that a generous amount of heat should be transferred from the Pacific side as happened this year, but by Summer 2016 la Niña will probably have already come...

PS. I remaked that sentence to make the post on-topic

BornFromTheVoid

  • Nilas ice
  • Posts: 1302
    • View Profile
  • Liked: 624
  • Likes Given: 235
Re: The 2015 melting season
« Reply #3838 on: August 31, 2015, 02:30:32 PM »
Here's the composite SLP anomaly for the Septembers that lost of most extent between now and their minimum.



The composite anomaly for those that lost the least.




Here's the predicted SLP anomaly for the next week.



I think the main thing we're looking for now is compression of the pack to enhance extent and area losses. A dipole would work well, but really anything that will push ice from the Pacific side of the Atlantic side would do
I recently joined the twitter thing, where I post more analysis, pics and animations: @Icy_Samuel

6roucho

  • Frazil ice
  • Posts: 296
  • Finance geek
    • View Profile
  • Liked: 0
  • Likes Given: 1
Re: The 2015 melting season
« Reply #3839 on: August 31, 2015, 02:34:13 PM »
Killian, 6roucho,

Critical slowing is a decreasing recovery from perturbations and it is considered to be a useful indicator of approaching a 'phase transition' or 'bifurcation'. The problem here is that the idea of tipping points has become a common meme which is cast about with little understanding and little regards for physical reality, thus the casting around of the loose idea is likely to lead to incorrect expectation.

Could this oscillation be conceivably interpreted as critical slowing? Yes it could, it could be viewed as a reduction of ability to recover from perturbations (interannual variation of extent or thickness) leading to an oscillation in August extent loss. However would it be expected to have a four year period? I suspect not, I would expect a random periodicity, which with a sufficently large dataset (e.g. parallel runs in a model), might have similar frequency domain charateristics but not a regular cycle (e.g. 4 years in every run of the model).

To invoke the idea of critical slowing and to present this as an early warning of an abrupt transition one would need to establish the causal mechanism. I have already suggested a mechanism, one that does not lead to the conclusion of an imminent crash. If you view this as critical slowing in prelude to a rapid transition, what is the mechanism that you propose?
ChrisReynolds, I happily admit I know zilch about the physics of sea ice. My comment was merely an aside. I like to read what the experts here write about the Arctic.

Although it has to said that people in my industry make large bets with as little knowledge of the fundamentals of markets. Unfortunately.

Gonzo

  • Frazil ice
  • Posts: 123
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3840 on: August 31, 2015, 03:29:30 PM »
This is just rough, but looking at the last few days satellite images, everything outside this red line is just mush or very vulnerable in my estimation.

Gonzo

  • Frazil ice
  • Posts: 123
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3841 on: August 31, 2015, 03:39:03 PM »

Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3842 on: August 31, 2015, 04:22:31 PM »
I figure this will be of general interest? A new paper by Pam Posey, Alan Wallcraft, Walt Meier et al. has been published in The Cryosphere today:

Improving Arctic sea ice edge forecasts by assimilating high horizontal resolution sea ice concentration data into the US Navy’s ice forecast systems

Quote
A method of blending ice concentration observations from the Advanced Microwave Scanning Radiometer (AMSR2) along with a sea ice mask produced by the National Ice Center (NIC) has been developed, resulting in an ice concentration product with very high spatial resolution. In this study, ACNFS was initialized with this newly developed high resolution blended ice concentration product. The daily ice edge locations from model hindcast simulations were compared against independent observed ice edge locations. ACNFS initialized using the high resolution blended ice concentration data product decreased predicted ice edge location error compared to the operational system that only assimilated SSMIS data.

A second evaluation assimilating the new blended sea ice concentration product into the pre-operational Navy Global Ocean Forecast System 3.1 also showed a substantial improvement in ice edge location over a system using the SSMIS sea ice concentration product alone.

On 2 February 2015, these two new data sources (AMSR2 and IMS) were added to the operational ACNFS and the pre-operational GOFS 3.1 jobstreams.
« Last Edit: August 31, 2015, 04:47:19 PM by Jim Hunt »
Reality is merely an illusion, albeit a very persistent one - Albert Einstein

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3843 on: August 31, 2015, 06:58:08 PM »
Iceman, Ktonine, Sea Ice Sailor,

It is, I concede rather sketchy...

Kevin, if the cycle between entry into Beaufort and re-entry into Beaufort is something like 4 years it would be an oscillation driven by a physical feedback of ice mass/state (i.e. FYI/MYI). Consider the first graphic here:
https://nsidc.org/cryosphere/seaice/processes/circulation.html
Say something like, into Beaufort and to Chukchi = 1 year. Assume the same rate, that's about 1/4 of the circle, which sugests something like four years (I can't find the reference I read the ~4 years from originally). This feedback need not necessarily be actual packet of ice doing the circuit, more like the presence of ice at the end of the season affecting the ice state of the next leg of the trip.

It is worth noting that when looking at data to 2006 Zhang finds a shift to Arctic Dipole dominant weather after about 2003, might this explain the emergence of the supposed cycles after 2005?

It might be worth a seperate thread here, but not now as I am busy with stuff unrelated to sea ice.

plinius

  • Frazil ice
  • Posts: 403
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3844 on: August 31, 2015, 07:37:21 PM »
Chris, I think it is by no means significant. If you have two or three extraordinary events in a dataset, they are very line up to simulate quasi-periods.

I think this here might also serve as a warning example:


And as long as we have not even a sketchy physical explanation why there should be a four-year cycle, I would rather not waste thought on this.

ktonine

  • Frazil ice
  • Posts: 363
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3845 on: August 31, 2015, 07:59:17 PM »
From the Beaufort Gyre project:
These studies indicate that the Arctic Ocean surface layer currents are consistent with the Arctic atmosphere surface layer motion, alternating between cyclonic and anticyclonic circulation regimes. Each regime persists from 4 to 8 years, resulting in a period of 8-–16 years. The cyclonic pattern dominated during 1989-1996. Since 1997 the dominant regime has fluctuated, with an anticyclonic pattern being more prevalent (Fig. 1). Figures 1-9 show annual simulated wind-driven surface ice and ocean motion for 2000 through 2008. The Arctic Ocean Oscillation index (bottom panel in Figure 1) illustrates alternation of circulation regimes at a period of 8-16 years. During anticyclonic circulation regimes the BG region accumulates fresh water and during cyclonic regimes the BG region releases fresh water and this water could be available for transportation to the North Atlantic via Straits of Canadian Archipelago and Fram Strait.

 It's possible we've seen a reduction in the circulation time post 2007.

plinius

  • Frazil ice
  • Posts: 403
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3846 on: August 31, 2015, 08:05:27 PM »
nice find, but they neither link any reasonable figures or data, nor is there anything else substantiating their claims. I do a bit hard to see anything more than random (and non-Gaussian) fluctuations, possibly with some memory terms.

greatdying2

  • Grease ice
  • Posts: 574
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The 2015 melting season
« Reply #3847 on: August 31, 2015, 09:01:06 PM »
2015 is getting very close to both 2011 and 2007 on this NSIDC chart.
The Permian–Triassic extinction event, a.k.a. the Great Dying, occurred about 250 million years ago and is the most severe known extinction event. Up to 96% of all marine species and 70% of terrestrial vertebrate species became extinct; it is also the only known mass extinction of insects.

Jim Hunt

  • Young ice
  • Posts: 4865
  • Stay Home, Save Lives
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 530
  • Likes Given: 50
Re: The 2015 melting season
« Reply #3848 on: August 31, 2015, 10:07:17 PM »
Every run of GFS looks very different so far out, and hence I have very little confidence in this 12:00 version, but....

Reality is merely an illusion, albeit a very persistent one - Albert Einstein

Killian

  • Frazil ice
  • Posts: 299
    • View Profile
  • Liked: 62
  • Likes Given: 8
Re: The 2015 melting season
« Reply #3849 on: August 31, 2015, 10:29:47 PM »
Killian, 6roucho,

Critical slowing is a decreasing recovery from perturbations and it is considered to be a useful indicator of approaching a 'phase transition' or 'bifurcation'...

Could this oscillation be conceivably interpreted as critical slowing? Yes it could

And was. Nothing more.

Quote
...However would it be expected to have a four year period? I suspect not, I would expect a random periodicity...

Well, yeah. I think you are making overmuch out of little much. Not sure how one instance got us to periodicity... At least, I never said it.

Quote
To invoke the idea of critical slowing and to present this as an early warning of an abrupt transition one would need to establish the causal mechanism.

False. You are talking about proving, or at least getting to viable theory. We are talking on a blog. Nobody said it *is* a coming bifurcation, only that it looks like one.

Quote
I have already suggested a mechanism, one that does not lead to the conclusion of an imminent crash. If you view this as critical slowing in prelude to a rapid transition, what is the mechanism that you propose?

The Arctic's response to AGW? Some of us eggheads spend too much time with our pocket protectors.

 ;) ;D

Really, it's a very long list. There is no one mechanism, there is just all kinds of bad stuff going on moving in and out of phase. Had we had 2007 weather this summer, new lows in area and volume, guaranteed. But we didn't. We had the third in a string of cool summers with relatively little transport via Fram - itself perhaps a developing hysteresis on melt rate.

The ice is preconditioned (from the bottom up, imo): It's popcorn throughout, with some exception in the upper center opposite Greenland/Canada. Thicker floes sitting in mush. I've talked about this and the bottom melt and thinning a lot, but not here. I think the volume models are way off.

Cheers