Support the Arctic Sea Ice Forum and Blog

Author Topic: The Slow Transition  (Read 233486 times)

jai mitchell

  • Nilas ice
  • Posts: 2357
    • View Profile
  • Liked: 207
  • Likes Given: 60
Re: The Slow Transition
« Reply #300 on: August 22, 2015, 10:54:06 PM »
Neven,

to be perfectly clear, I am not claiming geoengineering or even aerosol production caused the extremely unlikely cooling of the arctic (and California's drought!) during those years but the persistent high pressure ridge blocking pattern, that has since moved further north, has a causation.

I speculate a potential impact from intentional geoengineering or aerosol emissions but this is only speculation.  However, it is impossible to prove that geoengineering did NOT happen, and I certainly don't have proof that it did.  So it remains speculation (as with all claims that it did not happen)

Until one can clearly prove what the actual cause was, it is all speculation.
Haiku of Futures Passed
My "burning embers"
are not tri-color bar graphs
+3C today

Peter Ellis

  • Grease ice
  • Posts: 619
    • View Profile
  • Liked: 33
  • Likes Given: 14
Re: The Slow Transition
« Reply #301 on: August 23, 2015, 12:01:24 AM »
I can't prove it wasn't little green men from Pluto, so shall we speculate about that too?

anotheramethyst

  • Frazil ice
  • Posts: 141
    • View Profile
  • Liked: 0
  • Likes Given: 2
Re: The Slow Transition
« Reply #302 on: August 23, 2015, 05:00:12 AM »
in one of paul beckwith's more recent videos, he talks a little about the rrr, and from his analysis it really looks like the jet stream just shifted to a new stable point.

ktonine

  • Frazil ice
  • Posts: 363
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #303 on: August 23, 2015, 05:45:31 AM »
I can't prove it wasn't little green men from Pluto, so shall we speculate about that too?

Oh c'mon. Let's get real.  Martians are green. Plutonians are yellowish orange.

Michael Hauber

  • Nilas ice
  • Posts: 1114
    • View Profile
  • Liked: 168
  • Likes Given: 16
Re: The Slow Transition
« Reply #304 on: August 24, 2015, 12:16:22 AM »
I see no obvious theoretical reason to prefer an extrapolation of volume or extent to predict ice free state.  In theory they should predict the same date.  For instance an absurdly simple model would be a perfectly circular ice cap around the North pole, with thickness increasing linearly from the edge to a maximum at the north pole.  Then as the planet warms the radius of ice shrinks, and the maximum thickness at the pole also increases.  The volume of ice would be that of a cone, and proportional to radius cubed.  The extent would be a circle proportional to radius squared.  They'd both reach 0 when radius reaches 0.  The extent would require a second degree polynomial fit, and volume a third degree poly fit, and they'd both perfectly predict the same ice free date.

So why extent and volume extrapolate to different dates is a puzzle, and perhaps the solution would tell us a bit more about what is going on.  One possible answer is that the volume measures are not accurate enough and the extent measures give a more reasonable extrapolation. 
Climate change:  Prepare for the worst, hope for the best, expect the middle.

sofouuk

  • New ice
  • Posts: 82
    • View Profile
  • Liked: 0
  • Likes Given: 15
Re: The Slow Transition
« Reply #305 on: August 24, 2015, 09:42:00 AM »
I think this describes area and volume, not extent and volume, not that that likely makes any difference

sofouuk

  • New ice
  • Posts: 82
    • View Profile
  • Liked: 0
  • Likes Given: 15
Re: The Slow Transition
« Reply #306 on: August 24, 2015, 10:32:52 AM »
more seriously the ice isn't disappearing in the way described. if we assume (only a little more feasibly) that the ice cap is a semicircle with 2 m first year ice (in spring) along the curved edge and thicker multi year ice towards the centre of the straight edge (Canadian archipegalo), that can approximate the traditional ice pack in approximate equilibrium. then add extra heat, and the time series shows the semicircle getting smaller at summers end as years pass, but only slowly bcz the semicircle expands back to the maximum size every winter, and it's hard to melt all the way to the thicker ice in the short arctic summer. However the thicker ice disappears as the Beaufort gyre turns from ice maker to ice melter, meaning the volume graph appears to be crashing towards zero, but in fact it's easier to go from 4 m to 2 m than to go from 2 m to 0 m, bcz 2 m of ice is added wherever there is open water every winter. again, isn't that the whole point of this thread?
« Last Edit: August 26, 2015, 01:53:43 AM by sofouuk »

ktonine

  • Frazil ice
  • Posts: 363
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #307 on: August 26, 2015, 03:43:36 AM »
I see no obvious theoretical reason to prefer an extrapolation of volume or extent to predict ice free state.  In theory they should predict the same date. 

Michael, that would be true if the ice was uniformly thick.  It was not.  All the old, thick ice is gone.  So for an area of ice equal to X where uniformity would have you lose volume Y we instead lost volume Y+Z.  Now, given that the old ice is gone, we should start to see the two measures begin to converge. 

It's good to remember that Maslowski made his prediction of 'virtually ice-free' using *volume* not extent as his metric.  When he made his prediction he defined 'virtually ice-free' as 80% volume loss at minimum from the average 1990s volume at minimum.  2012 was just a few percent shy of reaching that mark.

Since that time 'virtually ice-free' has been adopted by others - but they almost always refer to extent and usually assign a < 1Mkm^2 as their meaning.

jai mitchell

  • Nilas ice
  • Posts: 2357
    • View Profile
  • Liked: 207
  • Likes Given: 60
Re: The Slow Transition
« Reply #308 on: August 26, 2015, 04:13:28 PM »
I can't prove it wasn't little green men from Pluto, so shall we speculate about that too?

A similar blocking ridge shows up under the GeoMIP scenarios.

http://climate.envsci.rutgers.edu/GeoMIP/
Haiku of Futures Passed
My "burning embers"
are not tri-color bar graphs
+3C today

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: The Slow Transition
« Reply #309 on: August 26, 2015, 07:58:10 PM »
Sofouuk,

Yes, that is basically the argument I was making.

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: The Slow Transition
« Reply #310 on: August 26, 2015, 09:27:41 PM »
Chris
Does the timing of the strengthening of both systems align?
It's a fascinating theory that could really aid forecasters if proven.
Terry

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: The Slow Transition
« Reply #311 on: August 30, 2015, 08:32:12 AM »
Sorry Terry,

Been busy, so I'm late in getting back. I don't know what systems you are referring to.

Michael Hauber

  • Nilas ice
  • Posts: 1114
    • View Profile
  • Liked: 168
  • Likes Given: 16
Re: The Slow Transition
« Reply #312 on: September 11, 2015, 07:57:37 AM »
Tamino does some Change point anlaysis on Arctic ice.  This analysis suggest that sea ice loss accelerated up to 2007, and has flattened since then:



However note that this includes data from all year round, and the summer loss has been somewhat faster than winter loss in recent years. 

Its also interesting to revisit his extrapolation of September sea ice extent in 2012.  This extrapolation predicted that the sea ice minimum in 2013 would beat the 2007 minimum, and by eyeball would be getting pretty close to the 2012 minimum if the curve was extrapolated to 2015.

Some have argued that the more aggressive extrapolation of the volume stat should be preferred over the more conservative extrapolation of the extent.  However even the extrapolation of the extent has proven to be overly aggressive for the last three years.
Climate change:  Prepare for the worst, hope for the best, expect the middle.

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: The Slow Transition
« Reply #313 on: September 11, 2015, 01:04:14 PM »
Tamino does some Change point anlaysis on Arctic ice. 

Off topic: Suspect Tamino has got the trump change point analysis wrong - wonder if my comment will get through.

Back on topic:
Nice result on the sea ice - statistically significant change in rates found for 2002 and 2006.

Peter Ellis wonders about different seasons. I wonder about whether different metrics like area and volume should also be considered and tested. Guess that if you get the same date this helps regarding significance but makes calculating significance tricky if not impossible.

Steven

  • Grease ice
  • Posts: 957
    • View Profile
    • Arctic sea ice data and graphs
  • Liked: 481
  • Likes Given: 19
Re: The Slow Transition
« Reply #314 on: September 18, 2015, 05:51:23 PM »
Tamino does some Change point anlaysis on Arctic ice.  This analysis suggest that sea ice loss accelerated up to 2007, and has flattened since then:
 


However note that this includes data from all year round, and the summer loss has been somewhat faster than winter loss in recent years. 

I played around a bit with sea ice extent data for separate seasons, using this R function for change-point analysis.

For the autumn data, the algorithm finds a change-point near the year 1992: see the red line in the graph below.  This change-point is statistically justified, if I'm reading the algorithm output correctly. 



Piecewise linear fits with 2 (or more) change-points can also be computed, but they are not statistically justified for these autumn data, if I'm reading the algorithm output correctly.

The situation is different for the other seasons.  E.g. for the spring data, the algorithm suggests that there are no statistically justified change-points.  So for the spring data there is no good reason to prefer a piecewise linear fit (with 1, 2 or more change-points) instead of a simple linear fit over 1979-2015.


jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: The Slow Transition
« Reply #315 on: September 18, 2015, 10:11:57 PM »
Tamino does some Change point anlaysis on Arctic ice.  This analysis suggest that sea ice loss accelerated up to 2007, and has flattened since then:



However note that this includes data from all year round, and the summer loss has been somewhat faster than winter loss in recent years. 

Its also interesting to revisit his extrapolation of September sea ice extent in 2012.  This extrapolation predicted that the sea ice minimum in 2013 would beat the 2007 minimum, and by eyeball would be getting pretty close to the 2012 minimum if the curve was extrapolated to 2015.

Some have argued that the more aggressive extrapolation of the volume stat should be preferred over the more conservative extrapolation of the extent.  However even the extrapolation of the extent has proven to be overly aggressive for the last three years.
Dang if I wasn't just looking at that myself, from the standpoint of "steps" of volume drop.

I see three volume "Plateaus" looking at the data from 1979 onwards.

The first  is 1979-1998,    varying around ~ 13.5K KM3 +/- ~2K KM3
The second is 1999-2006 "      "    "      "   ~  10.5K KM3 +/- ~1.5K KM3
The last (and current) since 2007.......... ~    5.25K KM3  +/- ~1.75K KM3

Numbers are very rough (hold up thumb eyeballing it), but I think present a starting point for discussion.

This space for Rent.

Archimid

  • Young ice
  • Posts: 3511
    • View Profile
  • Liked: 899
  • Likes Given: 206
Re: The Slow Transition
« Reply #316 on: April 23, 2016, 05:06:10 PM »
Greetings. I have read the first and last page of this thread. I will revive it because I believe that the most important point of the slow transition is this.

 There are some thermodynamic bounds for the refreeze season that "guarantees" a slow anual average growth rate.

This is a good picture of this years refreeze season.

http://cires1.colorado.edu/~aslater/ARCTIC_TAIR/index_80_t2m.html

To those that favors the "Slow Transition" hypothesis, do this year change your expectations for the thermodynamics bounds of the refreezing season?
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: The Slow Transition
« Reply #317 on: April 23, 2016, 05:56:57 PM »
Yes, much warmer than normal throughout this winter. Yes fully agree that this reduces the volume growth.

Looks like we are heading for FDD anomaly of -1000 to -1100 much larger than 2005/6 -800 FDDA

Yes it makes a difference but to keep it in perspective this means 4700 to 4800 FDD compared to 5000 to 5700 seen over last decade so about 10% less than typical for the last decade.

A 10% fall every year and we would soon have an ice free Arctic in summer. But this is likely just an exceptionally warm winter. The trend fall in FDD looks more like 450FDD over about 15 years.


I disagree.  If 0 sea ice volume is reached at any point,  it will only refreeze to a maximum volume of around 15k KM3.

I hope by now you have realised that winter volume growth is growing as the minimum volume declines. (Hence summer is declining faster than winter volume.) For example:
In 2012 with Sept volume ~3.8 K KM^3 the volume grew to 23.2 K Km^3 a growth of 19.4 K Km^3

Go back to Sept 2000 volume 11.1 grew to 27.6 the following April so only 16.5 K Km^3

Maximum volume of 15K Km^3 does not seem likely if you are aware of this trend.

Hopefully you have also followed the logic that the less ice there is, the less insulation there is,
so heat can be lost faster,
and heat lost faster means faster rate of ice growth.


Starting from 0 volume with current typical winter temperatures I would expect growth to be somewhere close to the 22.5K Km^3 winter minimum volume.

Now maybe you can put together a scenario where we start with low Sept Ice, followed by an exceptionally warm winters like this last one such that volume only gets up to 20K Km^3 and then we have really good melt weather such that we get down to really low ice volume. Such a scenario, while possible is far from what should be expected with typical years.

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #318 on: April 23, 2016, 06:23:29 PM »
Crandles.....I am absolutely in your camp.

When I first started visiting here and learning from you guys, the idea of a year round ice free Arctic Ocean was being discussed and I expressed that I could not imagine any near term future where the Arctic night and the still brutal cold (although less brutal) would not cause the Arctic to freeze over. I still feel this way and your exposition on the effects of the cold temps and ocean heat escaping to the atmosphere has me convinced. One of the current trends that supports this is the increasing spread between the annual minimum and the annual maximum, this despite the fact that warm anomalies, north of 80 degrees, have been strongest in the winter refreeze.

I wondered if tracking this spread between annual min and max might be useful and, when the tendency for this spread to increase transitioned into a shrinking spread, it would signal a tipping point.

I, half in jest, suggested that this spread measure should be called the Bifurcated Intra-Annual Chryosphere Oscillation Trend or BICOT for short. It would also be known as "Baby, it's cold out there." as research scientists huddled in their shelters, warming themselves in the long polar night.

Tor Bejnar

  • Young ice
  • Posts: 4606
    • View Profile
  • Liked: 879
  • Likes Given: 826
Re: The Slow Transition
« Reply #319 on: April 23, 2016, 06:53:57 PM »
I am strongly influenced by Chris's "Slow Transition" arguments.  I am also aware of "crocodiles in the Arctic" from the not-so-distant geologic past, and my recollection (has my memory gone bonkers?) is that CO2 concentrations then were not particularly different from what we have now or soon will have.  There were other differences (open 'Panama'?) that could influence the differences in climate, for sure, but some physics-oriented folks on these threads tell of the immense amount of heat stored in the Arctic Ocean that is currently isolated from the surface.  If that heat gets effectively mixed with the less dense surface waters, could this cause a 'quick' transition to functional ice-free Arctic?  I've gotten the impression that there is enough stored heat that it could radiate all winter keeping the temperatures high.  I do not know if "quickly" means 5 or 10 years or 500 years, though.
Arctic ice is healthy for children and other living things because "we cannot negotiate with the melting point of ice"

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: The Slow Transition
« Reply #320 on: April 23, 2016, 08:26:03 PM »
Tor
I think that as Arctic ocean and air temperatures rise we can assume increased cloud cover. The increasing clouds should act to trap summer heat more and more efficiently, until voila, very little ice cover during winter.
Terry


Archimid

  • Young ice
  • Posts: 3511
    • View Profile
  • Liked: 899
  • Likes Given: 206
Re: The Slow Transition
« Reply #321 on: April 24, 2016, 12:01:53 AM »
Yes, much warmer than normal throughout this winter. Yes fully agree that this reduces the volume growth.

Looks like we are heading for FDD anomaly of -1000 to -1100 much larger than 2005/6 -800 FDDA

Yes it makes a difference but to keep it in perspective this means 4700 to 4800 FDD compared to 5000 to 5700 seen over last decade so about 10% less than typical for the last decade.

The last ten years is not a good metric to compare it to. I would compare the FDD's to the average of the whole dataset and from there derive what is typical of a melting arctic. I'm afraid there is no dataset of what is typical of a growing arctic. Since the data is available it's been shrinking.

Quote


A 10% fall every year and we would soon have an ice free Arctic in summer. But this is likely just an exceptionally warm winter. The trend fall in FDD looks more like 450FDD over about 15 years.


I agree and hope. But this very winter could be sufficiently exceptional to fully melt arctic sea ice. On top of that since the world is warming, the probability of an event like this is higher every year.

Quote

I disagree.  If 0 sea ice volume is reached at any point,  it will only refreeze to a maximum volume of around 15k KM3.


I hope by now you have realised that winter volume growth is growing as the minimum volume declines. (Hence summer is declining faster than winter volume.) For example:
In 2012 with Sept volume ~3.8 K KM^3 the volume grew to 23.2 K Km^3 a growth of 19.4 K Km^3

Go back to Sept 2000 volume 11.1 grew to 27.6 the following April so only 16.5 K Km^3

Maximum volume of 15K Km^3 does not seem likely if you are aware of this trend.

Hopefully you have also followed the logic that the less ice there is, the less insulation there is,
so heat can be lost faster,
and heat lost faster means faster rate of ice growth.


Starting from 0 volume with current typical winter temperatures I would expect growth to be somewhere close to the 22.5K Km^3 winter minimum volume.

I came up with the number 15k  by looking at a sea ice volume chart (https://protonsforbreakfast.files.wordpress.com/2012/01/arctic-sea-ice-volume.jpeg) and averaging two or three growing seasons I picked at random. It seems like the numbers I picked were too low. Nevertheless, your numbers are aligned with my expectations.

I do realize that volume grows faster when there is less ice for the reason that that you have already pointed out. My point is that the faster volume rate only happens when there is sufficient sea ice already present. If there is close to 0 sea ice, the whole situation changes. 

To illustrate it I thought of this, ice begets ice. See this video for reference:



I looked at the growing season and observed that ice does not form in open ocean areas. Ice forms contiguos to already available sea ice or it forms from frozen and snow covered coasts. This is better seen by looking at the opened months of September and October of 2007 and 2012. The ocean has no ice on it until older ice grows and cover it.  So if the ice completely melts, there won't be any sea ice until at least October.

Best case scenario is that some month well after October, ice will start growing near the coasts of the arctic, possibly growing fast enough to completely cover the arctic ocean with a thin layer of ice. What month that happens depends mostly on the heat content of the ocean, waves and weather. But the ocean heat content will be much higher because there was such a lower albedo over the whole arctic for a long time, so I bet it wont be until December until the ice fully regrows and it will be thin ice.

Probably once it fully recovers it will grow fast, but it won't be enough to not completely melt again the following year.



Quote
Now maybe you can put together a scenario where we start with low Sept Ice, followed by an exceptionally warm winters like this last one such that volume only gets up to 20K Km^3 and then we have really good melt weather such that we get down to really low ice volume. Such a scenario, while possible is far from what should be expected with typical years.

Yes, but in a warming world, such scenario is more likely every year.



I am an energy reservoir seemingly intent on lowering entropy for self preservation.

marcel_g

  • Frazil ice
  • Posts: 127
    • View Profile
    • Art by Marcel Guldemond
  • Liked: 54
  • Likes Given: 425
Re: The Slow Transition
« Reply #322 on: April 25, 2016, 04:42:22 PM »
Archimid's argument here about ice needing an edge to start to grow onto makes some sense.

Am I right to guess that open water, even if it's giving off a lot of heat to the atmosphere in fall and winter, still has wave action that would draw up heat from warmer depths? So unless the open water doesn't have any waves or wind, it would be slower to freeze up? Or would this allow more heat to be released to the atmosphere at earlier stages in the fall because the water is continuously moving and bringing heat to the surface, setting up faster volume growth in the winter?

However, even if the water can release heat faster, wouldn't the increased greenhouse effect in the atmosphere then trap more of this heat, reducing the water to atmosphere difference?

And if most of the water is open in the fall, wouldn't there also be a lot more insulating clouds in the arctic than there would be in the fall? Wouldn't increased cloud cover also slow down re-freeze and volume growth?

I once knew someone who did some mining/prospecting work in the Arctic in the early 90s, and he corrected me when I thought there would be a ton of snow up there by October: he said they only got 6 inches or so early on, and after that not much, because all the open water froze up and the air was very dry for the rest of the winter.

Tor Bejnar

  • Young ice
  • Posts: 4606
    • View Profile
  • Liked: 879
  • Likes Given: 826
Re: The Slow Transition
« Reply #323 on: April 25, 2016, 05:27:13 PM »
Further to what marcel_g is asking, would the freezing of fresh (or brackish) water from rivers (and river ice being carried out to sea) provide the seeds for widespread ice growth?

Hudson Bay 'totally' melts every year and freezes up every year.  Where does its surface start freezing and how does it spread?
Arctic ice is healthy for children and other living things because "we cannot negotiate with the melting point of ice"

DoomInTheUK

  • Frazil ice
  • Posts: 221
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #324 on: April 25, 2016, 05:33:52 PM »
So many of these issues are inter-linked. The heat transfer from open water and cloud cover issue is described quite neatly (as an intro at least) in this discussion on equable climates.

http://www.seas.harvard.edu/climate/eli/research/equable/

The Hudson Bay freeze point has occurred to me before but I've never found an answer that I trust....although "from the shoreline outwards" is about as close as I ever got.

marcel_g

  • Frazil ice
  • Posts: 127
    • View Profile
    • Art by Marcel Guldemond
  • Liked: 54
  • Likes Given: 425
Re: The Slow Transition
« Reply #325 on: April 25, 2016, 05:35:43 PM »
Thanks for the link Doom, looks very interesting.

DoomInTheUK

  • Frazil ice
  • Posts: 221
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #326 on: April 25, 2016, 05:53:40 PM »

Hudson Bay 'totally' melts every year and freezes up every year.  Where does its surface start freezing and how does it spread?

After another quick search, the closest I can find is from the Canadian Coast Guard:

http://www.ccg-gcc.gc.ca/e0010736

The bays and harbours on the north shore of Lake Superior. So basically any water that is shallow and calm freezes over first and then expands out when it's cold and calm enough.

theoldinsane

  • New ice
  • Posts: 66
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #327 on: April 25, 2016, 06:49:10 PM »
About ice formation from

http://www.smhi.se/kunskapsbanken/oceanografi/is-till-havs-1.4105

(Swedish Meteorological and Hydrological Institute)

This is only in Swedish, but I did a quick translation with Google translate of some of it, hence the bad language. There is more in the link for anyone who want to translate.

In calm and cold weather settles the newly formed ice as a very thin film. This occurs primarily over the still waters of the archipelagos and the smaller bays. If the wind picks up, the film burst and underlying, warmer water coming up. If the weather however, remains favorable for icing, the film grows on one or a few days to a smooth and continuous ice.

The next stage of growth is the clear greyish solid ice. The ice need not be fixed, but rather is continuous and motionless. Solid ice consists of clear ice formed when freezing sea water on its underside, and partly of snowice formed when wet snow freezes on its upper side. On top of this is usually a more or less thick snow cover.

In the open sea the water is rarely so still that it can form a film of ice. The waves circulate the water in the surface layer so that it cools uniformly. In cold weather, the ice crystals formed at the surface, as result of the movement of water does not freeze together but being thrown around to several meters deep.

Gradually the ice crystals become increasingly more and more bunched together and form a liquid, swaying mourn, which gradually dampens the waves and thus accelerates the formation of ice.

The mourn are lumped together and finally the slush freezes together into round ice plates. These are tight and bobs and wear therefore against each other. Upstanding, snow-white edges formed and flak similar flat plates, hence the name. The aggregations may be between 30 cm and 3 m in circumference.

Such upstanding edges can also be formed in the drift ice that originally was a dense, continuous ice cover. If this breaks formed irregular polygons eventually becoming rounder. On the form, you can always see that there are plate-like ice and not "real" plate ice.

DoomInTheUK

  • Frazil ice
  • Posts: 221
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #328 on: April 26, 2016, 10:27:45 AM »
So once we get an ice free CAB, throw in a few autumn and winter storms to keep things churning, and I imagine the resulting ice will be a jumbled mess for the coming melt season.

Too many variables, not enough time. All I can do is sit back and watch in a fascinated horror.

epiphyte

  • Frazil ice
  • Posts: 387
    • View Profile
  • Liked: 22
  • Likes Given: 21
Re: The Slow Transition
« Reply #329 on: April 26, 2016, 05:21:55 PM »
Sometimes common sense gives a better feel for the way things are likely to work than any model operating outside it's comfort zone. We know that still water freezes on exposure to freezing wind, but the rolling ocean doesn't. A ship sailing in those conditions will eventually capsize from the weight of ice on its superstructure.

...And just as the weight of maritime experience tells us that that the bigger the waves, the less likely the surface is to freeze, it also reveals that the broader the reach of open ocean, the bigger the waves.

All of which to me is just one of the red flags flying in the face of the theory that in the long term, winter freezing is more stable and predictable than summer melting. That might be true starting from a placid or even frozen surface. But starting from a wave-tossed ocean?


A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: The Slow Transition
« Reply #330 on: April 26, 2016, 06:31:28 PM »
Quote
might be true starting from a placid or even frozen surface. But starting from a wave-tossed ocean?
Right. Not only that but the mean temperature of the top 100 m of the mixed Arctic Ocean, in view of the massive heat capacity and efficiency of heat transfer of water relative to air -- is way too high to support significant winter ice (which with snow on top, actually serves as a heat-conserving lid rather than a good radiator of ocean heat accrued during the summer.).

Jim Hunt has put together excellent coverage of waves and mixing impacts. Open water in summer, near-surface algal bloom, sunlight intercepted so solar heat is adsorbed higher in the water column? Slow transition doesn't add up for me,  whistling in the dark.

abbottisgone

  • Frazil ice
  • Posts: 297
  • "...I'm a rock'n'roll star,...... YES I ARE!!!!!!"
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #331 on: April 26, 2016, 09:32:06 PM »
Tamino does some Change point anlaysis on Arctic ice.  This analysis suggest that sea ice loss accelerated up to 2007, and has flattened since then:
 


However note that this includes data from all year round, and the summer loss has been somewhat faster than winter loss in recent years. 

I played around a bit with sea ice extent data for separate seasons, using this R function for change-point analysis.

For the autumn data, the algorithm finds a change-point near the year 1992: see the red line in the graph below.  This change-point is statistically justified, if I'm reading the algorithm output correctly. 



Piecewise linear fits with 2 (or more) change-points can also be computed, but they are not statistically justified for these autumn data, if I'm reading the algorithm output correctly.

The situation is different for the other seasons.  E.g. for the spring data, the algorithm suggests that there are no statistically justified change-points.  So for the spring data there is no good reason to prefer a piecewise linear fit (with 1, 2 or more change-points) instead of a simple linear fit over 1979-2015.


The last graph is visually interesting...
..
But I left school and grew my hair
They didn't understand
They wanted me to be respected as
A doctor or a lawyer man
But I had other plans..........

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #332 on: January 01, 2017, 10:32:38 PM »
I have been pulling some data and have started charting trends on Arctic ice to see if I can interpret what has been going on since 1979. I will be posting these as I complete them. None of these charts are complex and all or most will be looking at annual data points. Many will provide absolutely no insight at all while others may suggest and result in some rather speculative and spurious conclusions.

I've attached the first four charts which attempts to determine if melt and freeze season lengths correlates with the amount of growth and melt for both SIE and SIA. Melt and freeze season lengths are determined by when Minimums and Maximums occur.

One problematic issue with this is that the max or min for a specific year does not reflect when there is a lengthy dwell of this measure at either minimum or maximum. As I was collecting the data, I noticed quite a few years where the final minimum or maximum was nearly identical to one that occurred weeks before or after. This causes the length of Melt or Freeze season to take on a certain arbitrariness. I will think about how I might correct this, perhaps pick the midpoint date between two nearly identical measures.

Meantime, here are the 1st four charts looking at SIE and SIA. From my perspective, there seems to be no correlation. I am no smarter than before, merely cost myself several hours of my life.  >:(
« Last Edit: January 01, 2017, 10:44:12 PM by Shared Humanity »

SCYetti

  • New ice
  • Posts: 34
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: The Slow Transition
« Reply #333 on: January 01, 2017, 11:24:00 PM »
Shared Humanity
But you did learn something. There is no apparent correlation between length of freezing and melting seasons with the final totals of ice extent. I wonder what the correlation with freezing degree days and thawing degree days would be.

charles_oil

  • Frazil ice
  • Posts: 337
    • View Profile
  • Liked: 38
  • Likes Given: 74
Re: The Slow Transition
« Reply #334 on: January 01, 2017, 11:29:14 PM »
Thanks - looks interesting but as yet I cant see any trend either

It might be handy or show something if the latest few years were shown (different colour/s / size of dots)?

I guess a high/low chart indicator might allow for the range on the drawn out max / min times....  How about using a range like within +/- x% to define the max (or min)?

Also - per your calcs - how many days into the freeze are we in 2016/17 ?

Which data set are you using?

Regards

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #335 on: January 01, 2017, 11:31:01 PM »
Here is another attempt at evaluating the changing state of Arctic ice since 1979. This chart looks at dispersion at minimum, essentially comparing extent to area. The first trend line looks at the difference in terms of millions of square kilometers while the second trend line is tracking the percentage difference between extent and area.

This chart suggests there are two and possibly three distinct transition points in dispersion or compactness. The 1st seems to occur in the 1988 - 1989 seasons where there is a marked increase in dispersion and perhaps an increase in year over year variability. Another obvious transition point occurs in 2007 - 2008 where there is a marked increase in dispersion and perhaps a reduction in year over year variability. While less pronounced, there seems to be another transition point in 1998.

While not certain, this chart suggests that when a transition occurs it is not likely that we return to the previous state. It is as if some type of floor is set in place which locks in an increased level of dispersion.

(edit.....)

Since we all know what happened in 2007 with the collapse in SIE, SIA and volume, I decided to look at annual volume minimums to see if this might be that floor. I've attached the PIOMAS chart and, while I cannot be certain, it seems that volume might be contributing to both the increase in variability and stabilizing of dispersion at new, higher levels..
« Last Edit: January 02, 2017, 12:10:20 AM by Shared Humanity »

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #336 on: January 01, 2017, 11:35:29 PM »
SIE data taken from NSIDC

SIA data taken from Cryosphere Today.

DrTskoul

  • Nilas ice
  • Posts: 1455
    • View Profile
  • Liked: 210
  • Likes Given: 60
Re: The Slow Transition
« Reply #337 on: January 01, 2017, 11:38:47 PM »
Data are too noisy to say anything about change points.  See Tamino's work on temperature data series.

Cheers!
« Last Edit: January 02, 2017, 09:06:30 PM by DrTskoul »

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #338 on: January 01, 2017, 11:57:00 PM »
Data are too noisy to say anything about change points.  We Tamino's work on temperature data series.

Cheers!

Like I said...posting these charts clearly aware that many, if not most, will not reveal anything of import.

Certainly the year to year observations are, quite possibly, useless to draw conclusions but the overall, nearly 40 year trend is increased dispersion which speaks to the increasing mobility of a shattered ice pack.
« Last Edit: January 02, 2017, 12:07:33 AM by Shared Humanity »

oren

  • Moderator
  • First-year ice
  • Posts: 9805
    • View Profile
  • Liked: 3584
  • Likes Given: 3922
Re: The Slow Transition
« Reply #339 on: January 02, 2017, 12:10:44 AM »
I doubt that using some very general variables like total extent and melt season length would give you anything meaningful, as the arctic variability is growing, and noise around the minimum and maximum prevents good definitions.
Something that might have a better chance of working is when you have more localized phenomena and use definitions that are less sensitive to noise. For example some ideas, if you look at specific peripheral seas that have become seasonal, and try to:
Correlate frozen season total length (let's say from 90% ice extent to 90% ice extent), or total FDDs over the freezing season, with melting period length (let's say from 90% ice extent to 10% ice extent).
Or total ice-free season length correlated with final freeze date.
One of the problems is that these local seas are affected by the seas around them.
Related directly to this thread, maybe correlate FDDs and PIOMAS volume?

DavidR

  • Grease ice
  • Posts: 740
    • View Profile
  • Liked: 36
  • Likes Given: 3
Re: The Slow Transition
« Reply #340 on: January 02, 2017, 01:16:54 AM »
One problematic issue with this is that the max or min for a specific year does not reflect when there is a lengthy dwell of this measure at either minimum or maximum. As I was collecting the data, I noticed quite a few years where the final minimum or maximum was nearly identical to one that occurred weeks before or after.
I  looked at this from a slightly  different  perspecctive to see how the length of the maximum plateau, affected the minimum.  I  defined the plateau  as the period when extent was within 200K sq km of the maximum, including  any dips that  later recovered.
http://forum.arctic-sea-ice.net/index.php/topic,1211.0.html

It  appeared there was some correlation between the length of the 'plateau' and the loss of ice extent  but  it  wasn't really  strong enough to  warrant continuing with.  However the concept  of a plateau may be useful to identify the length of  the melting season.
Toto, I've a feeling we're not in Kansas anymore

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #341 on: January 02, 2017, 04:59:04 PM »
I am generally at a loss with discussions all over the forum as the contributions here (many experts in areas that they study) far surpass my very limited understanding. I visit here daily and often do not comment. As this thread is looking at the slow transition, I thought one contribution I might make is to provide various trend charts of specific metrics. These trend charts are capturing this slow transition in some manner. Some of these charts I am creating although I sincerely doubt that any of my creations do not already exist. I just don't know where to find them. In many other cases, I will repost trend charts that I find interesting on other threads in this forum.

I originally thought I might be able to discern elements in these trends that highlight specific processes in the melt that could predict future events but as oren and DrTskoul point out here this is, quite likely, far too optimistic/naive. I still hope that the trend charts I find or create will help inform the conversation.

I doubt that using some very general variables like total extent and melt season length would give you anything meaningful, as the arctic variability is growing, and noise around the minimum and maximum prevents good definitions.

Data are too noisy to say anything about change points.

Cheers!

I have reposted a chart that etienne has posted here http://forum.arctic-sea-ice.net/index.php/topic,382.2800.html#lastPost which tracks average annual extent behavior. What I find interesting is that there seems to be a noticeable period of relatively lower year to year variability from 1998 through 2006 even as the overall trend throughout is in decline. These two points match similar points in the dispersion chart I posted earlier where I suggested some sort of transition point had occurred. That volume also dropped to a new low in 2007 and remained low from this point forward when compared to prior years suggests to me that these are not simply coincidences but that they are telling us something. I just have no clue what.
« Last Edit: January 02, 2017, 05:46:30 PM by Shared Humanity »

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #342 on: January 02, 2017, 07:42:32 PM »
Another trend chart...this one compares the annual amount of melt for SIA and SIE. There are a couple of not very enlightening observations, a third that simply supports an argument for which Crandles has already provided more than enough support and a 4th observation that raises some interesting questions.

1. SIE and SIA seasonal melt has been increasing from 1979 to 2015.

2. There is a strong correlation between SIE and SIA annual melt trends. Of the 38 years charted, the SIE and SIA seasonal melt quantity trend line moved in opposition only four times (1988, 1999, 2001, 2005).

3. There is a shift upwards in seasonal melt which occurs in 2007. Crandles has, in my mind, clearly shown that it was the devastation of volume and thick MYI in 2007 which has driven this. We have a new, smaller MYI floor which serves to limit how low SIA and SIE can go.

4. In general SIA seasonal melt amounts are greater than SIE seasonal melt although this difference seems to have shrunk, beginning around 2004. From 1979 to 2003, there are only 2 years where SIE melt exceeded SIA melt (1985 9.613 to 9.602) (1990 10.141 to 10.059). Of the remaining 23 years, SIA melt exceeded SIE melt, on average, by 477 Sq Km. In the last 11 years from 2004 to 2015, SIE melt exceeded SIA melt 4 times (2005, 2007, 2011, 2012). Of the remaining 8 years where SIA melt did exceed SIE, the average was by only 186K Sq Km.

Many of these graphs clearly show a transition that occurred in or around 2007. Some of the graphs suggest additional transition points in 1988 and, if you have as vivid imagination as I have, 1998 as well. Finally, these are only trend charts of specific ways of measuring what is going on in the Arctic. What I am most interested in is what these changing trend lines suggest is actually going on with the ice. Things are changing. What is driving these changes in trend lines? Is it weather changes, characteristics of the ice, SST increases?

As always, I haven't a clue.
« Last Edit: January 02, 2017, 08:08:37 PM by Shared Humanity »

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #343 on: January 02, 2017, 09:46:30 PM »
Just wanted to say that I have the time to post these charts because I've been on vacation. I decided to create some charts purely as a discovery process, not sure what I would find.

Here is another, last for the day and perhaps the last for a while. Since I had created a trend chart for dispersion at minimum, I thought I would look at the trend for dispersion at maximum. There is one rather obvious observation. The dispersion at maximum is significantly less than at minimum. Nothing surprising here. The trend chart otherwise is rather weird and the only conclusion I am comfortable drawing is that DrTskoul is correct.

Data are too noisy to say anything about change points.  See Tamino's work on temperature data series.

Cheers!

What appears to be a very mysterious downward trend from 1979 to 1995, followed by a Lord Knows What from 1996 to 2000 and finally by an upward trend beginning in 2001 is simply noise. Claiming a trend here is no different than when a denialist cherry picks a decade and declares a warming hiatus. I have to believe that, if I had 200 years of data, this would be clear. This sudden realization causes me to question most of my observations about the previous charts I have posted.

As a discovery process, it has not been a complete waste of time. I have discovered that, with practice, I can improve my skills at creating graphs in Excel.  :)
« Last Edit: January 02, 2017, 09:53:26 PM by Shared Humanity »

oren

  • Moderator
  • First-year ice
  • Posts: 9805
    • View Profile
  • Liked: 3584
  • Likes Given: 3922
Re: The Slow Transition
« Reply #344 on: January 03, 2017, 12:59:23 AM »
Actually some of these graphs are quite interesting when presented this way. If you have more ideas, please be encouraged to post them.

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #345 on: January 03, 2017, 01:21:23 AM »
Actually some of these graphs are quite interesting when presented this way. If you have more ideas, please be encouraged to post them.

I can probably come up with other possible charts....don't have any right now. To turn them into actual ideas or insights, I'd need to link them to real changes in ice.....perhaps viewing ice animations of extent or looking at the charts of peripheral seas as oren has suggested. Perhaps these apparent transition points are when a new seasonal ice free regime has been established in peripheral seas. Might look at this.

anotheramethyst

  • Frazil ice
  • Posts: 141
    • View Profile
  • Liked: 0
  • Likes Given: 2
Re: The Slow Transition
« Reply #346 on: January 06, 2017, 11:50:05 PM »
Shared Humanity
But you did learn something. There is no apparent correlation between length of freezing and melting seasons with the final totals of ice extent. I wonder what the correlation with freezing degree days and thawing degree days would be.

And that's interesting, because I expected there to be a correlation.  And yet there's not one!  Fascinating!  Thanks for sharing that!

Shared Humanity

  • Guest
Re: The Slow Transition
« Reply #347 on: January 15, 2017, 04:11:08 PM »
Shared Humanity
But you did learn something. There is no apparent correlation between length of freezing and melting seasons with the final totals of ice extent. I wonder what the correlation with freezing degree days and thawing degree days would be.

And that's interesting, because I expected there to be a correlation.  And yet there's not one!  Fascinating!  Thanks for sharing that!

Actually, these charts are looking to determine whether there is a correlation between length of freeze or melt season and the actual increase or decrease in SIA and/or SIE during that season rather than the final totals of SIE or SIA. The Y axis on the chart is the total square kilometers of melt or freeze (the difference between min and max in SIA/SIE for a particular melt or freeze season) that occurred during that season.

icefisher

  • New ice
  • Posts: 60
    • View Profile
  • Liked: 6
  • Likes Given: 0
Re: The Slow Transition
« Reply #348 on: February 06, 2017, 03:45:36 AM »
This Nov 2016 paper from Dirk Notz and Julienne Stroeve equates 1000 additional gigatons of CO2 emissions with the total loss of September Arctic Sea Ice.  At the current annual rate of approximately 40 gigatons Arctic Sea Ice during late Summer will vanish by about 2040 just from CO2 alone.  With additional positive feedbacks -- sooner than later?
http://science.sciencemag.org/content/354/6313/747
doi:10.1126/science.aag2345
Email: dirk.notz@mpimet.mpg.de

longwalks1

  • Frazil ice
  • Posts: 204
    • View Profile
  • Liked: 65
  • Likes Given: 19
Re: The Slow Transition
« Reply #349 on: February 06, 2017, 06:48:09 AM »
Hmm ...

Quote
.  We  define  this  transition  period  to  start  when  the  30 year  mean  September 
Arctic  sea ice  area  in  a  particular  simulation  decreases  for 
the first time to an area that is 10% or more below the simulation’s  minimum  sea
ice  cover  during  the  period  1850–1900,  and  to  end  once  the  30 year  mean  September  Arctic 
sea ice area drops for the first time below 1 million km2  (see
table S1 for specific numbers).

Is this some sanity?  They are not using the 5 consequtive years definition.  i.e.

https://forum.arctic-sea-ice.net/index.php/topic,1559.0.html  and appear to be utilizing area and not extent.