Please support this Forum and Neven's Blog

Author Topic: The Slow Transition  (Read 102724 times)

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #50 on: July 26, 2014, 07:25:41 PM »
Thanks Steven,

I've not read that paper, very much appreciated.

greatdying2

  • ASIF Citizen
  • Posts: 387
    • View Profile
Re: The Slow Transition
« Reply #51 on: July 26, 2014, 09:18:49 PM »
Chris, thanks very much for your detailed reply. The data is very interesting and I agree with most of your conclusions.

It seems that we are moving towards a state where basically all the ice that melts out each summer grows back each winter, and conversely, where all the ice that grows each winter basically melts out each summer. And the volume of ice in this cycle will be less than is currently melting out, because it will be much thinner. But I don't see why this suggests that there will be a "slow transition". Or perhaps that's not your assertion?
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.

John_The_Elder

  • ASIF Lurker
  • Posts: 21
    • View Profile
Re: The Slow Transition
« Reply #52 on: July 27, 2014, 12:24:42 AM »
I have been lurking, reading, and learning for a few years and I thought I would join the discussion. With the transition from MYI to FYI are we not entering an era where the ice will melt at lower at lower temps. At the same time SSTs are rising at a very rapidly. As well " The Arctic-wide melt season has lengthened at a rate of 5days decade from 1979 to 2013" link...http://www.see.ed.ac.uk/~shs/Climate%20changeArctic%20iceChanges%20in%20Arctic%20melt%20season%20and%20implications%20for%20sea%20ice%20loss%20-%20Stroeve%20-%202014%20-%20Geophysical%20Research%20Letters%20-%20Wiley%20Online%20Library_files/grl51412_002.pdf
With the extra time to melt, lower melt point, less time for winter ice to grow are we not approaching the tipping point. All it will take is a big summer storm or a warm winter and the Fat Lady will climb onto the stage.
Back to lurking
John
John

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #53 on: July 27, 2014, 08:39:07 AM »
Great Dying,

Moving to the regions, rather than the hole Arctic Ocean. The following plots show April volume in red, and the melt season volume loss from April to September.

Beaufort has already transitioned to a seasonally sea ice free state. Note how after 2007 the melt season losses are nearly equal to the April volume in most years, which implies virtually no ice at the end of the season.


The same thing has happened since 2007 in the East Siberian Sea.


Meanwhile the Central Arctic is some way off summer losses equalling April volume.

If we take the quadratic fit for summer losses and April volume then the two curves meet around 2020.

However I've been arguing that the winter losses are unlikely to fall much further because first year ice is starting to dominate and as long as the summer extent is not zero ice will continue to be added to the multi-year ice off the CAA and northern Greenland which will thicken to more than 2m thick. If we assume that both of these factors keep central Arctic thickness at around 2.2m thick, it was 2.36m from 2010 to 2014 in April, but 2.2m is more typical of Beaufort and ESS (there is wiggle room here!), then volume will be around 10k km^3 and won't fall greatly below that in the short term. In that case summer losses equal april volume in around 2029 (plus or minus a few years)

But the thinning of ice between 1m and 2m across the whole PIOMAS domain has been (I estimate) around 1.5cm per year. Add this into the equation and the two lines meet in around 2026.


However, my previous comment's reasoning suggests that summer losses won't continue along the summer loss quadratic trend line I've drawn on the above two plots, so it is likely that the summer losses will not continue on the trend in the above two images.

It is possible that the winter volume will decrease more rapidly. The thinning of 1.5cm per year for ice between 1m and 2m is derived by comparing the periods 1986 to 1985 and 2009 to 2013. With continuing warming of ocean and the winter atmosphere it is possible that there will be rather more winter thinning recently - although this isn't easy to work out. This might offset the summer volume loss being less than the quadratic fit shown in the above two images and keep the earliest date of ice free in the last half of the 2020s, but equally it could be in the first half of the 2030s (similar to what Overland & Wang find). My gut feeling is it is not likely to be much later than that.

Of course virtually sea ice free (<1M km^2 NSIDC September Extent) will happen earlier, and I guess that a year like 2012 in a decade may be enough to give no ice by the end of the melt season, but that I don't think would be sustained.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #54 on: July 27, 2014, 08:43:19 AM »
I have been lurking, reading, and learning for a few years and I thought I would join the discussion. With the transition from MYI to FYI are we not entering an era where the ice will melt at lower at lower temps. At the same time SSTs are rising at a very rapidly. As well " The Arctic-wide melt season has lengthened at a rate of 5days decade from 1979 to 2013" link...http://www.see.ed.ac.uk/~shs/Climate%20changeArctic%20iceChanges%20in%20Arctic%20melt%20season%20and%20implications%20for%20sea%20ice%20loss%20-%20Stroeve%20-%202014%20-%20Geophysical%20Research%20Letters%20-%20Wiley%20Online%20Library_files/grl51412_002.pdf
With the extra time to melt, lower melt point, less time for winter ice to grow are we not approaching the tipping point. All it will take is a big summer storm or a warm winter and the Fat Lady will climb onto the stage.
Back to lurking
John


John I have argued in the past that we're approaching a tipping point.
http://dosbat.blogspot.co.uk/2013/12/go-on-say-something-outrageous.html
But if winter volume doesn't decline rapidly then peak thickness isn't going to descend into the region of rapid losses.

Nightvid Cole

  • ASIF Citizen
  • Posts: 421
    • View Profile
Re: The Slow Transition
« Reply #55 on: July 27, 2014, 03:14:02 PM »
I think the argument that the decline in max volume will flatten is likely correct.

But that's the only part I agree with. If increases in melt are due to more open water, are you claiming that there will be less open water? Because it seems the opposite is what is happening and will continue to happen. More open water earlier in the season, as well as many other factors all pushing the system towards more melt.

It seems to me that the argument that increased melt is caused by declining max volumes is putting the cart before the horse.

Not less open water but the rate of increase in open water will slow down if the max volume declines at a slower rate.

Why cart before the horse? A melt season starts at the max volume doesn't it? If max volume was only 15K km^3, I think we would all agree that the ice would melt out. If you don't think that declining max volume causes the increases in melt when the literature is full of talk of albedo feedback being the dominant cause of change during the summer, what do you think is the main cause?

A decrease in true ice concentration is *not* the only way to lose albedo. Earlier snowmelt is another. How do you know it is a true loss of ice rather than a loss of snow on top of the ice which accounts for the greater drops in volume? I am not convinced. The volume anomaly drop on PIOMAS occurs from approximately April 25 to June 25. At that time of year, there is very little true open water in the non-peripheral Central Arctic Basin. Melt ponds, sure (esp. in the second half of June), but actual surface of the deep blue sea being exposed, no.

As long as the SNOW melt continues to advance earlier into spring, more solar energy will be absorbed and ice thickness will drop by an increasing amount, even if the spring ice thickness flatlines.

Steven

  • ASIF Citizen
  • Posts: 361
    • View Profile
Re: The Slow Transition
« Reply #56 on: July 27, 2014, 03:18:05 PM »
I've not read that paper, very much appreciated.


Chris, I think it's interesting that the models in that paper suggest that the trend of declining snow depths has already been going on for some time now (although at a slow pace, which is expected to accelerate throughout the 21st century).  This seems consistent with observations at Soviet drifting stations on multi-year Arctic sea ice between 1954-1991, which suggest a small downward trend of snow depths.  Although of course it's possible that this is partly due to natural variability during that period.

Warren et al. 1999, Snow Depth on Arctic Sea Ice,
http://www.atmos.washington.edu/~sgw/PAPERS/1999_Arctic_snow.pdf


Weak negative trends of snow depth are found for all months. The largest trend is for May, the month of maximum snow depth, a decrease of 8 cm over 37 yr, apparently due to a reduction in accumulation-season snowfall.






On the other hand, 1991 is more than 20 years ago, and it seems difficult to find more recent papers of the same style as the above Warren et al. paper.  Anyway the percentage of first-year ice (compared to multi-year ice) increased notably in recent years.  Kurtz and Farrell 2011 suggests that spring snow depths on first-year ice in the Arctic Ocean are about 50% lower than the corresponding climatological values for snow depth on multi-year ice.
http://gsfcir.gsfc.nasa.gov/download/authors/22391/Journal%20Articles_22391
 
« Last Edit: July 27, 2014, 04:48:58 PM by Steven »

Comradez

  • ASIF Lurker
  • Posts: 26
    • View Profile
Re: The Slow Transition
« Reply #57 on: July 28, 2014, 03:43:48 AM »
I've been thinking about the impact of a freak melt-out year on subsequent years and whether such an event would be a permanent tipping point, or if the ice would quickly recover and return to the downsloping mean, and here's what I've come to conclude:

It depends very much on where the anomalously warm water at the end of the melt season is.  The CAA, Hudson Bay, or the Beaufort will always refreeze.  We saw this in 2012, where the water in parts of the Beaufort got up to 20 C during the summer, and sure enough, it re-froze quite solid over that winter.  These basins are relatively isolated from warm currents. 

On the other hand, I could imagine a year where there is a nearly complete melt-out, where the Chukchi, the Bering, the Barents, and the Kara soak up tons of heat and never properly re-freeze, setting the stage for a big melt the next year.  This happened a bit over the winter of 2011/2012, where the Kara sea never properly re-froze until early February (and then only to about 1 meter by April), which caused it to clear out abnormally quick during the summer of 2012 and soak up tons of heat for the whole Arctic basin.  These four basins are sufficiently connected to warm currents that I don't think it is a given that "the ice will always re-freeze to 2 meters" if given an early-summer wallop followed by clear skies during the peak of the season so that they really start to warm up.

jdallen

  • ASIF Upper Class
  • Posts: 2415
    • View Profile
Re: The Slow Transition
« Reply #58 on: July 28, 2014, 04:17:55 AM »
I have been lurking, reading, and learning for a few years and I thought I would join the discussion. With the transition from MYI to FYI are we not entering an era where the ice will melt at lower at lower temps. At the same time SSTs are rising at a very rapidly. As well " The Arctic-wide melt season has lengthened at a rate of 5days decade from 1979 to 2013" link...http://www.see.ed.ac.uk/~shs/Climate%20changeArctic%20iceChanges%20in%20Arctic%20melt%20season%20and%20implications%20for%20sea%20ice%20loss%20-%20Stroeve%20-%202014%20-%20Geophysical%20Research%20Letters%20-%20Wiley%20Online%20Library_files/grl51412_002.pdf
With the extra time to melt, lower melt point, less time for winter ice to grow are we not approaching the tipping point. All it will take is a big summer storm or a warm winter and the Fat Lady will climb onto the stage.
Back to lurking
John


John I have argued in the past that we're approaching a tipping point.
http://dosbat.blogspot.co.uk/2013/12/go-on-say-something-outrageous.html
But if winter volume doesn't decline rapidly then peak thickness isn't going to descend into the region of rapid losses.


I'm sure you have the right of it here Chris; the key to summer melt out will be winter volume.

The net annual energy budget of the arctic has not changed dramatically.  What has changed is the total sensible heat available at the end of the melt season, and the additional heat which is brought into the arctic, primarily by currents, but also by atmospheric exchange.  The atmospheric exchange does not increase heat in the winter, as much as it reduces heat lost to radiative loss.  The real 500lb gorilla heat-wise is the influx of heat coming in via the Gulf Stream, combined with heat picked up by the much larger areas of open water, which get sealed under ice at the end of the melt season.  There simply isn't enough time during the refreeze for all of that heat to escape.

That heat is what will take us to the tipping point.  It's a brain-twisting concept, but for ice to be retained, the heat taken up by melting during the summer must be *less* than the heat released by ice in winter when it freezes.  The available heat in the top 100 meters or so of the water column will be key to determining whether or not that happens.
This space for Rent.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #59 on: July 28, 2014, 06:44:44 PM »
Thanks for the comments everyone. I'll reply tomorrow night, it's been a long and busy day and I'm tired. Work tomorrow will be lighter.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #60 on: July 29, 2014, 06:32:12 PM »
Steven,

The far greater end of season open water and implied increases in evaporation and humidity within the Arctic Ocean do indeed mean that the Warren results may not be applicable now. However the modelling based paper Hezel et al is probably right, especially after 2007, the mechanism of later freeze is basic.

I am very surprised at the lack of results in Google Scholar on Arctic snow. The only additional paper I found seems to omit the Arctic Ocean, concentrating on the Arctic land.
ftp://ftp.cira.colostate.edu/ftp/Liston/papers/first_author/2011b.liston.JCLIM.pdf
The changing cryosphere: Pan-Arctic snow trends (1979-2009), Liston et al, 2011
It too is a model based study, although it uses two interlinked specialised models.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #61 on: July 29, 2014, 07:07:18 PM »
Comradez,

Tietsche et al, Recovery mechanisms of Arctic summer sea ice, 2011, is a paper that was discussed over at Neven's blog on several occasions before the forum started up.
http://onlinelibrary.wiley.com/doi/10.1029/2010GL045698/full

...abrupt partial loss of Arctic summer sea ice is a common feature of those runs [Holland et al., 2006]. Surprisingly, these abrupt partial losses are often followed by an equally rapid temporary recovery. This suggests that Arctic sea ice has a preferred equilibrium state that varies smoothly with the climatic forcing, and that there are recovery mechanisms that counteract the destabilizing ice–albedo effect after abrupt losses.

...we find that even dramatic perturbations of summer sea-ice cover in the Arctic are reversible on very short time scales of typically two years. This suggests that a so-called tipping point, which would describe the sudden irreversible loss of Arctic summer sea ice during warming conditions, is unlikely to exist.


Their experiment involves removing the sea ice from a model at several points in the model projection. They find that: "...sea-ice free summer conditions cause the ocean to gain excess heat through the surface during summer, but they also cause enhanced heat loss through the surface in the following autumn and winter, when the insulating sea-ice cover is anomalously thin."

Copy of their figure 1 attached. Each decade a run has ice reduced to zero, on 1 July the ice in the model is converted to water, the blue lines show what happens after this is done. The reference run ensemble, in which sea ice zeroing is not done is shown with its standard deviation of the ensemble in dashed lines. When ice is removed it rebounds to the local equilibrium extent.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #62 on: July 29, 2014, 07:27:31 PM »
JD Allen,

I agree that the heat retained, and its impact on winter thickness is critical. But Tietsche et al, and the PIOMAS based model study Zhang et al, (2010, Arctic sea ice response to atmospheric forcings with varying levels of anthropogenic warming and climate variability.), also the models used in the Overland and Wang* studies that find sea ice free is likely in the 2030s, do not find that heat retention is significant compared to heat loss. I say that because if the modelled ocean were retaining heat leading to winter thinning such that we could expect a fast transition to a seasonally ice free state, then such a fundamental feature would be expected to drive such a transition in the models.

But when I look at the following graph, bearing in mind what PIOMAS shows...



I can't help but feel that what we have been seeing is not just weather driven (stochastic) loss, but is a fundamental process. I still think that despite heat loss in the autumn dispensing of most of the energy that has previously gone into volume loss* we may seee a faster transition than that suggested by the three studies mentioned above. But I think that earlier than the mid 2020s, as implied by thinning of ice between 1m and 2m thick, is looking very unlikely.

*Overland & Wang: http://www.agu.org/pubs/crossref/pip/2012GL052868.shtml#content
*energy involved in volume loss, as outlined earlier: http://forum.arctic-sea-ice.net/index.php/topic,933.msg32318.html#msg32318

jdallen

  • ASIF Upper Class
  • Posts: 2415
    • View Profile
Re: The Slow Transition
« Reply #63 on: July 29, 2014, 07:39:48 PM »
Thinking about Tietsche, there are good arguments there which suggest against a "abrupt" tipping point; however, that is a perception based on human time frames and bias.  In geological scale, two years or two hundred is a lighting strike.

To reach a persistently ice free state in summer will require a much greater reservoir of heat at depth in the Arctic Ocean. I am coming to think that is actually much more important than seasonal ice volume. 

I think the net change in heat has reached the point that we will see a high degree of volatility in coverage - modest changes in weather will unlock or lock away the heat on a seasonal basis, reflected in metaphorical "highs" and "lows" sloshing over the edges of a statistical bucket which is filling steadily.  At some point the flow out will become continuous and match inputs. That's the point we reach our new stable state.
This space for Rent.

wili

  • ASIF Upper Class
  • Posts: 1936
    • View Profile
Re: The Slow Transition
« Reply #64 on: July 29, 2014, 08:02:39 PM »
"To reach a persistently ice free state in summer will require a much greater reservoir of heat at depth in the Arctic Ocean"

Could you explain a bit more clearly what you mean by this? IIRC, the water below the relatively fresh (for now) surface water gets warmer and saltier as you go down into the Arctic Ocean (although it's more complicated than that with a number of different strata, but you get the general idea).

The point is, it is  the coldest and freshest surface layer that just needs to get a bit warmer and/or saltier and freezing becomes much more difficult. We don't have to warm the whole ocean to get there, since much of that ocean is already fairly warm compared to the surface.

(Others, better informed than I, can now step in and correct me or add nuance...)
"A force de chercher de bonnes raisons, on en trouve; on les dit; et après on y tient, non pas tant parce qu'elles sont bonnes que pour ne pas se démentir." Choderlos de Laclos "You struggle to come up with some valid reasons, then cling to them, not because they're good, but just to not back down."

Peter Ellis

  • ASIF Middle Class
  • Posts: 530
    • View Profile
Re: The Slow Transition
« Reply #65 on: July 29, 2014, 08:30:16 PM »
Thinking about Tietsche, there are good arguments there which suggest against a "abrupt" tipping point; however, that is a perception based on human time frames and bias.  In geological scale, two years or two hundred is a lighting strike.

"Tipping point" in this context has nothing to do with the speed of the transition and everything to do with the reversibility of the transition.  If there is a "tipping point" then once the ice has been lost it will not return even if the climate cools back down again.

A good example of a tipping point would be the Greenland ice sheet, which is sustained not just by cold temperatures but also its high elevation.  If it melted down to the bedrock, then even chilling the climate back to pre-industrial levels wouldn't make it regrow again, because any accumulation would have to start from (warmer) zero altitude.  It would quite literally never get off the ground again.

The model experiment Tietsche et al did demonstrate conclusively that (at least in this model) there is no tipping point for Arctic sea ice.

In terms of applicability to the real world, it says two things:

1) If there's an unexpectedly low year due to weather promoting high melt, then the ice will recover back to the trend, it's not "locked in" to the lower level.  This has been well borne out looking at 2007-2009 and 2012-2013/14

2) If the climatology recovers to pre-industrial levels, so will the ice.  This unfortunately is a fantasy: the CO2 we've kicked out to date is going to stay there for thousands of years.  About all you can say is that there is a slim chance that geo-engineering could work, as opposed to no chance.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #66 on: July 29, 2014, 08:33:24 PM »
I think the greatest issue in increasing variability is thinner ice. So for a given summer melt, it is easier to reveal open water, but equally, weather variability from year to year can cause greater variation in extent in September than for thicker ice.

I'm not persuaded about the role of deep Atlantic Water, it seems to me that the depth of that water in the deeper parts of the Arctic Basin presents a substantial barrier to interaction with the ice. Where the Atlantic does interact with the ice, Barents and Kara, we do see its effects. Chukchi is similarly influenced by the Pacific, but the gateway there is much smaller.

I wonder sometimes about AGW and Hollywood Movies. Are we not only driven by evolution to only be able to deal with immediate threats, but are we also conditioned by the Movie industry to expect disasters that will unfold in a short human timescale? Taking a step back and a transition that takes from 1995 to 2035, 40 years, or a bit more than a human generation, is very fast.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #67 on: July 29, 2014, 08:37:41 PM »
Peter,

Minor niggle. A bifurcation is a tipping point with hysteresis, and is often (in a practical sense) irreversible, but in the literature there are 'reversible tipping points', which exhibit no hyteresis.


Lenton 2012 "Arctic Climate Tipping Points".

Peter Ellis

  • ASIF Middle Class
  • Posts: 530
    • View Profile
Re: The Slow Transition
« Reply #68 on: July 30, 2014, 12:52:15 AM »
That's why I said "in this context", because the whole point of the Tietsche et al thought experiment / model is to check for reversibility - i.e. that's the definition of tipping point they happen to be using for that paper.

jdallen

  • ASIF Upper Class
  • Posts: 2415
    • View Profile
Re: The Slow Transition
« Reply #69 on: July 30, 2014, 02:27:41 AM »
"To reach a persistently ice free state in summer will require a much greater reservoir of heat at depth in the Arctic Ocean"

Could you explain a bit more clearly what you mean by this?

<snippage>

The point is, it is  the coldest and freshest surface layer that just needs to get a bit warmer and/or saltier and freezing becomes much more difficult. We don't have to warm the whole ocean to get there, since much of that ocean is already fairly warm compared to the surface.

(Others, better informed than I, can now step in and correct me or add nuance...)


You are exactly correct in your assessment of the current structure of the arctic water column.

That said, increased heat will change the thickness of the layers between the thermoclines and haloclines, to the point eventually where they increasingly will become vulnerable to changes at the ocean surface.

For example, lets look at a buoy:

http://www.whoi.edu/page.do?pid=132476

You can see we have very deep warm water below 300M which is close to 1C

You have a mid layer, which depending on the dominant current, can be cooler if flowing out of the region, or in the case of the buoy below, much warmer:

http://www.whoi.edu/page.do?pid=120916

Then, the upper 100 meters or so can be split into 2 or three layers with cold fresher water on top above a warmer layer starting at 25M or so, transitioning  back to colder water below 50M.

Thickness will vary.

The key to my thinking has to do with the heat in the regions below 200 meters.  Right now, there is typically a fairly thick region of cooler water between deeper warmer layers, which when the surface is disturbed still tends to protect the ice from the huge reservoir of heat that exists at depth.

As the net heat content of that deeper water increases, as it has steadily for the last 3 decades, that heat will tend to move up.  At some point, it will be at shallow enough depths that both conductive heating and mixing will make more of it available at the surface with requisite impact on the ice.  The potential for sudden melt outs (as happened in 2012) will increase.

Warm enough - if that upper layer intruding from the Atlantic gets to be 2-4C as I expect it will - that mass of heat could cause melting during storms late into the winter.
This space for Rent.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #70 on: July 30, 2014, 10:22:23 AM »
"
Critical is understanding what is leading to the increased loss of volume during melt seasons. It is increased amounts of ice in thinner categories leading to greater open water formation and greater ice albedo feedback. If thickness merely tends to a nominal 2m in April then further gains through April thinning will not happen. The process will instead be more gradual.
"

[Ignoring minor details like the fact that I don't know what I'm talking about...]

The big areas of open water are the Lena River outflow in the Laptev Sea; the MacKenzie river outflow in the Beaumont, and the gulf stream and Ob river outflow.  These rivers and ocean currents will continue to efficiently transfer increased southern heat into the arctic.  Thus, these areas will melt faster and grow larger.  Because these pools of open water area are relatively small, small growth in the perimeter of these areas represents a large percentage increase in open water area.  [We can counter-argue that the Gulf Stream is a large area.  However, there's a small portion of this area near the central arctic ice.]  The larger open water areas will more efficiently capture solar energy, keeping summer melt volume on an increasing trajectory.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #71 on: July 30, 2014, 07:31:08 PM »
That's why I said "in this context", because the whole point of the Tietsche et al thought experiment / model is to check for reversibility - i.e. that's the definition of tipping point they happen to be using for that paper.

Understood.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #72 on: July 30, 2014, 07:37:57 PM »
Cesium62,

But the regions into which those rivers empty tend to melt out largely in recent years. So I don't see any gain in loss. The process you outline is probably already playing a role.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #73 on: July 31, 2014, 08:33:03 AM »
[Continuing to just chat here to chew up excessive free time...]

Allow me to paraphrase your argument, Chris, and you can point out where I'm completely missing the boat:

The winter arctic is going to be sufficiently cold for a hecka long time that ice will form pretty much everywhere inside the arctic circle except may about 60 degrees of arc north of Europe.  And once ice starts to freeze, it quickly reaches maximum volume.  And there's not much thick multi-year ice left; so we won't see much decrease in April maximum ice volumes.  Meanwhile, the ice won't melt much faster.


The argument that the April ice volume will mostly plateau sounds like a reasonable argument to me.  I think there's room for the volume to drop another 10%, but the basic argument looks nice to me.  I don't particularly understand the part about the ice not melting much faster.

The parts that largely melt out in recent years are next to the central arctic basin.  So if the parts that do melt out melt out a little bit faster, it allows more heat to accumulate next to the part that currently doesn't melt out.  Meanwhile, the regions that largely melt out have relatively small areas today.  So relatively minor increases in the absolute area melted out early in the season lead to relatively large percentage increases in the amount of heat being absorbed early in the season.

I'm arguing that the heat flow toward the central arctic basin will continue to increase rapidly and thus we will continue to see increasing melt volumes.

It's not a question of whether the river delta areas melt out.  It's a question of when they melt out and how much heat is absorbed by the open water when that melt out occurs.  The integral of the open water area over time "near" the central arctic basin will continue to grow.


On a completely different subject...  I greatly enjoyed the graph showing the amount of ice that would need to melt to get to an ice-free September.  Visually, the longer red lines tend to look further away from their blue neighbors than the shorter red lines do.  This suggests that the average is doing screwy things.  I'm wondering what the graphs would look like using the 10-day medians instead of the 10-day average.

 

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #74 on: July 31, 2014, 04:23:07 PM »
[Continuing to just chat here to chew up excessive free time...]

Allow me to paraphrase your argument, Chris, and you can point out where I'm completely missing the boat:

The winter arctic is going to be sufficiently cold for a hecka long time that ice will form pretty much everywhere inside the arctic circle except may about 60 degrees of arc north of Europe.  And once ice starts to freeze, it quickly reaches maximum volume.  And there's not much thick multi-year ice left; so we won't see much decrease in April maximum ice volumes.  Meanwhile, the ice won't melt much faster.


That's OK.

The argument that the April ice volume will mostly plateau sounds like a reasonable argument to me.  I think there's room for the volume to drop another 10%, but the basic argument looks nice to me.  I don't particularly understand the part about the ice not melting much faster.

The parts that largely melt out in recent years are next to the central arctic basin.  So if the parts that do melt out melt out a little bit faster, it allows more heat to accumulate next to the part that currently doesn't melt out.  Meanwhile, the regions that largely melt out have relatively small areas today.  So relatively minor increases in the absolute area melted out early in the season lead to relatively large percentage increases in the amount of heat being absorbed early in the season.


But the volume loss is the key issue. If winter volume doesn't continue to drop on the same trajectory as it has from 1995 to 2010 then a much greater summer volume loss is needed to make summer volume loss equal to winter volume implying no (or very little) ice at the end of the season in September.

I've posted a graphic showing past ten day losses of volume from 1979 to 2014, and in red, the scaled up volume loss that would be needed to give zero ice volume by the end of the melt season.
http://forum.arctic-sea-ice.net/index.php/topic,933.msg32325.html#msg32325

The recent behaviour of melt has resulted in at worst 2012. That did not result in an ice free ocean, but it did happen from the same sort of baseline of winter maximum volume that we've seen over the last four years.

I'm arguing that the heat flow toward the central arctic basin will continue to increase rapidly and thus we will continue to see increasing melt volumes.


Maybe they will, but do you have evidence for this?

It's not a question of whether the river delta areas melt out.  It's a question of when they melt out and how much heat is absorbed by the open water when that melt out occurs.  The integral of the open water area over time "near" the central arctic basin will continue to grow.


In the long term it will. Winter volume will decline in the long term, so a given rate of loss of volume will reveal more open water as time goes on due to winter thinning. Also as the region warms marginally greater volume loss can be extpected in summer. I just doubt this will happen so fast it will cause a 'fast transition'.


On a completely different subject...  I greatly enjoyed the graph showing the amount of ice that would need to melt to get to an ice-free September.  Visually, the longer red lines tend to look further away from their blue neighbors than the shorter red lines do.  This suggests that the average is doing screwy things.  I'm wondering what the graphs would look like using the 10-day medians instead of the 10-day average.


I think that's the graph I linked to above. It's not ten day averages but ten day differences in volume. Basically it shows the rate of change of volume. Using medians of the ten day periods would just be a smoothed plot of the typical volume plots.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #75 on: July 31, 2014, 07:07:43 PM »
Maybe they will, but do you have evidence for this?

Nope.  I'm just chatting here, enjoying the hard work the rest of you do.

I think that's the graph I linked to above.

Yes, that one.  You wrote:  "I've then taken the average ten day change of volume for 2007 to 2013, and have manually scaled this until I get zero volume on 18 September. "  I'd like to replace the word "average" in that sentence with the word "median".  Partly, it's a sanity check: if the graph with medians doesn't look like the graph with averages, we may be looking at distributions where the average is not a good description of the distributions. 

Partly it's because it looks to me like some of the outside red lines might get a bit longer and the inside red lines might get a bit shorter.  'course, I can't really tell just eyeballing the tiny pixels.

But...  it's probably time for me to figure out how to extract the data, see if I can figure out how to reproduce your graph, and then draw my own graphs to my heart's content.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #76 on: July 31, 2014, 07:26:34 PM »
I didn't keep the workings for that graph, but here's median compared to average:

Period   Avg   Median
12-Mar   0.6999   0.7270
22-Mar   0.5747   0.5720
01-Apr   0.4861   0.4670
11-Apr   0.2401   0.2990
21-Apr   0.0584   0.0270
01-May   -0.3290   -0.2900
11-May   -0.4794   -0.4580
21-May   -0.8656   -0.8690
31-May   -1.2284   -1.1790
10-Jun   -1.6310   -1.7020
20-Jun   -2.3131   -2.2920
30-Jun   -2.4406   -2.4280
10-Jul   -2.3321   -2.3360
20-Jul   -2.0313   -2.0450
30-Jul   -1.6837   -1.7730
09-Aug   -1.2233   -1.1460
19-Aug   -0.9271   -0.8500
29-Aug   -0.5400   -0.5240
08-Sep   -0.3089   -0.2800
18-Sep   -0.0509   0.0100
28-Sep   0.1786   0.2020

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #77 on: August 01, 2014, 08:35:46 AM »
I didn't keep the workings for that graph, but here's median compared to average:

Thanks!  Yep, those are darn close.

Steven

  • ASIF Citizen
  • Posts: 361
    • View Profile
Re: The Slow Transition
« Reply #78 on: August 02, 2014, 04:01:12 PM »
I am very surprised at the lack of results in Google Scholar on Arctic snow.


Coincidentally, there's a new paper out today about interdecadal changes in snow depth on Arctic sea ice.  The paper analyzes March/April snow depth data from Operation IceBridge between 2009 and 2013.  The paper is paywalled  (Chris I've sent you a PM).

Webster et al. 2014, Interdecadal Changes in Snow Depth on Arctic Sea Ice,

http://onlinelibrary.wiley.com/doi/10.1002/2014JC009985/abstract


Abstract.

... This study assesses spring snow depth distribution on Arctic sea ice using airborne radar observations from Operation IceBridge for 2009-2013..... we compared the recent results with data from the 1937, 1954-1991 Soviet drifting ice stations. The comparison shows thinning of the snow pack, from 35.1 ± 9.4 cm to 22.2 ± 1.9 cm in the western Arctic, and from 32.8 ± 9.4 cm to 14.5 ± 1.9 cm in the Beaufort and Chukchi seas. These changes suggest a snow depth decline of 37 ± 29% in the western Arctic and 56 ± 33% in the Beaufort and Chukchi seas. Thinning is negatively correlated with the delayed onset of sea ice freeze-up during autumn.


1954-1991 snow depth distribution for March/April (left) vs. 2009-2013 IceBridge snow depth (right):



Figure 7. The snow depth distribution resulting from the two-dimensional quadratic equation (Equation 1) fitted to the 1937, 1954-1991 Soviet drifting ice station data (Left panel) and the 2009-2013 IceBridge snow depth products (Right panel) for March and April. Point measurements are indicated by the small circles; snow depths are indicated by color. The grey shading indicates areas where no data are available.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #79 on: August 03, 2014, 09:10:17 AM »
Thanks Steven,

That really is a much greater decline than I had expected, unless one takes the lowest percentage thinning implied by the uncertainty bounds. Assuming the central values are nearer the 'truth', that's got to have a thickening effect on the ice. But the ice hasn't thickened, first year ice has thinned (or at least not changed much) - I covered that on my blog at the end of one post, but can't locate it now.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #80 on: August 04, 2014, 08:18:13 AM »
Going back to comment #5 with the three graphs of april volume and yearly ice loss...

The second and 3rd graphs are not realistic.  If yearly volume melt increases as suggested by those graphs, then april volume will decrease as it does in the first graph.  The multi-year and first year ice is fairly well stirred.  Summer melt is currently melting 60% of first year ice, 60% of second year ice, 60% of each category of multi-year ice.  If summer melt increases to 65% of ice volume, the April volumes will correspondingly decrease.  You would need a weather pattern that preferentially melts first year ice (or preferentially moves first year ice out of the central arctic) in order to produce either graph two or three.

So, the slow transition hypothesis requires that summer melt volume slow down.  But not only does the summer melt volume need to slow down, we can't have any more step function years like 2007 and 2010.  These are step functions because they just happened to be particularly good melt years.  The weather conditions just happened to come together to melt a bit more of the ice than usual, leaving the remaining ice weaker.  We'll continue to get years that are harder on the ice than average. 

Richard Rathbone

  • ASIF Citizen
  • Posts: 340
    • View Profile
Re: The Slow Transition
« Reply #81 on: August 04, 2014, 09:42:46 AM »
Snow thinning negatively correlated with late freezing.

Does this show up as an autocorrelation in local time series? Thick snow => late melt => early freeze => thin snow => early melt => late freeze => thick snow ...

 


crandles

  • ASIF Upper Class
  • Posts: 1704
    • View Profile
Re: The Slow Transition
« Reply #82 on: August 04, 2014, 12:51:37 PM »
Snow thinning negatively correlated with late freezing.

Does this show up as an autocorrelation in local time series? Thick snow => late melt => early freeze => thin snow => early melt => late freeze => thick snow ...


I would have thought
Thick snow => late melt => large minimum area => able to take snow earlier => thick snow

also
Thin snow => early melt => low minimum area => late being able to take snow => thin snow

but how would it switch between these two regimes?

How does your 'early freeze => thin snow' work?

Steven

  • ASIF Citizen
  • Posts: 361
    • View Profile
Re: The Slow Transition
« Reply #83 on: August 04, 2014, 07:13:02 PM »
Snow thinning negatively correlated with late freezing.


The negative correlation is with the snow thickness instead of the snow thinning.  This is a bit ambiguous in the abstract of the Webster et al. paper.  So a later freeze-up in autumn gives rise to reduced snow depths in winter/spring. 

Here's Figure 11 of the paper.  The Figure shows the correlations between sea ice freeze-up dates (left panels) and snow depth distributions (right panels) for the 2009-2013 and 1954-1991 periods.  Greenland is at the bottom right of each panel.  Freeze-up dates are given in Days of Year, ranging from approximately late August to late November.  Snow depths are for spring (March/April).   

E.g., the late freeze-up in the Beaufort and Chukchi seas is related to the reduced snow depths in these regions.



Edit: corrected error
« Last Edit: August 04, 2014, 08:00:17 PM by Steven »

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #84 on: August 04, 2014, 07:43:54 PM »
Going back to comment #5 with the three graphs of april volume and yearly ice loss...

The second and 3rd graphs are not realistic.  If yearly volume melt increases as suggested by those graphs, then april volume will decrease as it does in the first graph.  The multi-year and first year ice is fairly well stirred.  Summer melt is currently melting 60% of first year ice, 60% of second year ice, 60% of each category of multi-year ice.  If summer melt increases to 65% of ice volume, the April volumes will correspondingly decrease.  You would need a weather pattern that preferentially melts first year ice (or preferentially moves first year ice out of the central arctic) in order to produce either graph two or three.

So, the slow transition hypothesis requires that summer melt volume slow down.  But not only does the summer melt volume need to slow down, we can't have any more step function years like 2007 and 2010. These are step functions because they just happened to be particularly good melt years.  The weather conditions just happened to come together to melt a bit more of the ice than usual, leaving the remaining ice weaker.  We'll continue to get years that are harder on the ice than average.


Those extrapolation are merely dumb statistics, they contain no mechanism. As I've explained further up in the thread summer melt is not likely to continue to increase as it has done because it is (IMO) best interpreted as a small increase upon which have been superimposed step increases due to the volume loss events of 2007 and 2010.
http://forum.arctic-sea-ice.net/index.php/topic,933.msg32494.html#msg32494
With a mainly first year pack and a small contingent of MYI, with FYI rebounding rapidly to thermodynamic equilibrium thickness, the ability of MYI to continue a 'memory' of impacts is limited. Thus after future 2012 type crashes the ice will rebound to equilibirum, as has been the case with 2012, there won't be a carry over dragging the ice down to a new equlibirum.

The whole process of volume and area/extent loss will be limited by winter thermodynamic growth. Inflection in loss of all three metrics will result.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #85 on: August 04, 2014, 07:56:21 PM »
Snow thinning negatively correlated with late freezing.

Does this show up as an autocorrelation in local time series? Thick snow => late melt => early freeze => thin snow => early melt => late freeze => thick snow ...


I would have thought
Thick snow => late melt => large minimum area => able to take snow earlier => thick snow

also
Thin snow => early melt => low minimum area => late being able to take snow => thin snow

but how would it switch between these two regimes?

How does your 'early freeze => thin snow' work?

Surely...

Thick snow -> thinner ice -> less volume to melt in melt season -> more end season open water.

Because snow insulates ice reducing heat flux through it.

Thinner snow should have led to increased volume, but evidently that effect isn't as strong as whatever factors (e.g, ocean/atmosphere warming) are driving ice loss and thinning of FYI. However perhaps we could suggest that we would have seen greater thinning of TET were it not for the snow?

crandles

  • ASIF Upper Class
  • Posts: 1704
    • View Profile
Re: The Slow Transition
« Reply #86 on: August 04, 2014, 08:52:50 PM »
Surely...

Thick snow -> thinner ice -> less volume to melt in melt season -> more end season open water.

Because snow insulates ice reducing heat flux through it.

Certainly 'snow insulates ice reducing heat flux through it'.

Perhaps we have:

Thick snow -> thinner ice -> less volume to melt in melt season
but also Thick snow -> albedo high and slower to fall so slower start to melt season

which leads to ambiguous effect on end season open water.


Thinner snow should have led to increased volume, but evidently that effect isn't as strong as whatever factors (e.g, ocean/atmosphere warming) are driving ice loss and thinning of FYI. However perhaps we could suggest that we would have seen greater thinning of TET were it not for the snow?

Yes that certainly makes sense to me.

So if max volume begins to levels out now that MYI is close to as low as it can go, so that melt volume in summer starts to increase at only a slow rate, meaning little snow thickness change then TET might start decreasing at a faster rate. That might keep the max volume declining rather than levelling out and that means the melt volume keeps increasing.

So does that mean the max volume recently is just in a pause but we should now expect it to continue its decline?

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #87 on: August 06, 2014, 07:20:01 PM »
Crandles,

Sorry, busy with my PIOMAS post last night.

Perhaps we have:

Thick snow -> thinner ice -> less volume to melt in melt season
but also Thick snow -> albedo high and slower to fall so slower start to melt season

which leads to ambiguous effect on end season open water.

But I'd have thought (don't know for sure) that once temperatures get above zero snow's albedo falls and as it has less volume it thins much faster than ice.

So if max volume begins to levels out now that MYI is close to as low as it can go, so that melt volume in summer starts to increase at only a slow rate, meaning little snow thickness change then TET might start decreasing at a faster rate. That might keep the max volume declining rather than levelling out and that means the melt volume keeps increasing.

So does that mean the max volume recently is just in a pause but we should now expect it to continue its decline?

It's possible. But bear in mind, the models show thinning winter snow, together with stabilisation of volume loss due to the growth/thickness feedback.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #88 on: August 09, 2014, 02:55:41 PM »
Chris:

I took your data and made graphs for each region using the technique you used to display April volume for the central arctic and melting season losses.

https://docs.google.com/document/d/198DPkjMNK1YQ5ZzgSW41TuS3Xlv-O1eN5tZMMa4TdaQ/pub

On the one hand, the rate at which the trend lines have been converging for the arctic as a whole is higher than the rate at which the trend lines are converging for the central arctic basin.  This agrees with your suggestion of a Slow Transition:  the easy to melt stuff around the edges is gone, and we just have the hard stuff in the middle left to get rid of.


On the other hand, the graphs tell me a different story than what I'm hearing from you.  In the central arctic, the summer ice losses were relatively constant as you point out from 1979 to 2002.  Thin ice was completely melting away; thick ice was partially melting away.  During the winter, what had been completely melted away reappeared as FYI.  What was partially melted away became MYI, but wasn't as thick as before -- once we reach first year ice thickness, additional thickening is slow.

I would expect to continue to see this part of the trend.  MYI still accounts for 40% of the ice.  Ice more than two years old is still 24% of the ice.  The ice still has a memory.  The gap between the April high and the September low is precisely the thicker MYI that can still be lost.

In 2006 you suggest there is a step change.  Although I'm leery of reading too much into random points on a graph, there are a couple of mechanisms that could produce a step change.  The Kara, the Laptev, and the ESS started completely melting away by the end of summer around this time.  The lack of protection on its flanks may have allowed the CAB to drift more easily when the wind was blowing to the south, or for waves to blow up on the CAB ice when the wind was blowing north. 

http://onlinelibrary.wiley.com/store/10.1029/2011GL048970/asset/grl28491.pdf;jsessionid=0E87241F82E517C47B4B2F31B342C3F9.f03t01?v=1&t=hymji8qi&s=b6006ab31e2a3b651206af8a8a7d600901882caf&systemMessage=Wiley+Online+Library+will+be+disrupted+9th+Aug+from+10-2+BST+for+essential+maintenance.+Pay+Per+View+will+be+unavailable+from+10-6+BST.

This paper suggests that it's gotten windier in the winter since 2002 in the CAB increasing the amount of ice drift in winter.  Elsewhere, thinner ice seems more susceptible to drift, suggesting that when the CAB ice thins further, another step change could occur.

http://www.sciencedaily.com/releases/2013/01/130114092521.htm also suggests that thinner ice will be more easily exported through the Fram strait.


I would expect the momentum of the CO2 trapped heat to continue to provide linear increases in the volume of April ice lost during the melting season.  On top of that, the various Arctic feedbacks look to be net positive:  soot, more open water to trap sun and let large waves build up, thinner ice more easily transported, loss of protective ice in areas around the CAB.  It's hard to see how, in the face of a harsher environment, sea ice volume loss would decrease for many years.

The ESS in 1990, the CAA in 1998 and 2010, the Beaufort in 1993...  These all suggest that sea ice can collapse rapidly when the weather changes suddenly.  Still, the CAA should provide an early warning.  I wouldn't expect the CAB to be ice free until after the CAA started being consistently ice free.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #89 on: August 16, 2014, 10:38:03 PM »
Oh noze!  My favorite thread (well, after the secret geoengineering thread of course) has fallen off the first page.  Time to ping it.

http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn_nowcast_anim365d.gif
seems to show that *all* of the ice gains about a meter in thickness over the course of the winter.  This seems to disagree with the model/research that the thicker the ice, the more slowly it adds additional thickness.

Do we think this is a hycomARC model artifact?  Or maybe I'm reading more into the animation than is warranted?  Or maybe I'm confused about something else?

crandles

  • ASIF Upper Class
  • Posts: 1704
    • View Profile
Re: The Slow Transition
« Reply #90 on: August 17, 2014, 01:01:57 AM »
I am confused as to why you linked an animation from March 2014 to Aug 2014 for the purpose of
'seems to show that *all* of the ice gains about a meter in thickness over the course of the winter'.

Why not link
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2013091518_2013091600_153_arcticictn.001.gif
and
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2014031518_2014031600_038_arcticictn.001.gif

While there is thickness growth along edge of CAA and Greenland - this is dynamics eg compression, ridging, slabbing ...

The yellowly green area clearly advects and does not change thickness much. Meanwhile areas that melt out clearly grow 1.5 - 2m.

>"This seems to disagree with the model/research that the thicker the ice, the more slowly it adds additional thickness."

That applies to growth in thickness to thermal equilibrium thickness and not to dynamics. So I don't really see any disagreement with the research.

So I am not sure what you are seeing, or, if it is just confusion, what is causing that.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #91 on: August 17, 2014, 07:52:37 AM »
I am confused as to why you linked an animation from March 2014 to Aug 2014 for the purpose of
'seems to show that *all* of the ice gains about a meter in thickness over the course of the winter'.


Interesting.  Yesterday it was a 365 day animation.  Looks like when they updated the animation to the current date, they also truncated the first N days....
http://www7320.nrlssc.navy.mil/hycomARC/arctic.html is the preceding page.

Why not link
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2013091518_2013091600_153_arcticictn.001.gif
and
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2014031518_2014031600_038_arcticictn.001.gif

'cause the 365 day animation was sooo cool.

While there is thickness growth along edge of CAA and Greenland - this is dynamics eg compression, ridging, slabbing ...

The yellowly green area clearly advects and does not change thickness much. Meanwhile areas that melt out clearly grow 1.5 - 2m.


Aha.  Yes, I hadn't been thinking about that...


Now I'm going to wonder how people come up with maps of two year old ice versus three year old ice.  When you compress a bunch of 1 year old ice with a bunch of two year old ice over the course of a year, how old is the end result if, say, the one year old ice gets pushed on top of the two year old ice...

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: The Slow Transition
« Reply #92 on: August 17, 2014, 10:27:09 AM »
Cesium,

I would expect to continue to see this part of the trend.  MYI still accounts for 40% of the ice.  Ice more than two years old is still 24% of the ice.  The ice still has a memory.  The gap between the April high and the September low is precisely the thicker MYI that can still be lost.


As long as there is substantial open water from Beaufort to the ESS there won't be such a memory. Export of MYI into these regions due to the Beafort once led to the 'Beaufort Flywheel'.
http://www.whoi.edu/page.do?pid=66596

Now that most years, most of the MYI exported into that region melts out, the MYI is not circulating and ageing. There will always be MYI as long as end of season the Arctic Ocean is not ice free. So we can expect a residual of MYI, and if I am correct this will remain the case for some decades. But the ice in the Central Arctic will continually be being replenished, while export out of Fram and into Beaufort will be removing it.



The situation is like that shown in one of Wipneus's plots (which I can't locate). Wipneus did a plot of the year in which by current trends (back in 2012 I think) the grid cell would be ice free. There was a region along the transpolar drift near the pole were the year was much later than the rest of the pack. Similarly look at your plot of Greenland, there the intersection of the loss and April trend lines is much further in the future than in any of the other regions outside the Arctic Ocean.

In both cases this behaviour is because of continual influx of new ice, and export of ice to other regions/outside the Arctic.

********

Fram export area may have gone up. But PIOMAS shows that because of the reduced thickness volume has not increased as much as area. Zhang et al 2012 "Recent changes in the dynamic properties of declining Arctic sea ice: A model study"
From which I quote:
The simulated reduction in ice volume export from
the Arctic Ocean (calculated by integrating r  hu over the
Arctic Ocean, where h and u are ice thickness and velocity,
respectively; equivalent to export from all gates including
Fram Strait) during 2007–2011 is 15% (Table 1 and
Figure 1h). This is expected given the 33% decline in ice
volume. This means that the reduction of ice volume export
from the Arctic tends to slow down the decline of ice volume
in the Arctic. However, the 2007–2011 mean volume export
normalized by ice volume is higher than climatology
(Figure 1h). This suggests that without the decrease in ice
thickness in recent years the increase in ice speed would
have increased the ice volume export out of the Arctic.

My emboldening.

********

I would expect the momentum of the CO2 trapped heat to continue to provide linear increases in the volume of April ice lost during the melting season.  On top of that, the various Arctic feedbacks look to be net positive:  soot, more open water to trap sun and let large waves build up, thinner ice more easily transported, loss of protective ice in areas around the CAB.  It's hard to see how, in the face of a harsher environment, sea ice volume loss would decrease for many years.


Further increase of summer volume loss is inevitable. However I still think, as argued above, that the increases so far are best explained as a more modest trend of increase together with two step jumps, one at 2007, and one at 2010. The graph below showing melt season volume losses, and interannual melt season losses (melt season loss = max - min (daily))



However in this post:
http://forum.arctic-sea-ice.net/index.php/topic,933.msg32304.html#msg32304
The meeting of melt season loss and April volume in the mid 2020s is dependent on a quadratic fit to the whole melt season loss data. So it represents the earliest possible based on the assumptions that I am right about April volume, and the that quadratic holds into the future (I don't think it will).

There has been a net positive feedback. I am arguing it will be opposed by the strong negative feeback of FYI growth.

Wipneus

  • ASIF Governor
  • Posts: 3358
    • View Profile
    • Arctische Pinguin
Re: The Slow Transition
« Reply #93 on: August 17, 2014, 11:36:09 AM »
The situation is like that shown in one of Wipneus's plots (which I can't locate). Wipneus did a plot of the year in which by current trends (back in 2012 I think) the grid cell would be ice free. There was a region along the transpolar drift near the pole were the year was much later than the rest of the pack.

Maybe this plot?


crandles

  • ASIF Upper Class
  • Posts: 1704
    • View Profile
Re: The Slow Transition
« Reply #94 on: August 17, 2014, 02:53:53 PM »
Maybe this plot?



It is a wonderful plot. Wondering if you will at some point do an update to include 2013 or 2013 and 2014?

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #95 on: August 17, 2014, 02:56:15 PM »
It is a wonderful plot. Wondering if you will at some point do an update to include 2013 or 2013 and 2014?

+1

viddaloo

  • ASIF Upper Class
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
Re: The Slow Transition
« Reply #96 on: August 17, 2014, 05:28:40 PM »
I've studied the PIOMAS data a bit to try to find the most consistent change patterns over the period (1979—2014), and I've found it is the melt percentage. Ie the share of April maximum volume that melts away during spring and summer.

This graph isn't really ready or presentable, but you have 1985 to the left and a guestimate for 2014 at the right:

[]

Michael Hauber

  • ASIF Middle Class
  • Posts: 661
    • View Profile
Re: The Slow Transition
« Reply #97 on: August 17, 2014, 11:19:20 PM »

As long as there is substantial open water from Beaufort to the ESS there won't be such a memory. Export of MYI into these regions due to the Beafort once led to the 'Beaufort Flywheel'.
http://www.whoi.edu/page.do?pid=66596

Now that most years, most of the MYI exported into that region melts out, the MYI is not circulating and ageing. There will always be MYI as long as end of season the Arctic Ocean is not ice free. So we can expect a residual of MYI, and if I am correct this will remain the case for some decades. But the ice in the Central Arctic will continually be being replenished, while export out of Fram and into Beaufort will be removing it.



How sure are we about this?  Particularly the lack of memory?  I've been thinking that maybe some role for the beaufort flywheel may help explain some of the recovery over recent years.  2007 wiped out the flywheel.  Read the following as speculation, not certainty: 

In 2008 and 2009 milder conditions allowed some recovery, but with nothing there in 2008 we saw 2008 almost challenge 2007 despite a poor start.  Then in 2010 the flywheel had recovered to the point of a significant tongue of multi-year ice extending into beaufort and chukchi.  But 2010 was unusually hot, saw a major volume loss and this tongue wiped out.  2011 was another strong melting year and kept the pressure up.  Then in 2012 with the flywheel area still week with solid melting weather we saw major weakness appear throughout this area with large amounts of dispersed open water appearing throughout the pack very early in the melting season, eventually leading to flash melting in August - with the GAC 2012 arriving in time to finish the job, but not being the primary cause.  Then in 2013 and 2014 the flywheel has been rebuilding with thicker multi-year ice again extending into this area.  This is why 2014 has had a slower melt despite better melt weather.  Now we have a strong period of late season melting weather.  Will this have a similar impact as 2010 and thin the flywheel area without necessarily resulting in significant extent/area losses?  Maybe less so as it is late in the season, whereas the really hot weather in 2010 was at the height of summer.

A key question for 2013/2014 is how much of the multi-year ice melted as it moved into the beaufort and around towards chukchi, and whether significant amounts survived to circulate around towards greenland in the next year or two.  By numbers there has been very little extent reduction in beaufort in either year, but by obervation it appears there has been a steady transport of ice into the region roughly balanced by melt at the edges.  HYCOM seems to show that there is still a good tongue of thick ice in that region, but it is right up against the edge of the ice pack, and we've seen in past years the same type of set up and the melt edge push through the 2-3 metre ice tongue like nothing much is there....
Climate change:  Prepare for the worst, hope for the best, expect the middle.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #98 on: August 18, 2014, 07:36:47 AM »
I've studied the PIOMAS data a bit to try to find the most consistent change patterns over the period (1979—2014), and I've found it is the melt percentage. Ie the share of April maximum volume that melts away during spring and summer.

This graph isn't really ready or presentable, but you have 1985 to the left and a guestimate for 2014 at the right:
I really like this graph.

I'm getting slightly different graphs.  It looks to me like the central arctic is on a different trendline than the whole arctic (based on applying a linear fit from 1996 through 2014, not shown), so we might expect the arctic as a whole to fall off of its current trend, and/or the central arctic to increase it's trend.

cesium62

  • ASIF Citizen
  • Posts: 239
    • View Profile
Re: The Slow Transition
« Reply #99 on: August 18, 2014, 07:50:34 AM »
There has been a net positive feedback. I am arguing it will be opposed by the strong negative feeback of FYI growth.

The graphs of FYI growth may be interesting (assuming they weren't yet published and I just overlooked them...) 

[I read the graphs as suggesting a single step transition in 2007...]