Support the Arctic Sea Ice Forum and Blog

Poll

What will the NSIDC 2013 Arctic SIE September minimum be?

More than 5.0 million km2
2 (3.8%)
Between 4.75 and 5.0 million km2
4 (7.7%)
Between 4.5 and 4.75 million km2
4 (7.7%)
Between 4.25 and 4.5 million km2
7 (13.5%)
Between 4.0 and 4.25 million km2
10 (19.2%)
Between 3.75 and 4.0 million km2
13 (25%)
Between 3.5 and 3.75 million km2
5 (9.6%)
Between 3.25 and 3.5 million km2
2 (3.8%)
Between 3.0 and 3.25 million km2
2 (3.8%)
Between 2.75 and 3.0 million km2
2 (3.8%)
Between 2.5 and 2.75 million km2
0 (0%)
Between 2.25 and 2.5 million km2
0 (0%)
Between 2.0 and 2.25 million km2
0 (0%)
Between 1.75 and 2.0 million km2
0 (0%)
Between 1.5 and 1.75 million km2
0 (0%)
Between 1.25 and 1.5 million km2
0 (0%)
Between 1.0 and 1.25 million km2
0 (0%)
Between 0.75 and 1.0 million km2
0 (0%)
Between 0.5 and 0.75 million km2
0 (0%)
Between 0.25 and 0.5 million km2
0 (0%)
Between 0 and 0.25 million km2
1 (1.9%)

Total Members Voted: 49

Voting closed: August 11, 2013, 06:10:15 PM

Author Topic: NSIDC 2013 Arctic SIE September minimum: August poll  (Read 150270 times)

deep octopus

  • Grease ice
  • Posts: 559
    • View Profile
  • Liked: 6
  • Likes Given: 17
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #150 on: August 25, 2013, 04:13:56 AM »
Not only cool weather in the Arctic basin, but a delayed spring in Eurasia that not only, it seems, boosted CO2 readings for an additional week in Mauna Loa, but kept snow extents elevated until June. By the time spring arrived in earnest, insolation had peaked and still much of the energy was going into snow melt. Evidently, May 2013 was by and large one of the weakest months for sea ice melt in a long time. Almost every year, ice melt should be accelerating from April to June/July. This is purely for diurnal reasons. Not only did ice area melt about 1 million fewer km^2 this May compared to last, but it actually decelerated from April. By the end of May, 2013 was behind most years of the last decade. If there was a turning point that harmed this year's chances of serious melt, look to what happened in May.

Nevertheless, in spite of this, it's very likely that 2013 will finish with an area close to 2009, which is still within the 2007-2012 cluster for our purposes. And volume could realistically end in the 2010-2012 cluster--2010, as Chris Reynolds often reminds us, was a year of serious losses in volume. The ice is in bad shape going into 2014. This much we should remember.

Bob Wallace

  • Young ice
  • Posts: 3855
    • View Profile
  • Liked: 41
  • Likes Given: 5
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #151 on: August 25, 2013, 05:57:53 AM »
Patterns are not physical forces that cause things to move back and forth.  This would be like the denier favorite claim that present warming is just a 'normal back and forth temperature swing', i.e., a magic force at work.

Most people on this site are very aware of how melting has been happening over the last few years and most expected a continuation of that melt pattern - unless the weather was weird.

This year the weather was weird.  The ice did not melt as much as would normally be expected largely, it seems, because early season cloud cover blocked sunlight.

I suspect everyone made their prediction based on the unwritten caveat of "this is what I think will happen if the weather performs more or less like it has been performing".  Had the poll been for predictions given a very cloudy melt season or for predictions given an extreme 2007-type season I'm certain you would have seen different sets of numbers.

The big question is not why predictions were off.  IMO predictions were invalidated by the unusual amount of cloud cover.  The big question is why the unusual cloud cover?

Did the unusually thin ice fracture, release warmth from underneath and create its own clouds?

If so, this is "new physics", a new weather pattern that could change how the final melt phase continues.  Or maybe something else triggered the clouds.  We'll probably have a better idea later on and that could change the nature of the 2014 predictions.

mabs

  • New ice
  • Posts: 28
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #152 on: August 25, 2013, 07:35:55 AM »
Bob,

I think you misunderstood the argument. We, or at least I, was not talking causation. That belongs in another thread dedicated to causes. We were discussing, in not too technical terms, simple probability density functions, which is appropriate given that this topic deals with predictions. We, I believe, were trying to extrapolate the shape of the probability density function that would produce the line of best fit, given basic properties of probability functions and observed outcomes. That's what all statistical models do: estimate the line of best fit and its slope. All observed phenomena have an underlying, unobserved, probability function that defines them. Once you figure out what the slope and the function of best fit is for past data, you can then turn around and extrapolate into the future to make predictions about what you are likely to see if our estimates and the underlying function is correct. I believe I concluded that we simply don't yet have enough information to correctly estimate (to be read here "eyeball") the slope of the function underlying past observations, therefore it is hard to say when exactly a particular event is likely to be observed (i.e. a SIA smaller than 1 million sq. km.). However, certain basic properties of common probability functions appear to still hold true even in these troubled times, so there is no reason to throw out the window centuries of mathematical modeling.

I usually do not pay that much attention to what 'deniers' do with basic statistical or mathematical concepts because their understanding of them is very poor. It's possible that their use of "patterns" is more or less, somewhere in the vicinity of the universe, where a pattern means line of best fit or probability function. But I have my doubts.... I think they mean the line a 5-year old would draw to connect some dots on a page, which in this case happen to be representations of some data. I will certainly not stop using centuries-old methods and tools because 'deniers' are confused and confusing.
No god and no religion can survive ridicule. No church, no nobility, no royalty or other fraud, can face ridicule in a fair field and live.
-Mark Twain, Notebook, 1888

Neven

  • Administrator
  • First-year ice
  • Posts: 7986
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1165
  • Likes Given: 575
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #153 on: August 25, 2013, 10:19:36 AM »
AndrewP, I took an hour to read through the americanwx forum, and you're clearly not the person I was thinking of, so I take back and apologize for insinuating you were one of the denier types there. That WUWT-objectivity remark got my hackles up a bit too much (there, now you know my weak spot  ;) ).

Again, I have no problem whatsoever with everything else you said.
« Last Edit: August 25, 2013, 12:08:51 PM by Neven »
Il faut comparer, comparer, comparer, et cultiver notre jardin

johnm33

  • Guest
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #154 on: August 25, 2013, 01:16:40 PM »
Nothing to do with AndrewP but as one of the worst offenders/most pessimistic I've been thinking of what I'll say on a 'Why was I so wrong'  thread. But not yet.

Neven

  • Administrator
  • First-year ice
  • Posts: 7986
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1165
  • Likes Given: 575
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #155 on: August 25, 2013, 01:55:55 PM »
We'll have a nice bout of autoflagellation as soon as the minimum hits.  ;D
Il faut comparer, comparer, comparer, et cultiver notre jardin

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #156 on: August 25, 2013, 04:03:38 PM »
NSIDC updated: 5.71142 dropping -28k2

I calculate -29k, one "pixel" off.

Details:
Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -8.7                     8.8                    -5.2
               Kara Sea             Barents Sea           Greenland Sea
                    8.4                    -8.4                    -6.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   42.2                     0.0                   -33.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -23.3                     8.2                   -11.5
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.0                     0.0                   -29.0

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -33.4                     4.9                    -3.4
               Kara Sea             Barents Sea           Greenland Sea
                    3.8                    -2.6                     0.9
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   11.4                     0.0                    -9.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -7.7                     6.5                    -2.3
             Bering Sea          Sea of Okhotsk              Total Area
                    0.0                     0.0                   -31.3

TeaPotty

  • Frazil ice
  • Posts: 319
    • View Profile
  • Liked: 77
  • Likes Given: 119
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #157 on: August 25, 2013, 04:13:06 PM »
I apologize if my post was rude to some here.

I simply have no patience for anyone mentioning WUWT in a climate change argument. Not after my family personally lost so much from Hurricane Sandy.

Btw AndrewP, you mentioned WUWT again after Neven asked you not to.
Please just stop.

6roucho

  • Frazil ice
  • Posts: 296
  • Finance geek
    • View Profile
  • Liked: 0
  • Likes Given: 1
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #158 on: August 25, 2013, 04:13:25 PM »
Speaking as a statistician, I don't think anyone should be beating themselves up about reasonable estimates based on the trend.

Clearly, people with a tendency to guess low are more likely to be right in low years. That doesn't mean they're generally right: it simply means the variables suited their tendencies this time.

What we could beat ourselves up about is misunderstanding the nature of the trend.

Have we done that. Who knows? These are unusual times.
« Last Edit: August 25, 2013, 04:20:46 PM by 6roucho »

Shared Humanity

  • Guest
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #159 on: August 25, 2013, 05:47:41 PM »
NSIDC updated: 5.71142 dropping -28k2

I calculate -29k, one "pixel" off.

Details:
Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -8.7                     8.8                    -5.2
               Kara Sea             Barents Sea           Greenland Sea
                    8.4                    -8.4                    -6.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   42.2                     0.0                   -33.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -23.3                     8.2                   -11.5
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.0                     0.0                   -29.0

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -33.4                     4.9                    -3.4
               Kara Sea             Barents Sea           Greenland Sea
                    3.8                    -2.6                     0.9
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   11.4                     0.0                    -9.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -7.7                     6.5                    -2.3
             Bering Sea          Sea of Okhotsk              Total Area
                    0.0                     0.0                   -31.3


SIA in the CAB appears to be shrinking while SIE is holding up. Has this been the pattern for the last few days? Is this SIA drop mainly occurring in the thinning area that stretches across the center of the CAB? Could we actually finish the melt season with a large area of ice nearly separating from the remaining ice along Greenland and the archipelago?
« Last Edit: August 25, 2013, 06:29:51 PM by Shared Humanity »

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #160 on: August 25, 2013, 07:03:51 PM »
I apologize if my post was rude to some here.

I simply have no patience for anyone mentioning WUWT in a climate change argument. Not after my family personally lost so much from Hurricane Sandy.

Btw AndrewP, you mentioned WUWT again after Neven asked you not to.
Please just stop.

Hurricane Sandy had little, if anything, to do with climate change. It was a rare weather event that could have occurred with or without climate change. It's possible that climate change has slightly increased the probability of such an event. It's also possible it has decreased the probability. Far worse hurricanes have hit Long Islands before climate change. Such as the Long Islands express in 1938.
« Last Edit: August 25, 2013, 07:20:13 PM by AndrewP »

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #161 on: August 25, 2013, 07:10:47 PM »
Speaking as a statistician, I don't think anyone should be beating themselves up about reasonable estimates based on the trend.

Clearly, people with a tendency to guess low are more likely to be right in low years. That doesn't mean they're generally right: it simply means the variables suited their tendencies this time.

What we could beat ourselves up about is misunderstanding the nature of the trend.

Have we done that. Who knows? These are unusual times.

I would say that the biggest chance of that is that people are neglecting the power of the thickness/growth feedback. In a modelling study using PIOMAS Zhang finds that winter growth of ice in response to record lows reduces the rate of volume loss in the future. Whether net Arctic energy gain beats that is something that will only be answered with more years of data.

Quote
The rate of annual mean ice volume decrease relaxes approaching 2050.
This is because, while increasing SAT increases summer ice melt, a thinner ice cover increases winter ice growth. A thinner ice cover also results in a reduced ice export, which
helps to further slow ice volume loss. Because of enhanced winter ice growth, arctic winter ice extent remains nearly stable and therefore appears to be a less sensitive climate indicator.
Zhang et al, 2010, Arctic sea ice response to atmospheric forcings with varying levels of anthropogenic warming and climate variability.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #162 on: August 25, 2013, 07:17:34 PM »

SIA in the CAB appears to be shrinking while SIE is holding up. Has this been the pattern for the last few days? Is this SIA drop mainly occurring in the thinning area that stretches across the center of the CAB? Could we actually finish the melt season with a large area of ice nearly separating from the remaining ice along Greenland and the archipelago?

SIA calculated this way is essentially what CT does, so can be followed on this graph:


 and the concentration maps on the sites of CT and NSIDC.


AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #163 on: August 25, 2013, 07:28:02 PM »
Speaking as a statistician, I don't think anyone should be beating themselves up about reasonable estimates based on the trend.

Clearly, people with a tendency to guess low are more likely to be right in low years. That doesn't mean they're generally right: it simply means the variables suited their tendencies this time.

What we could beat ourselves up about is misunderstanding the nature of the trend.

Have we done that. Who knows? These are unusual times.

I would say that the biggest chance of that is that people are neglecting the power of the thickness/growth feedback. In a modelling study using PIOMAS Zhang finds that winter growth of ice in response to record lows reduces the rate of volume loss in the future. Whether net Arctic energy gain beats that is something that will only be answered with more years of data.

Quote
The rate of annual mean ice volume decrease relaxes approaching 2050.
This is because, while increasing SAT increases summer ice melt, a thinner ice cover increases winter ice growth. A thinner ice cover also results in a reduced ice export, which
helps to further slow ice volume loss. Because of enhanced winter ice growth, arctic winter ice extent remains nearly stable and therefore appears to be a less sensitive climate indicator.
Zhang et al, 2010, Arctic sea ice response to atmospheric forcings with varying levels of anthropogenic warming and climate variability.

Yes this is why extrapolating linear or exponential volume trends has little physical basis in reality. Most of the feedbacks operate in response to SIA.

Actual SIA is determined by externally forced temperature and positive and negative feedbacks. The primary positive feedback is summer albedo loss. The primary negative feedback is winter ice growth.

It is clear that net feedbacks are less than 1. The net feedback is not runaway positive, or else we likely already would have zero ice, and there would never be a bounce back from bad ice years like 2007 and 2012. Bad years would simply accelerate the net runaway positive feedback. This means that equilibrium can be reached and further sea ice losses must be externally driven by external forcing (warming).

The equilibrium at the moment appears to be around 3 million sq km. Further external forcing must occur to drive this equilibrium lower.

Extrapolating SIA trends would have much more basis in physical reality and representing the externally forced trend.
« Last Edit: August 25, 2013, 07:37:01 PM by AndrewP »

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #164 on: August 25, 2013, 08:12:23 PM »
Andrew,

I may post a thread about this topic. I am not convinced 'external' forcing is needed, where  by 'external' I mean heat flux external to the Arctic ice/ocean system. But unless I can see a way through I don't want to derail this thread further - hence may post a separate thread. It also might help to be more specific about 'ice dynamics'.

BornFromTheVoid

  • Nilas ice
  • Posts: 1339
    • View Profile
  • Liked: 679
  • Likes Given: 299
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #165 on: August 25, 2013, 08:41:43 PM »
Update for the week to August 24th

The current 1 day extent is 5,711,420km2, while the 5 day mean is on 5,737,840km2
 
The daily anomaly (compared to 79-11) is at -1,036,980km2, a decrease from -1,127,280km2 last week. The anomaly compared to the 07, 11 and 12 average has increased from +897,450km2 to +1,115,130km2 this week. We're currently 7th lowest on record, compared to 6th last week.
 
The average daily loss over the last 7 days was 33.8k/day, compared to the long term average of 46.7k/day, and the average of the last 5 years of 72.1k/day.

The average long term loss over the next week is 34.6k/day, with the average of the last 5 years being 45.5k/day.

The loss so far this August is the 17th lowest on record, at -1,385,010km2. We need to average a loss of 190.3k/day for the remainder of the month to record the largest August extent drop on record.

The daily minimum extent so far this year is below the annual daily minima of 26 out of 34 previous years, and within 1 million km2 of another 3.





I recently joined the twitter thing, where I post more analysis, pics and animations: @Icy_Samuel

Steven

  • Grease ice
  • Posts: 670
    • View Profile
  • Liked: 273
  • Likes Given: 17
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #166 on: August 25, 2013, 09:04:26 PM »
Yes this is why extrapolating linear or exponential volume trends has little physical basis in reality. Most of the feedbacks operate in response to SIA.

Actual SIA is determined by externally forced temperature and positive and negative feedbacks. The primary positive feedback is summer albedo loss. The primary negative feedback is winter ice growth.

It is clear that net feedbacks are less than 1. The net feedback is not runaway positive, or else we likely already would have zero ice, and there would never be a bounce back from bad ice years like 2007 and 2012. Bad years would simply accelerate the net runaway positive feedback. This means that equilibrium can be reached and further sea ice losses must be externally driven by external forcing (warming).

The equilibrium at the moment appears to be around 3 million sq km. Further external forcing must occur to drive this equilibrium lower.

Extrapolating SIA trends would have much more basis in physical reality and representing the externally forced trend.

Andrew,

Thanks for the information.  I also think that there is no physical motivation for extrapolating the current volume trends, and that the SIA/SIE trends may be a better predictor for the date of an (essentially) ice free Arctic in September.

Chris,

It would be interesting indeed to have a separate thread about this topic, and about the role of ice dynamics, external forcing etc.

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #167 on: August 26, 2013, 01:42:47 AM »
Steven
 Why would you prefer SIA and especially SIE to PIOMAS?
It's very easy to cover the surface of your drink with a tiny amount of crushed ice, it just won't last very long.
Terry

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #168 on: August 26, 2013, 08:03:23 AM »
Steven
 Why would you prefer SIA and especially SIE to PIOMAS?
It's very easy to cover the surface of your drink with a tiny amount of crushed ice, it just won't last very long.
Terry

For the reasons listed above. There is a physical reason that the trend in SIA will continue. There is not a physical reason for the trend in SIV to continue. Instead, there are physical reasons it will not continue.

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #169 on: August 26, 2013, 08:54:29 AM »

Quote
For the reasons listed above. There is a physical reason that the trend in SIA will continue. There is not a physical reason for the trend in SIV to continue. Instead, there are physical reasons it will not continue.
The overall pattern of the last 13 years has been the lowering of SIV, SIE & SIA, therefore since SIV has been headed in the correct direction in 10 of those years while SIE & SIA have only been correct in 7 (barely better than a coin toss) i much prefer SIV.

SIV is, in my opinion a much more reliable measure of what is happening to the Arctic Sea Ice. the SIA and SIE are more likely to be displaying "noise".
The only reasonable explanation for preferring either over volume would be if you believe that in a particular year we have moved past some invisible line that would separate past performance from the expected performance in the coming year.
Not possessing the foresight to know if such a line exists, let alone whether it has been crossed I prefer to base my prognostications on extrapolations that have a better past track record.
Terry

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #170 on: August 26, 2013, 09:13:20 AM »

Quote
For the reasons listed above. There is a physical reason that the trend in SIA will continue. There is not a physical reason for the trend in SIV to continue. Instead, there are physical reasons it will not continue.
The overall pattern of the last 13 years has been the lowering of SIV, SIE & SIA, therefore since SIV has been headed in the correct direction in 10 of those years while SIE & SIA have only been correct in 7 (barely better than a coin toss) i much prefer SIV.

SIV is, in my opinion a much more reliable measure of what is happening to the Arctic Sea Ice. the SIA and SIE are more likely to be displaying "noise".
The only reasonable explanation for preferring either over volume would be if you believe that in a particular year we have moved past some invisible line that would separate past performance from the expected performance in the coming year.
Not possessing the foresight to know if such a line exists, let alone whether it has been crossed I prefer to base my prognostications on extrapolations that have a better past track record.
Terry

The fact that in individual years SIV follows the trend more frequently is irrelevant. Both variables have overall been following long-term independent trends. There is simply more annual variability around the SIA trend. This does not make that long-term trend any less likely to continue or to be the more powerful of the two trends. Both trends cannot continue. The only question is which trend is more strongly grounded in its relationship to temperature, positive and negative feedbacks.

The answer is clearly SIA. The two primary feedbacks (albedo and open-water ice creation) have a direct relationship to SIA. They only have an indirect relationship to SIV through SIA.

In short, over the last 30 years SIA has existed in a changing equilibrium determined by externally forced temperature, positive feedbacks, and negative feedbacks. When warming occurs, SIA decreases to a new equilibrium where the strength of the negative open-water ice creation feedback is strong enough to create enough volume to prevent further SIA loss. More warming occurs. SIA must drop again to another equilibrium where there is enough fall open water to counterbalance the slower rate of ice production and faster rate of destruction.

We are currently averaging about 3 million sq km in SIA over the last 7 years. If this were to drop to 2 million, there would be an incredible increase in fall ice production which would increase volume and bring us right back to 3 million. Thus there needs to be an external source of warming to drive SIA lower.

We witnessed exactly this in 2012-13. Area dropped to 2.2 million, and fall volume production was the the greatest ever by far, bringing us right back to where we were before. I have been saying exactly this for a couple of years now (to you as well), and 2012 provided a perfect illustration.

Oyvind Johnsen

  • New ice
  • Posts: 20
    • View Profile
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #171 on: August 26, 2013, 09:48:53 AM »
AndrewP:
You have an interesting theory. But I have some questions:
Why did the large drop in SIA in 2007 create a new equilibrium-SIA? (Or was 2007 just a manifestation of a new equilibrium about to be established? The question of explanation remains the same.)
Why will the drop in 2012 not? (Apart from the fact that 2013 seems to show this.)
From your general theory, it seems that you would just as well have predicted a rebound from the low SIA-level in 2007, to the "old" equilibrium. Unless you can explain the difference between 2007 and 2012.
By the way, this discussion is obviously OT in this thread.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #172 on: August 26, 2013, 03:58:44 PM »
NSIDC update 5.65206 down 59k4

I calculate -63k9, 7 "pixels" off.

Details:

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -13.8                     6.5                   -18.2
               Kara Sea             Barents Sea           Greenland Sea
                   10.1                     9.4                    -3.9
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -30.3                     0.0                     6.2
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -17.8                    -1.9                   -10.0
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.0                     0.0                   -63.9

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -7.4                     5.9                    -9.7
               Kara Sea             Barents Sea           Greenland Sea
                    4.2                     1.5                    -0.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -5.1                     0.0                     6.2
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -9.3                    -6.2                    -1.8
             Bering Sea          Sea of Okhotsk              Total Area
                    0.0                     0.0                   -21.8

Steven

  • Grease ice
  • Posts: 670
    • View Profile
  • Liked: 273
  • Likes Given: 17
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #173 on: August 26, 2013, 08:37:17 PM »
Steven,
Why would you prefer SIA and especially SIE to PIOMAS?

Terry, here's an addition to the reasons that were mentioned above.

In the last few decades and years, we've seen a transition where much of the thickest MYI (say 6 meters thick) disappeared and most of the ice is now thinner at the start of the melt season (say 2-3 meters thick for the central ice pack).  It's almost inevitable that this seriously biases downward the volume trends, especially for the September values, since the relative difference in thickness between these two types of ice becomes maximal at the end of the melt season.

But this difference in ice thickness is September doesn't mean much for the albedo feedback.  Sunshine is already weak in September.  The albedo feedback is strongest during early summer, when the central ice pack is still about 1.5-3 meters thick and its albedo isn't much different from the albedo of 6 meters thick ice, at least initially.

As for SIE: I don't pay much attention to it during summer, except for its effect on the concentration indices such as CAPIE.  Area is obviously more useful than extent during early summer.  The fact that CT is sensitive to melt ponds makes it particularly useful for keeping track of the albedo feedback (which is why I use CT for making predictions).  But again, albedo is not so important anymore in September.  Therefore I think both September mean extent and September mean area are robust and useful measures.
« Last Edit: August 27, 2013, 07:37:52 AM by Steven »

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #174 on: August 26, 2013, 08:57:12 PM »
I agree with your reasons Steven. Ice production/creation is controlled by area, not thickness. Temperature, albedo, and open-water thermal radiation care very little weather the ice is 1.5m or 6m. There is a tremendous difference between open water and ice. Thus our "buffer" between an "ice-free" state is the remaining 3 million sq km of area. Our "buffer" is not the 4,000 km3 of volume.

Now to Oyvind's questions


AndrewP:
You have an interesting theory. But I have some questions:
Why did the large drop in SIA in 2007 create a new equilibrium-SIA? (Or was 2007 just a manifestation of a new equilibrium about to be established? The question of explanation remains the same.)
Why will the drop in 2012 not? (Apart from the fact that 2013 seems to show this.)
From your general theory, it seems that you would just as well have predicted a rebound from the low SIA-level in 2007, to the "old" equilibrium. Unless you can explain the difference between 2007 and 2012.
By the way, this discussion is obviously OT in this thread.

I believe that 2007 was the manifestation of a new equilibrium in the arctic. The arctic is constantly manifesting, or attempting to manifest (it often can't keep up) new equilibriums between SIA and temperature. Incredible warming of the arctic took place from the late 90s to mid-2000s, with some warming continuing even after that. 2007-2010 was the manifestation of this warming.

At the end of the 2012 melt season I was not sure that 2012 was not a new manifestation. Warming of the arctic had slowed, but perhaps we weren't quite at equilibrium yet and the equilibrium lay near 2013 (or even lower). I thought there was >50% we had overshot the equilibrium, but I could not be sure. What convinced me we had overshot it at least a little bit, was the record ice production from October-January, bringing us right back to where we had been. The continuing gains in volume relative to 2010-2012 during April and May reinforced the probability of at least some overshoot, possible a moderate amount of overshoot.

Again none of these are absolutes. One still probably cannot be 100% sure that our current equilibrium lies near or below 2012, and that 2013 was just a fluke. Although I believe I can be at least 90 or 95% confident the equilibrium is above 2013, and further warming will be required to bring the multi-annual mean SIA below 2012.

Neven

  • Administrator
  • First-year ice
  • Posts: 7986
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1165
  • Likes Given: 575
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #175 on: August 26, 2013, 09:10:27 PM »
What convinced me we had overshot it at least a little bit, was the record ice production from October-January, bringing us right back to where we had been.

Andrew, did you have a look at the 2012/2013 winter, and compare it to previous years?
Il faut comparer, comparer, comparer, et cultiver notre jardin

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #176 on: August 26, 2013, 09:47:30 PM »
Winter ’12-’13 didn’t exactly prelude any reset of volume.
Remember:

In a way, it is also boring to reproduce this again and again. But it confines the conditions spawning the retarded melt season to the period after February.

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #177 on: August 26, 2013, 09:58:56 PM »
Well if you look at the Oct-February period as a whole, 2012 and 2013 look roughly similar in terms of temperature. Oct-Jan 2013 is actually a bit warmer. I chose these two periods because the ice was nearly recovered to 2012 by the end of January, and tied by the end of February.

Better weather might have played some role, but it doesn't look that much better on the whole. January and especially February of 2013 were colder than of 2012 which closed the deficit. But on the other hand, the preceding period was warmer, but the deficit did not broaden. On the whole 12-13 was able to catch 11-12 despite its considerable initial deficit. The open water and thin ice readily thickened, as theoretically expected. Given identical weather, you wouldn't even expect the deficit to be totally erased. You'd expect the deficit to be halved or something. But it was completely erased. Probably partially weather, and partially the greater radiation of open water and thin ice than thick ice.

Neven

  • Administrator
  • First-year ice
  • Posts: 7986
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1165
  • Likes Given: 575
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #178 on: August 27, 2013, 06:05:22 AM »
Looking back at the winter analysis I did on the ASIB, there were a lot of clear skies to radiate through:



In my conclusion I wrote:

Quote
My feeling right now is that despite a couple of differences, this melting season looks similar to last year's melting season and so it will probably beat the 2007 records. Whether it will also beat last year's records is too early to tell, as weather is still an important factor. We used to see rebounds in the Arctic whenever a record low was hit (also known as 'recovery! recovery!' on fake skeptic blogs), but I'm not sure how bouncy the Arctic is anymore.

I don't think we'll see sea ice area numbers dive below 1 million km2, even though we have now entered the Maslowski-period. He went out on an educated limb about 7-8 years ago by saying that the Arctic could be ice-free by 2016, ± 3 years. It probably won't happen this year (although nothing should be ruled out in the Arctic), but another big melt could pave the way for more crashes to come. Or something unforeseen happens and gives the ice some respite. Either way, we'll be here to see it.
Il faut comparer, comparer, comparer, et cultiver notre jardin

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #179 on: August 27, 2013, 08:47:26 AM »
I've been testing regression models for predicting minimum SIA for the years 2005-2012 using early season SIA and SIV. The results are interesting and confirm that a SIA minimum of 3.25 was the best prediction by the 160th day of the year (June 9th).

For the June 9th regression:

The R-squared value is .74, indicating that the model is able to explain 74% of the variance in minimum SIA. This is significantly better than a volume only regression which can only explain 63% of the variance. The P value for the regression is .03, indicating we can reject the null hypothesis that the correlations occur by chance with 97% confidence. The standard error of the regression is 371,000. The model predicted that 2013 would have a minimum SIA of 3.26 million with 95% confidence the minimum would be above 2.65 million. There was 99.7% confidence that the minimum would be above 2012's.

Below are the line fit plots. The first is for SIA. The second is for SIV. The green dot represents the model predicted value for 2013 of 3.26 million (+/-600,000).

« Last Edit: August 27, 2013, 06:41:12 PM by AndrewP »

jdallen

  • Young ice
  • Posts: 3291
    • View Profile
  • Liked: 574
  • Likes Given: 216
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #180 on: August 27, 2013, 09:47:28 AM »
<snippage>
Better weather might have played some role, but it doesn't look that much better on the whole. January and especially February of 2013 were colder than of 2012 which closed the deficit. But on the other hand, the preceding period was warmer, but the deficit did not broaden. On the whole 12-13 was able to catch 11-12 despite its considerable initial deficit.

I guess my question here, is why would 2012/13 winter re-freeze SIA *not* catch up to 2011/12?  We aren't yet at a place where the energy input during the summer/melt season combined with existing enthalpy is even close to preventing a complete refreeze of most of the basin by February.  The only thing that might affect that would be serious circulation between the surface and deeper layers warming the surface vigorously at a rate high enough it would keep the surface about -2C.  That's a pretty tall order, and unlikely except directly over where the relict North Atlantic flow enters the basin (interestingly enough, directly under the low-concentration band sweeping across the middle of the CAB...)

Further feed back question - once the refreeze is complete, we've put a insulating cap on the largest source of heat available, and directly limited the amount of heat that can be radiated out of the atmosphere.  One line of logic here that we missed in spring might be that the February/March break up may have allowed enough heat out of the basin to slow the onset of melt.  That's a bit tough to justify though, because of how quickly the leads refroze, and how rather small area they opened to radiation.  So the question, reached circuitously is, how would a large expanse of generally thinner ice, of approximately the same area make any difference in the melt rate?


The open water and thin ice readily thickened, as theoretically expected. Given identical weather, you wouldn't even expect the deficit to be totally erased. You'd expect the deficit to be halved or something. But it was completely erased. Probably partially weather, and partially the greater radiation of open water and thin ice than thick ice.

Here I'm going to disagree - the physics isn't with you, I think.  Given identical weather as you put it, I'd absolutely expect the refreeze to make up the SIA deficit, as it isn't really as dependent on air temperature, as it is on the lack of insolation.  As I mentioned above, I don't think sufficient enthalpy or inputs exist yet outside of insolation that would significantly slow the refreeze.  Raise the temperature of the entire water column a degree C or so, and I think then we'd see something very different, but that absolutely isn't going to happen year over year.

I think to understand this year, we need to go back to one of your other factors - albedo - and less directly, cloud cover.  One of the notable aspects of late May and June was the storms and cloud cover that entered the arctic, and for the most part, stayed.  That, much more than the end of winter SIA had the most to do with slowing melt.  The lower air temperature isn't enough to explain it, as the ability of warm flow from the continents simply doesn't transport enough heat to make much of a difference other than to remove snow cover and a few tens of CM of surface ice.

I think we need to look more at the interaction between total heat content and salinity in the arctic basin, and what effect that has on circulation and the very narrow dynamic temperature range driving melt.  Thinking in terms of scales of energy, that and total insolation are key, not so much weather. 


If it were clear, even if a degree or so cooler, I think insolation would have overwhelmed that anomaly.  This year, with as much persistent cloud cover as we have had, I think the albedo of the ice itself was far less material to what we saw happening, and as such means the refreeze is less important than other factors.

Regarding volume, I'm coming to think that is less meaningful a factor as practically affecting melt.  As we saw last year, large volumes of thick ice shuttled into the Beaufort did not hinder it melting out.  Rather, volume I think is more a measure of the coherent strength of the pack, and its ability to resist other forces being applied to it.  I think we need to settle the question of just how much difference a "mesh" pack vs. "cottage cheese" pack makes in terms of net loss or capture of heat over time, and whether it is a symptom or contributor to reduction of the pack.  That it *is* having a dramatic, negative effect on terrestrial animal life and sea mammals is almost a certainty.
This space for Rent.

jdallen

  • Young ice
  • Posts: 3291
    • View Profile
  • Liked: 574
  • Likes Given: 216
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #181 on: August 27, 2013, 09:52:43 AM »
I've been testing regression models for predicting minimum SIA for the years 2005-2012 using early season SIA and SIV. The results are interesting and confirm that a SIA minimum of 3.25 was the best prediction by the 160th day of the year (June 9th).

For the June 9th regression:
<snippage>

I believe I understand your numbers here, and I can see they were predictive.  I'll still wondering, is it correlation, or cause?  I'm still dubious in that regard.  That you've been able to find a pattern, there is no question.  I'll leave it to the more math-savvy to evaluate your statistics.
This space for Rent.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #182 on: August 27, 2013, 04:02:26 PM »
NSIDC update, 5.55733 down -94k7

I calculate -96k2, two "pixels" off.

The details, and the area numbers:

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -12.5                   -10.1                    -9.8
               Kara Sea             Barents Sea           Greenland Sea
                  -12.5                   -28.9                     3.9
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -8.6                     0.0                   -10.6
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -8.3                    -2.5                     3.2
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.6                     0.0                   -96.2

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -22.5                     3.7                   -13.7
               Kara Sea             Barents Sea           Greenland Sea
                   -4.9                    -5.9                    -9.9
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -3.0                     0.0                    -7.0
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -9.0                    -1.6                     0.8
             Bering Sea          Sea of Okhotsk              Total Area
                    0.1                     0.0                   -72.8

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #183 on: August 27, 2013, 06:33:22 PM »
I've been testing regression models for predicting minimum SIA for the years 2005-2012 using early season SIA and SIV. The results are interesting and confirm that a SIA minimum of 3.25 was the best prediction by the 160th day of the year (June 9th).

For the June 9th regression:
<snippage>

I believe I understand your numbers here, and I can see they were predictive.  I'll still wondering, is it correlation, or cause?  I'm still dubious in that regard.  That you've been able to find a pattern, there is no question.  I'll leave it to the more math-savvy to evaluate your statistics.

The P-value indicates that there is only a 3% chance the correlation occurs by chance.

And actually the regression using volume alone only gives a 1% chance the correlation occurs by chance. On the other hand, it only explains 63% of the variation in the minimum, while a regression using both SIA and SIV explains 74% of the variation, which is a considerable improvement.

The regression for area alone has an R-squared of 60% and a P-value of .02.

All three regressions are significant at the .01-.03 level, and explain 60-74% of the variance. The regression using both area and volume explains the most variance (74%) and I would consider it to have the most predictive power.
« Last Edit: August 27, 2013, 06:39:27 PM by AndrewP »

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #184 on: August 27, 2013, 06:57:54 PM »
Given identical weather, you wouldn't even expect the deficit to be totally erased. You'd expect the deficit to be halved or something. But it was completely erased. Probably partially weather, and partially the greater radiation of open water and thin ice than thick ice.

I've recently posted this graphic on another thread about another issue.



This is from Thorndike 1975, it shows the sea ice growth rates and their strong dependency on thickness. This is because ice grows by accreting new ice to its base from the ocean, to do this there has to be a heat flux through the ice to the surface, the primary limiting factor here is the insulation provided by ice as it thickens. I can post calculations and graphics if needs be, but the take home message is that the sort of temperature differences seen from year to year in the Arctic have a small effect on heat flux through the ice compared to the thickness/flux relationship, therefore they have a negligible effect on the bulk of the thickening of ice.

Temperatures can impact the equilibrium thickness in the later months of the winter (MAM), but the bulk of the thickening is driven by the thickness of ice. After 2012 there was massive open water, so a massive area of ice was undergoing the high growth rates associated with ice thickness in the curve above. Weather had nothing to do with record volume gain in the post 2012 freeze season - ice dynamics was the main mover.

Neglecting the relatively small losses of MYI volume, given identical weather one would expect a identical April volume maximum. And this would be despite the two years in question having very different preceding area/extent minima and consequent open water.

This is the point of the growth/thickness feedback, as more open water occurs at the end of the melt season so growth of new ice is able to respond, potentially stemming winter thinning.

Shared Humanity

  • Guest
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #185 on: August 27, 2013, 07:40:16 PM »
Thanks Chris. It is a simple post like this (they need to be simple if I am to understand them) that keep me coming back to this site.

Isn't this, then, an argument against a perennially ice free Arctic anytime soon?

AndrewP

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #186 on: August 27, 2013, 07:59:28 PM »
Yes, it is the primary reasons models do not predict an ice free arctic soon, as demonstrated in Tietsche 2011:

http://www.seas.harvard.edu/climate/seminars/pdfs/Tietsche_GRL_2011.pdf


Chris,

I see what you are saying and that is a wonderfully illustrative graphic. However, much of the ice does thicken close to 2m in its first year, and whether it thickens to 2m or simply 1.5m I presume is somewhat weather dependent. Winter volume creation has remained nearly constant for 30 years has it not, despite far more open water at the start of freeze season? I presume that the shorter and warmer freeze season has been cancelling the negative open water feedback.

You're probably right that given identical weather we would have recovered more than half way back to 2012. But given the thinning of the remaining ice in 2012, which is now MYI, it might have taken more than one year of identical weather for a complete recovery. We can see this in the loss of ice in the 2.5m+ bins you've shown. The thickness of FYI created can be held as a constant given identical weather, meaning that winter 12-13 was destined to create more volume of FYI because of the higher area of FYI. But the deficit in the volume and areal coverage of MYI meant that not all of the deficit would be made up, given identical weather. Only with somewhat more favorable weather was that deficit made up.

In short, the average thickness of FYI in February 2013 was greater than in February 2013. This indicates that the weather was slightly better for FYI thickening. The areal coverage, and thickness, of MYI was lower.

We can test this theoretical expectation using PIOMAS. We see quite clearly that the areas of FYI were thicker in 2013 than 2012.

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #187 on: August 27, 2013, 08:20:06 PM »
Thanks Chris. It is a simple post like this (they need to be simple if I am to understand them) that keep me coming back to this site.

Isn't this, then, an argument against a perennially ice free Arctic anytime soon?

Possibly.

Zhang's 2010 paper using PIOMAS for forward projection finds that this effect of rapid growth of ice over winter stems the winter thinning and allows ice to persist much longer than naïve extrapolation of summer volume loss might suggest.
http://onlinelibrary.wiley.com/doi/10.1029/2010GL044988/abstract
However we are now entering a phase where April thickness is entering a region of non-linear relationship with September open water formation (see my last graphics post on the NSIDC _July_ thread). And more open water implies more energy gain during the summer.

It might be thought that I should defer to Zhang's paper, but in that paper where they transition from volume loss using observed weather to volume loss using randomly shuffled weather, there is a clear levelling of the volume loss - which implies to me that shuffling the weather is neglecting the role of weather in the process of volume loss - as I think this year shows.

jdallen

  • Young ice
  • Posts: 3291
    • View Profile
  • Liked: 574
  • Likes Given: 216
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #188 on: August 27, 2013, 09:33:23 PM »
Yes, it is the primary reasons models do not predict an ice free arctic soon, as demonstrated in Tietsche ...

Very nice article, Andrew.

 I agree we are a rather long way from a year round ice free arctic; fundamentally, the total heat content of arctic water needs to come up considerably to offset winter radiative losses.

I do think because of increasingly hostile summer conditions we will see an ice free summer (< 1.1 million KM2) sooner than later.  I'm less certain now of my 2016 prediction for that, but i think that is easily possible now with the right early season conditions... or even later, in some circumstances. That in itself will not prevent a complete refreeze, and will tend to moderate somewhat climate impact. 

Unfortunately, that sort of extreme variability will wreak absolute havoc on arctic ecology.

This space for Rent.

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #189 on: August 27, 2013, 09:54:30 PM »
Andrew,



It is only in recent years that the growth/thickness feedback has become apparent, for most of the series there is substantial weather variability in the relationship between volume gain over autumn/winter and the preceding CT Area minimum.

But this is because for most of the period volume gain has not been from the process of very thin ice thickening or ice growth from open water.

Using the thickness/volume breakdown data that I presume you have access to (given some of your comments): I have calculated the volume difference between September and April of the following year, given for the year in which April falls. This is then broken down into buckets of 1m each, i.e. 0m is 0 to 0.999m, 1m is 1.000 to 1.999m etc. Such that we can see the difference in volume in the thickness bands between April and September.



So during much of the period concerned the growth/thickness process could not be at work because most volume gain was coming from ice thicker than 2m. Which would be more subject to weather variations affecting temperature and hence heat flux through the ice. Looking back at the Thorndike graphic this thicker ice doesn't exhibit such strong thermodynamic growth.

After 2007 it is seen that volume growth in ice less than 2m thick increases substantially, it is this that is subject to the rapid thermodynamic growth outlined in Thorndike's graphic.

So the role of weather variability in previous years has declined as seen in the above scatter plot where pre-2007 years splatter above and below the trend, but subsequent years are closer to the trend line. And this has happened precisely because thermodynamic growth is becoming a stronger factor.

Peter Ellis

  • Grease ice
  • Posts: 619
    • View Profile
  • Liked: 33
  • Likes Given: 14
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #190 on: August 27, 2013, 10:55:14 PM »
I am not sure you can say that without knowing a lot more about the internals of the ice model.  On your own blog, you confirm that you're using the "effective thickness" of the ice - however we have no understanding of what that means.  What is the "effective thickness" of a pixel that has 50% MYI at 4m and 50% FYI at 1m?  If the FYI thickens up to 2m, how does that change the "effective thickness"? When FYI forms ridges, how does that alter the "effective thickness"?

Moreover, while we know that first year ice _in reality_ doesn't get thicker than ~2m (well, closer to 2.2m if you look at the current set of buoys), we don't know whether that's also true _within the model_.

Simply put, I do not believe you are safe to conclude that <2m "effective thickness" indicates FYI and that >2m indicates MYI.

ChrisReynolds

  • Nilas ice
  • Posts: 1714
    • View Profile
    • Dosbat
  • Liked: 2
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #191 on: August 28, 2013, 07:46:53 AM »
Peter,

Effective thickness is the thickness to which the sub grid thickness distribution is applied. We do have understanding of what effective thickness means (it is the thickness to which the sub-grid thickness distribution is applied), and how it can be used: Effective thickness is what the PIOMAS team provide as their thickness maps. And is used along with concentration to produce their thickness distribution, also to calculate their volume data.

The plots of effective thickness for Sept 2012 and for April 2013 show that the large region of open water translated into ice of around 2m thick. Showing that the model complies with the typical ~2m thickness for a season's thermodynamic growth.

Furthermore the comparisons between ASCAT, QuikScat, and PIOMAS support the interpretation that early winter (Dec/Jan) 2m thickness represents ice that contains a large MYI fraction, and ice thinner than that represents FYI in the process of thickening. The agreement between MYI from the satellite data and is over 2m thick from PIOMAS is too good to be by chance.

Quote
What is the "effective thickness" of a pixel that has 50% MYI at 4m and 50% FYI at 1m?  If the FYI thickens up to 2m, how does that change the "effective thickness"? When FYI forms ridges, how does that alter the "effective thickness"?

1) The model would not present the sort of 4m/1m situation you outline, the model would have an effective thickness of a certain value, to which it would apply the thickness distribution. The thickness distribution is a bell shaped curve therefore it could not represent your example. The model does not model FYI and MYI separately and explicitly.

2) Thickening FYI? The bell curve thickness distribution is applied to an effective thickness, as the ice thickens the curve of thickness distribution rises along with the effective thickness. Like wise with ridging, a set of equations handle mechanical deformation.

EDIT - Buoys, these are concentrated amongst the MYI in the centre of the pack. They tell nothing about conditions across the Siberian/Eurasian half of the pack. 2.2m is near enough to 2m to be described as a 2m thickness. The 2m thickening due to thermodynamics is a nominal figure.
« Last Edit: August 28, 2013, 07:53:48 AM by ChrisReynolds »

helorime

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 3
  • Likes Given: 17
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #192 on: August 28, 2013, 08:07:58 AM »
What is the subgrid thickness distribution?
Any sufficiently advanced technology is indistinguishable from magic.

Richard Rathbone

  • Nilas ice
  • Posts: 1169
    • View Profile
  • Liked: 249
  • Likes Given: 14
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #193 on: August 28, 2013, 08:37:27 AM »
What is the subgrid thickness distribution?

For some operations, it is only necessary to know what the average thickness is in a grid cell, and it doesn't matter whether that thickness is uniform or not. An example is calculating the volume.

For some operations it makes a big difference whether the ice is of uniform thickness, or has some bits thinner than others. An example is calculating the volume gained during a timestep.

For the operations where the calculation is sensitive to thickness within a cell, a distribution is used. Ice might be half 3m MYI and half 1m FYI, or it might be a mean 2m with standard deviation 0.5m or it might be 80% at 2.5m and 20% at 0m. These are all different ways in which the thickness might be modelled at a finer level than the grid cell, and hence subgrid thickness distributions.

Arctic sea ice models generally use subgrid thickness distributions in some form or other because the rate of volume gain is highly non-linear with thickness. Computationally it pays to use a more complicated description in a larger cell than a single thickness variable at a much finer grid scale.

helorime

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 3
  • Likes Given: 17
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #194 on: August 28, 2013, 08:50:26 AM »
Thank you.  I take it that cell size depends on the particular analysis and that the resolution of the ice structure in an individual cell is always substantially higher than the cell size?  So by "subgrid" what is actually meant is complex categorization of cell types based on their varied ice thickness and structures within the grid?

Are model predictions robust enough to differentiate (for example) between a cell that has 3 meter ice on one side and slants to 1.5 meters on the other vs ones that have 3 meter thick ribbons separated by 1.5 meter thick runnels  throughout a cell?
« Last Edit: August 28, 2013, 08:58:45 AM by helorime »
Any sufficiently advanced technology is indistinguishable from magic.

jdallen

  • Young ice
  • Posts: 3291
    • View Profile
  • Liked: 574
  • Likes Given: 216
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #195 on: August 28, 2013, 09:00:01 AM »
Thank you.  I take it that cell size depends on the particular analysis and that the resolution of the ice structure in an individual cell is always substantially higher than the cell size?  So by "subgrid" what is actually meant is complex categorization of cell types based on their varied ice thickness and structures within the grid?

Are models fine enough to differentiate between a cell that has 3 meter ice on one side and slants to 1.5 meters on the other vs ones that have 3 meter thick ribbons separated by 1.5 meter thick runnels  throughout a cell?

I kind of doubt it... unless there's a satellite out there able to resolve the state of the ice better than the 250M resolution we see on Lance Modis.

That's actually one of the questions I'd raise considering the current state of large areas of ice - we have huge areas covered by floes which are under the minimum resolution of our measuring systems; the granularity has become to fine for us to really see the actual detail over large areas which are no longer dominated by a "mesh" pack.

I think the current Bremen image speaks to that pretty eloquently - the pack is currently split in two, maybe three major uneven pieces, separated by a jumble of disconnected ice cubes of varying size.  And looking at some of the images, I'm not so sure the two northern sections of pack ice are all that robust either.

I think the tools we have may no longer be able to capture key small-scale events which may be having a profound effect on the system's mechanics.
This space for Rent.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #196 on: August 28, 2013, 09:11:49 AM »
"Effective thickness" times area (on the grid) gives you "PIOMAS volume". So for all practical purposes it is the average thickness of the grid cells.

50% 4m thickness and 50% 1m, gives an effective thickness of 2.5m, if the percentages are percentage by area. 

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #197 on: August 28, 2013, 10:39:59 AM »
Morning all,

Concerning JDAllen’s remark “I'm not so sure the two northern sections of pack ice are all that robust either.”, I’ve taken this 100x100km2 detail from MODIS r05c04 this morning:



It is located at app. 75N 170E over what UniBremen represents this morning as compact 95+ concentrated sea ice, like in the “mesh pack”.
The true nature of this ice is better cleared on MODIS in an adjacent stretch 100 km to the South:



There is no reason to doubt that under a persistent layer of low clouds and fog the state of the ice is any different in the First detail.
This is all structureless swaths of floes, with the largest 5-10 km wide within endless rubble.

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #198 on: August 28, 2013, 10:54:19 AM »
And in the light of what Chris Reynolds posted above:

“Effective thickness is what the PIOMAS team provide as their thickness maps. And is used along with concentration to produce their thickness distribution, also to calculate their volume data.”

I find the detailed MODIS maps compared to UB concentration very alarming. How would you effectively calculate thickness out of concentration when you’re actually confronted with a slush of debris?

Phil.

  • Frazil ice
  • Posts: 382
    • View Profile
  • Liked: 49
  • Likes Given: 1
Re: NSIDC 2013 Arctic SIE September minimum: August poll
« Reply #199 on: August 28, 2013, 02:22:10 PM »
In short, the average thickness of FYI in February 2013 was greater than in February 2013.

I assume that this should be: 'the average thickness of FYI in February 2013 was greater than in February 2012'?