Cryosphere Today 2013 Arctic SIA maximum

[crandles]

Wouldn't water be at the freezing point temperature of about -1.5C or is it a lot fresher?
I would have thought brine rejection would make it saltier rather than fresher even if salty water will tend to sink away from the ice/water boundary.

If -1.5C, the calculation would then be
(-29 - -1.5) / (-25 - -1.5) = 1.17
&
1/1.17 = .854
not enough difference to really worry about.

[Ice Cool Kim]

Apologies Chris.  I had not read carefully enough what you wrote about flux ( I have a lot of difficulty with white on black web sites).

It was your initial phrase that was wrong:

NCEP/NCAR shows average absolute temperature was around -25degC for February to April 2010, this being some 4 degrees warmer than average, that's a factor of 1.16

The change in absolute temp is 1.016, not 1.16 , that led me to think you'd slipped up.  You seem to have meant that the temp diff changes by a factor of 1.16.

heat flux through the ice is inversely proportional to the difference in temperature

I think you meant directly proportional to , not inverse.

[crandles]

The 'absolute' is as opposed to a temperature anomaly. Not absolute as in kelvin scale.

If the ice is thinner more heat can flow though the ice. So it is inversely proportional.

[Ice Cool Kim]

That's called "temperature" not absolute temperature. Not the first time I've seen that misused. You are right though, that probably was what he meant. I just read what was written.

 Inverse...?  He said difference in temperature not thickness !

[Neven]

That's called "temperature" not absolute temperature. Not the first time I've seen that misused. You are right though, that probably was what he meant. I just read what was written.

 Inverse...?  He said difference in temperature not thickness !

Lucia, is that you? 

[Ice Cool Kim]

Lucia, is that you?

No why? Does someone called Lucia like things to be called correctly as well? It does help in science. Nice to know I'm not alone

It's only bloggin' but it does help to use correct terms and know your inverse/direct stuff and not confuse thickness and temp diff. etc.

I wouldn't have commented originally but it just happened that the absolute temp change was 1.016 and Chris stated 1.16 . It looked like a slip that would have mislead the rest of what he was doing.

[gfwellman]

I.C. Kim - you're right about the original blog post you referred to.  There were two distinct and significant misstatements, but one didn't affect the math presented, and the other only turned out to affect the math in the small way that crandles worked out.

I can't remember where, but I remember a big laugh being had about a year ago about some denialist type taking ratios of temperatures ... in Fahrenheit.  Neven's joke about Lucia ... I'll let him explain that one, but the conversation you were having with crandles was taking on a "Who's on First" flavor, and that is reminiscent of things at Lucia's Blackboard.  I laughed, but it was a little unfair because you were only trying to increase clarity.

[ChrisReynolds]

I've corrected the inversely to directly. That was the only error.

The use of the term absolute temperature was not in error. It was necessary to show I was talking not about anomaly but about actual temperature, and as the term absolute was followed by degC and at no point was K referred to, the meaning was clear.

[Frivolousz21]

I see the Navy ice thickness forecast for the next week (run 3/22) suggests thinning of the pack -- less black, and the red is changing to yellow, even north of 80N.

Just a heads up, I wouldn't trust those maps in any way, not even VS each other.

[DungeonMaster]

I think Crandles' approach of using periods of change in volume/area is more useful in giving answers to the question: "What if year X happened from now?" Although it's still a poor cousin to having a model like PIOMAS and running forward, because it still doesn't fully account for initial conditions.

Uh, now I realise what I was asking - pretty naive, I stop harassing you with this 

Can't PIOMAS teams do this - running the model in forward mode ? I guess that they can but won't publish anything, fearing a reputational damage if they're wrong... We'll probably have to wait for SEARCH to see some estimates.

[Ice Cool Kim]

PIOMAS is not a prediction tool. It is an attempt to model ice in the context of input of climatic data. It is getting interesting but don't forget that it requires SST and atmospheric data as inputs to do what it does.


@Chris.
Absolute temperature has a precise meaning in science. It is better not to re-use it to mean something else. "Actual temperature" may be better if your feel that someone is going think you are referring to temperature anomaly when you say temperature.

Also, whether or not you add "absolute", you said temp not temp difference which it turns out was what you meant. I could have guessed that from the rest but it was not what you wrote.

I already said I had not read the rest carefully enough. White on black does funny things with my eyes and I did not want to struggle through reading the whole thing as carefully as I should. When I saw absolute temperate (rather than temp difference) I took it to mean what it said , I thought you were going to consider some radiative effects where that would have been relevant.

It was only the coincidence of 1.016 change in abs temp being the same as 1.16 temp diff in this case that lead me to comment. I was trying to be helpful, not pedantic.

Nothing for anyone to get upset about.

[ChrisReynolds]

Ice Cool Kim,

Sorry, but you seem to be mistaking brevity for annoyance.

When I post in the morning I have time for one or two posts and they tend to be brief, notelike and to the point. Thanks for picking up the ice heat flux error. It's an unfortunate coincidence that in degC the calculation yielded 0.16, where as the same calculation in Kelvin yielded 0.016.

Truly I wasn't annoyed, in fact I like people to pick up any errors I make, and I like people to challenge my assumptions. Without this happening there's no telling how much time I'd waste on dead end pursuits.

[Ice Cool Kim]

Thanks, I'm glad it was helpful and taken in the right spirit. So many people are so touchy with anything climate related I'm always half expecting people to get shirty.

I was starting to get that ' here we go again feeling". Glad I was mistaken and thanks for the post.

I always learn more from those who take a critical look at what I've done than those who applaud.

 

[crandles]

PIOMAS is not a prediction tool. It is an attempt to model ice in the context of input of climatic data. It is getting interesting but don't forget that it requires SST and atmospheric data as inputs to do what it does.

I would like to know more about the direction and size of the adjustments made as a result of assimilating data.

To run in a forward prediction manner you may well need to extrapolate how the general size and direction of adjustments would change. I have no idea how predictable trends in size, direction and distribution of such adjustments would be. I would guess that is harder and more prone to error than extrapolating the sea ice volume trend so not a lot of point.

[Ice Cool Kim]

I would like to know more about the direction and size of the adjustments made as a result of assimilating data.

I could be wrong because I don't have access to source code but my understanding is that PIOMAS takes climatic data inputs (primarily form other models that do attempt predication) and runs a physical thermodynamic model to calculate how water freezes and how this propagates. 

I don't think it's a case of applying adjustments. They take a known measured state of the ice as a initial boundary condition and integrate a model over time.

this is a reciprocal relationship with the global models since the formation of ice will affect how the GCM predictions evolve.

[lanevn]

What's with that predictions http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticice/nowcast/ice2013032618_2013040300_035_arcticice.001.gif ? Are they really expect super late max?

[crandles]

I would like to know more about the direction and size of the adjustments made as a result of assimilating data.

I could be wrong because I don't have access to source code but my understanding is that PIOMAS takes climatic data inputs (primarily form other models that do attempt predication) and runs a physical thermodynamic model to calculate how water freezes and how this propagates. 

I don't think it's a case of applying adjustments. They take a known measured state of the ice as a initial boundary condition and integrate a model over time.

http://psc.apl.washington.edu/wordpress/research/projects/arctic-sea-ice-volume-anomaly/

For the ice volume simulations shown here, sea ice concentration information from the NSIDC near-real time product are assimilated into the model to improve ice thickness estimates and SST data from the NCEP/NCAR Reanalysis are assimilated in the ice-free areas.

There aren't enough measurements of ice thickness, even with IceSat2 it takes a month to cover all areas to get an initial state. Therefore thickness has to be modeled and adjusted when data shows the model calculations are going wrong.

So I am taking above quote to mean that if the model has ice in a cell where NSIDC show 0 ice concentration then the thickness of ice in that cell is adjusted to 0 and if the other way around I assume they adjust the thickness to some minimal thickness amount.

Is there a different interpretation which doesn't involve 'adjustments'?

[ChrisReynolds]

Kim, Crandles

PIOMAS doesn't take current thickness adjustments. The submarine DRA (Data Release Area) freeboard/thickness data was split into two subsets, one was used to fine tweak variables for which the physics were uncertain, the other was used for intercomparison. More on validation here:
http://psc.apl.washington.edu/wordpress/research/projects/arctic-sea-ice-volume-anomaly/validation/

Schweiger et al 2011 list the variables assimilated:

Daily mean NCEP/NCAR reanalysis data are used as atmospheric forcing, i.e., 10-m surface winds, 2-m surface air temperature (SAT), specific humidity, precipitation, evaporation, downwelling longwave radiation, sea level pressure, and cloud fraction. Cloud fraction is used to calculate downwelling shortwave radiation following Parkinson and Washington [1979].

PIOMAS uses these forcing data to drive the ocean model and the ice model.

The PIOMAS release history solely concerns the assimilation procedure:
http://psc.apl.washington.edu/wordpress/research/projects/arctic-sea-ice-volume-anomaly/release-history/
The basic physical model therefore does not seem to have changed since Zhang 2003, "Modeling Global Sea Ice with a Thickness and Enthalpy Distribution Model in Generalized Curvilinear Coordinates"
http://psc.apl.washington.edu/zhang/Pubs/POIM.pdf
And, I can't stress this too much, references therein. 

[ChrisReynolds]

Crandles,

So I am taking above quote to mean that if the model has ice in a cell where NSIDC show 0 ice concentration then the thickness of ice in that cell is adjusted to 0 and if the other way around I assume they adjust the thickness to some minimal thickness amount.

That's my understanding.

[ChrisReynolds]

Crandles,

I've just finished reading Winton 2000 "A Reformulated Three-Layer Sea Ice Model"
http://www.gfdl.noaa.gov/bibliography/related_files/mw0001.pdf

You've suggested in relation to some matter that FYI freezing onto the base of MYI and the subsequent easier melt of the FYI might play a role - in the Spring melt IIRC.

That Winton paper details the two layer ice model used in PIOMAS (three if you count snow). The two ice layers are of equal thickness, so as the ice thickens or thins the partition between the two layers shifts. Brine inclusion is treated not as a separate variable but is included implicitly within some other calculations, I'm still trying to get my head around how the brine inclusion is handled because this is a major difference between young and old ice. However from what I've read I can't see aging being a factor that makes ice 'fresher'. So the 2010 and following spring melt seems to be related to thinner ice, not older less briny ice.

So while ice growth on the base is modelled - explicit brine content isn't except in that the upper and lower layers have different temperatures, which impacts saline content parametisation. I'll see if this makes more sense after Breakfast.

The following passage is fascinating:

The treatment of the brine content of the ice is one demarcating feature distinguishing simple and sophisticated models. Since sea ice contains salt, it also contains an amount of liquid water needed to dilute its salt to a brine with freezing temperature equal to the sea
ice temperature. Thus, as the ice is heated, for example, two changes occur: 1) the temperature of the ice and its brine increases, and 2) the higher temperature allows
for a more dilute brine and so some ice melts, increasing the brine content of the ice.

I'd never thought of the issue as being one of temperature dependent equilibria.

Maybe we need a separate thread to discuss PIOMAS?

[ChrisReynolds]

Thanks, I'm glad it was helpful and taken in the right spirit. So many people are so touchy with anything climate related I'm always half expecting people to get shirty.

I was starting to get that ' here we go again feeling". Glad I was mistaken and thanks for the post.

I always learn more from those who take a critical look at what I've done than those who applaud.

 

Kim,

Happy to have cleared that up, I wouldn't want to deter you from posting here. 

[crandles]

Thanks Chris,

Interesting. Yes it would appear PIOMAS does not distinguish layers of different ages. Even if there was an effect in reality to cause the ice volume anomaly pattern, it wouldn't be modelled by PIOMAS nor does it seem possible for assimilation adjustments to reality to cause such an effect to show up in PIOMAS data.

So thanks for the link to the paper and for debunking that possibility.

A separate brine variable is no longer needed because the brine content in the new model is completely determined by the upper ice temperature and the (predetermined) ice salinity.

So this paper doesn't help us understand the '(predetermined) ice salinity'. However, if the upper layer ice gets warmed then more brine forms and presumably at some level the brine starts to 'leak' out of the ice. The more frequently and the bigger the extent by which the ice is warmed past the point where brine begins to leak, the fresher the ice gets. So ice doesn't get fresher by age specifically but by number of times and extent by which ice is warmed past point where brine leaks. Of course the number of times it is so warmed will be related to age and season.

Adding FYI to MYI will add to the ice salinity but the salt all goes into one pot rather than keeping different ages of ice separate.

That is my interpretation/guesses which could easily be wrong.

[TerryM]

Would this also indicate that a particularly warm summer might freshen (age) the surviving ice more rapidly than a relatively cool summer?


What I'd noticed, particularly last year, was the extension of the summer warm period north of 80. Temperatures over all remained at the freezing point, but the number of days that the Arctic stayed at this temperature increased. Would this extended period of warmth also act to prematurely age the remaining ice?


If either or both of these suppositions are true, could this act as a negative feedback, possible helping to explain the "rebound" after the large drop off in 2007.


Terry

[Ice Cool Kim]

I think surface reflection (melt pools and open water) needs to be accounted for and I have not been able to find any info on that.

[mhampton]

I think surface reflection (melt pools and open water) needs to be accounted for and I have not been able to find any info on that.

The connections between age of ice, salinity, and melt pools seem to be very complex from what I can tell (I am a mathematician but not an ice or climate expert).  I guess some people like Ken Golden are working on modelling those interactions but feedback with percolation parameters seems like a nasty problem to me.

[ChrisReynolds]

Crandles,

The salinity of the ice is stated in table 1 to be 1 per mil, thus brine leakage and freshening as ice ages simply doesn't seem to occur in PIOMAS, as the Zhang 2003 paper cites Winton 2000 with regards the 2 layer (+1 layer snow) sea ice component.

This means that the post 2010 spring melt must be related solely to thinning, not to impacts on age or implied physical characteristics. There are three papers, the maths is borderline beyond me - it's over 20 years since I dealt with a differential equation. Otherwise the best way would be to use the papers to make a column model of 1 cell, and apply warming to various different initial thicknesses.

The thing is we know brine rejection with aging strengthens and freshens ice in the real world. Does this suggest that this isn't an important process in the overall picture, given that PIOMAS seems to be doing quite well? I suspect an implicit inclusion of this in the treatment of ice thickness, but that's a really thorny batch of interlinked equations.

Terry M,

I'm not sure about that. There is no accounting for ice ageing and freshening in PIOMAS, but treatment of thickness may have some implicit mechanical effects. The Winton paper I've just read really only deals with thermodynamics. I didn't think temperature affected ice salinity, I thought it was mechanical deformation and age that allow the brine to drain.

Ice Cool Kim,

In Hibler 1980, Modelling a Variable Thickness Sea Ice Cover, albedo is in two states: ice or snow. I don't recall reading about further complexity being added in the other papers PIOMAS is based on. This might explain why after the Spring Melt the PIOMAS anomalies indicate less ice loss than climatology, which in turn might imply more volume loss during the late summer than PIOMAS shows.


However as with that Hibler paper, a lot of the older papers are image PDFs so I can't search the pdf easily for key words. And my memory isn't excellent.

[crandles]

The salinity of the ice is stated in table 1 to be 1 per mil, thus brine leakage and freshening as ice ages simply doesn't seem to occur in PIOMAS, as the Zhang 2003 paper cites Winton 2000 with regards the 2 layer (+1 layer snow) sea ice component.

I would guess you are more likely to be right than I am.

Nevertheless, it seems to me that table 1 was provided for notation. Salinity of ice is predetermined per earlier quote; if it was fixed wouldn't it say fixed so doesn't predetermined mean variable but established earlier in the calculations?

Anyway from what little I have seen, it still seems possible that the values given in table 1 are simply the values used in the comparison:

To compare the new model with Semtner’s model
both were forced with climatological seasonal fluxes of
downward short- and longwave radiation, and sensible
and latent heat fluxes. The values used were the same
used by Semtner to compare his model with the Maykut
and Untersteiner model.

You may well have seen something more clear cut in another paper and so you are more likely to be right even if I haven't misread this paper.

Edit: also

Constant relating freezing temperature to salinity 0.0548C per mil

Why would this be needed if the salinity was fixed? Surely you would just have a fixed freezing temp?

[ChrisReynolds]

Winton states that:

A separate brine variable is no longer needed because the brine content in the new model is completely determined by the upper ice temperature and the (predetermined) ice salinity.

My emboldening.

Yet...

I read the values in Table 1 as constants used in the model. Earlier a variable/constant mu is defined as "a constant equal to minus the freezing temperature (in degC) divided by the salinity" But this is always used in a term mu*S, where S is salinity. Now it doesn't seem to me to make sense if both instances of salinity are equal, because that would imply mu = 'the freezing temperature'. Although where net energy fluxes would seek to increase temperature above the freezing point of the ice the constant mu makes another appearance as the temperature is set to -mu*S, this is stated to be equal to the freezing point of the ice. In other words this does imply that mu = freezing temperature.

I'm going to drop the matter today and sleep on it.

[crandles]

I read the values in Table 1 as constants used in the model. Earlier a variable/constant mu is defined as "a constant equal to minus the freezing temperature (in degC) divided by the salinity" But this is always used in a term mu*S, where S is salinity. Now it doesn't seem to me to make sense if both instances of salinity are equal, because that would imply mu = 'the freezing temperature'. Although where net energy fluxes would seek to increase temperature above the freezing point of the ice the constant mu makes another appearance as the temperature is set to -mu*S, this is stated to be equal to the freezing point of the ice. In other words this does imply that mu = freezing temperature.

I'm going to drop the matter today and sleep on it.

The definition of mu seems to me to be saying that (-mu * S) = freezing point.
(not that mu = freezing temperature.)

So I am having difficulty following what you are trying to say and why you would think mu=freezing point.

My edit above was trying to say that if salinity was constant then you would just have a constant for the freezing temperature rather than have this definition of mu=-freezingtemp/S

[ChrisReynolds]

mu = -Tf/S

Where S = salinity, and Tf = Freezing point.

Therefore mu*S = -Tf*S/S = -Tf.

[crandles]

Therefore mu*S = -Tf*S/S = -Tf.

Agreed that is same as my  (-mu * S) = freezing point.

Now it doesn't seem to me to make sense if both instances of salinity are equal, because that would imply mu = 'the freezing temperature'.

both instances of salinity??
(Also don't understand what you think doesn't make sense.)

I assume reference to the previous model's variable for brine content is a variable for the volume of liquid brine contained in the ice (or maybe the amount of salt in liquid water). This isn't needed in this model because you can work it out from temperature of upper layer and the salinity. If salinity was fixed then this would say you can work out the brine content from the temperature of the upper layer.

[ChrisReynolds]

What I'm saying is that I don't understand why mu*S is needed at all, because it's always going to be equal to Tf*1, which is what was making me wonder if the salinity (S) used to make mu is the same as the salinity used in mu*S.

Why have mu*S when mu*S = -Tf*S/S = -Tf. Why not just dispense with mu*S and use -Tf (minus freezing point, or as is stated in the paper minus freezing temperature of ice).

The answer is that mu*S does not equal -Tf*S/S

I am now increasingly convinced that this is a result of ambiguity in Winton's wording. Mu does not equal -Tf/S, because S is salinity (which varies) and in the paper mu*S is stated as a constant. CICE uses Winton as a reference, and the documentation for that ice model, as in Winton table 1, it states mu as a constant of proportionality: 0.054degC/per mil, which relates salinity to the freezing point of the ice at salinity S. Bitz 1999 seems to support this, although the only copy I can find of that is a post script file, so equation formats are corrupted, However Bitz attributes this figure to a 1968 paper.

Anyway what is going on is a scaling of melt temperature dependent on salinity, which was what I was expecting.

I think others are looking at CICE in the programming thread, I'll go and have a look.

[wanderer]

Ice Thickness Comparison:

vs.

[Jim]

Hi Wanderer,

Could you give us a bit of context, regarding these two images?
Clearly, they are different, although they are for the same day (29th March 2013).
The only difference on the image writing is:

1st image
"Ice Thickness:"
"0 to 7.6"

2nd image
"Ice Thickness (m):"
"0 to 7.4"

What does "(m)" mean, and why only on one of the images?
What does "0 to 7.6/7.4" mean? What is being measured here, and what is the significance of this?

Yours inquiringly,
Jim

[Ice Cool Kim]

A bit context about what the data source is and whether it's modelled or measured / a mix of both?

Not much point in banging up a picture without saying what it is.
 

[ChrisReynolds]

Sounds interesting, I wish the images were still up.

[crandles]

Can still look at the links:

Code: [Select]

http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticict/nowcast/ict2012032818_2012032900_035_arcticict.001.gif
that is 2012 not 2013 and the other is current.

[ChrisReynolds]

Thanks Crandles,

They're up now, must have been a blip at HYCOM.



What do I see from those images?

Well the FYI is thicker this year than for the same data last year. But there is less MYI (thickest region bright colours).


Jim,

I suspect those thicknesses are the peak thickness, it would tally with thickest ice in grid boxes for PIOMAS.

[Ice Cool Kim]

Thanks Chris,

here's some explanation of the project. Seems like "real-time" model output :?
http://www7320.nrlssc.navy.mil/hycomARC/prologue.html

[johnm33]

I happened to be looking through the archive yesterday
http://www7320.nrlssc.navy.mil/hycomARC/navo/arc_list_arcticictn.html
and found that whilst if you looked 20 odd days ahead you could get an arguable match for the extent/thicknesses you had to go to about the 5th of august for the thick ice to match [ish]

[wanderer]

Sorry about that - it should have been a comparison between March 29 2012 and March 29 2013.
Maybe it is easier if you check it out yourself:
http://www7320.nrlssc.navy.mil/hycomARC/arctic.html

Like Chris Reynolds said - it seems that the FYI is thicker, but there seems to be much less MYI.

How much will this have impact over the summer melt season?

[Jim]

Ah, I see Wanderer - This year vs last year (sorry, missed that) 
So, there is a significant drop in thick ice this year and it looks to be more 'mixed' and distributed as well.
I do notice that the ice in the Kara sea and north Barents sea is thicker this year against last years thinner ice, but I would think this will still melt out pretty quickly. A good storm in mid-summer will certainly stir-up the CAB and could well do in a large chunk of the old ice, now that it is more spread out. As I've said before, this summer melt will be extremely interesting!