Freezing Season Basics?

[JimboOmega]

So I'm new to watching the ice during the freezing season...

Can anyone explain the basics? Like what areas freeze first, how the weather influences freezing? What sort of set up from the melting season has impact on the freezing season? What should we expect this year different from previous years?  Is there "freezing momentum"? 

All of those kind of things, for a newbie like me!

[crandles]

So I'm new to watching the ice during the freezing season...

Can anyone explain the basics? Like what areas freeze first

I am no expert but some quick non expert answers:

Spreads outwards from existing ice.

how the weather influences freezing?

Increased temperatures reduce the maximum thickness that FYI grows to thermo-dynamically.

Thin ice thickens much faster than thick ice. So Oct Nov doesn't matter too much as any deficit or surplus is quickly reduced by normal temps in Jan - Mar.

What sort of set up from the melting season has impact on the freezing season?

Ice formed is pretty much determined by quantity of heat lost from ocean. Thus cloudy weather slows down heat loss and reduces amount of ice while clear sky increases heat loss and ice formed.

More heat transfer from lower latitudes is also relevant and that can depend on whether polar vortex is displaced or stays central and winds zonal.

What should we expect this year different from previous years?  Is there "freezing momentum"? 

More difficult questions for me. Over to someone more expert.

[seaicesailor]

So I'm new to watching the ice during the freezing season...

Can anyone explain the basics? Like what areas freeze first, how the weather influences freezing? What sort of set up from the melting season has impact on the freezing season? What should we expect this year different from previous years?  Is there "freezing momentum"? 

All of those kind of things, for a newbie like me!

If melting momentum is understood as the initial build up of conditions by melting  (open water, melt ponds) that are going to reduce albedo early enough so that heat absorption -> melting -> heat absorption feedback be amplified during the season ..... then there is no freezing momentum. There is no mechanism that enables a heat release -> freezing -> heat release amplifying feedback

[RoxTheGeologist]

In the same way that dark bodies absorb radiation, they also releases it more efficiently. In terms of heat loss, lots of open ocean being mixed will lose far more heat than ocean covered with a thin white layer of ice.

[Cate]

Jimbo, thank you for this thread. I'm new to freezing season too and appreciate the chance to pick up some info.

As for what freezes first. I've been noticing on Explorer that some of the fjords on Ellesmere and northern Greenland are freezing over. No idea how thick that ice would be at this point but it shows up white---could be just a skim with snow on top? I wonder whether and how this early freezing in the fjords figures into the extent/area calculations of the minimum?

[seaicesailor]

In the same way that dark bodies absorb radiation, they also releases it more efficiently. In terms of heat loss, lots of open ocean being mixed will lose far more heat than ocean covered with a thin white layer of ice.

Yes, but that is early attenuated heat release rather than a self-amplifying nechanism to make freezing much faster. That is not what the "momentum" word used by Neven conveys to me. It would be more like "freezing deceleration"
In any case it would be a negative feedback mechanism to slow down the refreezing.
But I am still not convinced by the argument "late refreezing beneficial to ice recovery". Would like to elaborate on that with more time.

[icy voyeur]

Salt water is different from fresh in that the density of salt water is driven by temperature in a linear manner. In fresh water, 4C is the most dense, so as fresh water cools below 4C it floats on denser water.

In the oceans, as water cools at the surface it sinks (and then exchanges heat and warms). This produces some mixing of the top layer of water. Salinity also plays a role complicating things. When salt water freezes, there's often some salt trapped in the ice but if temps are near the freezing point the salt causes microscale local melting and the salt weeps out as very salty cold water which then is very dense and sinks while of course becoming diluted as things mix.

Now if cooling was  very gradual, cold water would keep sinking and all that latent heat would mix and forestall freezing. But with rapid cooling, ice forms on a surface layer and floats. It insulates the water below it from as rapid a heat loss and the ice continues to form by growing bottom down. There's still a sort of turbulence cause by cooling the top layer of water and salt exclusion from forming ice but it potentially more sedate than that which occurs with open water. And of course the existence of surface ice dampens wave action which also contributes to mixing of the top ocean layers. I believe the net effect is that an ice cap slows the loss of heat.

It might be that these dynamics will help the arctic to exhaust more heat as the arctic becomes more ice free. That warmer water will delay the ice cap formation and increase the rate at which heat is lost compared with seasons where there was less heat to begin with so the ice cap formed earlier. But things might also come down to the irregularities of weather and how soon a significant cold snap manages to initiate an ice cap.

How things shake out practically is, I think, largely unknown. And there are other confounding factors. Worst is perhaps that less ice means more waves and those waves abusing shallow shores could radically increase methane releases with significant global consequences. That's a bit beyond your question however.

[Andreas T]

In the same way that dark bodies absorb radiation, they also releases it more efficiently. In terms of heat loss, lots of open ocean being mixed will lose far more heat than ocean covered with a thin white layer of ice.

there is a widespread misunderstanding which generalizes the observation that absorptivity of a surface at a given wavelength equals the emissivity of that wavelength.
Since you are talking about the absorption of short wave radiation from the sun and the much longer wavelength of emitted IR at low temperatures the behaviour especially of ice and snow surfaces is very different. Fine snow has an emissivity at 10micrometer which similar to that of soot, but at visible wavelengths the absorptivity is obviously much much lower. To make snow emit at visible wavelengths one would have to heat it to temperatures where it would clearly no longer be snow, but it should be a poor emitter of these wavelengths.
source https://cimss.ssec.wisc.edu/itwg/itsc/itsc15/presentations/session2/2_2_knuteson.pdf

[RoxTheGeologist]

Yes, but that is early attenuated heat release rather than a self-amplifying nechanism to make freezing much faster. That is not what the "momentum" word used by Neven conveys to me. It would be more like "freezing deceleration"
In any case it would be a negative feedback mechanism to slow down the refreezing.
But I am still not convinced by the argument "late refreezing beneficial to ice recovery". Would like to elaborate on that with more time.

I think that anything that promotes increased heat loss is beneficial to countering the affects of global warming. I would think the 'worst' scenario is a ocean covered with a nice thin layer of white ice and with insulating clouds and no precipitation. Water freezing in clouds releases the heat of fusion, a dark ocean with clear skies radiates more heat.

I'd theorize that a feedback mechanism may be more storms dumping snow, increasing heat loss with freezing water in the atmosphere, increasing albedo and reinforcing the halocline with fresh water ice.

this is a great resource for the properties of water and ice:
http://linkingweatherandclimate.com/ocean/waterdensity.php

[epiphyte]

Yes, but that is early attenuated heat release rather than a self-amplifying nechanism to make freezing much faster. That is not what the "momentum" word used by Neven conveys to me. It would be more like "freezing deceleration"
In any case it would be a negative feedback mechanism to slow down the refreezing.
But I am still not convinced by the argument "late refreezing beneficial to ice recovery". Would like to elaborate on that with more time.

I think that anything that promotes increased heat loss is beneficial to countering the affects of global warming. I would think the 'worst' scenario is a ocean covered with a nice thin layer of white ice and with insulating clouds and no precipitation. Water freezing in clouds releases the heat of fusion, a dark ocean with clear skies radiates more heat.

I'd theorize that a feedback mechanism may be more storms dumping snow, increasing heat loss with freezing water in the atmosphere, increasing albedo and reinforcing the halocline with fresh water ice.

this is a great resource for the properties of water and ice:
http://linkingweatherandclimate.com/ocean/waterdensity.php

This year we're starting from somewhere new...

   - Highly granular ice = very high area: ice/water interface ratio.
   - Much open water = rapid ocean heat loss to atmosphere -> surface freezing...

   but...

   - rapid heat loss -> convection -> mixing from warm southern latitudes...
                                                 ->  wind -> waves -> mixing from warmer depths?
                                                  -> imported moisture -> snow -> insulating layer over new ice?

... whatever happens, it will be as new to the models as it is to us amateur observers.


   

[sofouuk]

the 'late refreezing beneficial to ice recovery' theory is based on the idea that most snow falls in early autumn; if the areas where it falls are already covered with new, thin ice, the snow forms an insulating layer with slows heat loss from the ocean through the ice, and hence the ice thickens more slowly; vice versa if the autumn snow falls into the ocean. crandles posted a graph n brief explanation of this near the start of last year's freezing thread I think. it's not hard to critique the idea (as crandles just indicated here first year ice always tends to grow to about 2 m thick by the end of winter, for example) but the idea that big melt years will tend to be followed by 'recovery' years is in the literature, due to the snow into the ocean effect

[seaicesailor]

What happened in 2013? Ice volume in spring was as low as in spring 2012.
By looking at sea temperatures in fall 2012 no wonder it took so long for refreezing in some areas, it had to lose the heat excess but that does not imply that it ended up with lower heat available for the next season. It is clear by now what caused the 2013 recovery.
Was 2008 season such a recovery, or are we talking about an "old-normal" year that simply did not match the exceptional 2007 weather during the melting season?

[sofouuk]

exactly what happens in specific years will obviously depend on conditions specific to that year - the theory doesn't say big melt/late refreeze years will invariably be followed by a recovery, only that a recovery is more likely than another big melt

http://forum.arctic-sea-ice.net/index.php/topic,1377.0.html

crandles' comments start towards the bottom of the first page

[Tealight]

This years freezing season could be similar to 2012, where ice advanced from the ice edge and the coast. Neven called it Bilateral freezing and made a blog post about it:
http://neven1.typepad.com/blog/2012/11/bilateral-freezing.html

[seaicesailor]

exactly what happens in specific years will obviously depend on conditions specific to that year - the theory doesn't say big melt/late refreeze years will invariably be followed by a recovery, only that a recovery is more likely than another big melt

http://forum.arctic-sea-ice.net/index.php/topic,1377.0.html

crandles' comments start towards the bottom of the first page

If you mean this post
http://forum.arctic-sea-ice.net/index.php/topic,1377.msg62121.html#msg62121
I found it really compelling, but keep in mind it relies on a model of a regional sea.
I found then this other comment very convincing on how difficult it is to assess the effect of snow
http://forum.arctic-sea-ice.net/index.php/topic,1377.msg63295.html#msg63295

[epiphyte]

This years freezing season could be similar to 2012, where ice advanced from the ice edge and the coast. Neven called it Bilateral freezing and made a blog post about it:
http://neven1.typepad.com/blog/2012/11/bilateral-freezing.html

...but where is the edge? This year it's bordering on fractal...

[Iceismylife]

Fun to watch.

[SCYetti]

You're right that is fun to watch. This happens in shallow placid water in the Antarctic when it has ice forming on the surface. In most of the ocean the water is much deeper and more turbulent so  brinicles don't form. But doesn't the brine still absorb heat from the surrounding water? So does the water remain -1.8C but with a lower enthalpy?

[Iceismylife]

That brine can have a salinity of 27%

https://nsidc.org/cryosphere/seaice/characteristics/brine_salinity.html

Fresh water freezes at 0 degrees Celsius (32 degrees Fahrenheit), but the freezing point of sea water varies. For every 5 ppt increase in salinity, the freezing point decreases by 0.28 degrees Celsius (0.5 degrees Fahrenheit); thus, in polar regions with an ocean salinity of 35 ppt, the water begins to freeze at -1.8 degrees Celsius (28.8 degrees Fahrenheit).

It may well freeze some of the surrounding water with the ice floating to the surface.  But the brine should fall some distance before it mixes with enough sea water to equalize in salinity. It would be colder and higher density and fall some distance in the water column.

[seaicesailor]

If anybody wants to discuss this, here or in the freezing season thread: why the refreezing wave traveled so fast outwards within the two CAB wedges? I find it amusing and no idea how could it be.

[binntho]

If anybody wants to discuss this, here or in the freezing season thread: why the refreezing wave traveled so fast outwards within the two CAB wedges? I find it amusing and no idea how could it be.

I agree, I find it very surprising - but then again this is my first "start-of-refreeze" experience. On Nullschool, the last 24 hours show fairly slow winds, but for the most of the area where refreeze is happening (judging from Uni-Bremen maps(, temperatures are well above -10.

[Peter Ellis]

The -10 figure has passed into "background knowledge" here without real examination.  It is only a rule of thumb, and has no physical basis.  Water freezes at 0 degrees if fresh, or around -1.4 degrees for sea water. If the air above sea water is held steady at -1.4 degree for <i>long enough</i>, then the surface will freeze. That's just thermodynamics.

But of course, the devil is in the detail. The sea has rather a lot (!) of thermal inertia to overcome, so "long enough" might well be months or even years if the air temperature really was held at only just below freezing. In reality, air temperature is never steady - instead it continues to drop throughout the onset of the freeze season.

The figure of -10 degrees is simply a heuristic from Wayne Davidson - he has observed that in his local area, by the time the surface freezes over, temperatures have already dropped to around -10 degrees.  Conditions elsewhere in the Arctic may well be completely different.

(In replies, please can we have no bullshit about supercooling, nucleation of ice crystals and so forth - that only applies to extremely stable laboratory conditions)

[johnm33]

" why the refreezing wave traveled so fast" I'm thinking the simplest explanation is the emergence, of bottom melt fresh water still in contact with it's 'parent' ice, thus already at 0c and primed to freeze. As more emerges and freezes it dampens mixing and a larger percentage of the emergent freshwater freezes. [?]

[JimboOmega]

The -10 figure has passed into "background knowledge" here without real examination.  It is only a rule of thumb, and has no physical basis.  Water freezes at 0 degrees if fresh, or around -1.4 degrees for sea water. If the air above sea water is held steady at -1.4 degree for <i>long enough</i>, then the surface will freeze. That's just thermodynamics.

But of course, the devil is in the detail. The sea has rather a lot (!) of thermal inertia to overcome, so "long enough" might well be months or even years if the air temperature really was held at only just below freezing. In reality, air temperature is never steady - instead it continues to drop throughout the onset of the freeze season.

The figure of -10 degrees is simply a heuristic from Wayne Davidson - he has observed that in his local area, by the time the surface freezes over, temperatures have already dropped to around -10 degrees.  Conditions elsewhere in the Arctic may well be completely different.

(In replies, please can we have no bullshit about supercooling, nucleation of ice crystals and so forth - that only applies to extremely stable laboratory conditions)

I assume that most of the heat loss comes from the water itself radiating heat to space, rather than losing the heat to the atmosphere convectively/conductively, right?

[Iceismylife]

I assume that most of the heat loss comes from the water itself radiating heat to space, rather than losing the heat to the atmosphere convectively/conductively, right?

I'd put most of the heat loss as evaporation.

[jdallen]

I assume that most of the heat loss comes from the water itself radiating heat to space, rather than losing the heat to the atmosphere convectively/conductively, right?

I'd put most of the heat loss as evaporation.

It's not monolithic, and you actually have multiple layers emitting, but eventually the loss is governed by the rate at which IR is emitted out to space from the top of the atmosphere.

Loss from the water is three fold - IR/black body radiation from the top layers of ocean, loss via transport from evaporation and re-condensation, and conductive via direct transfer to atmosphere.  For sure evaporative loss is huge, but I'm not sure it's the biggest, as it depends on vapor pressure which will be rather low, and that evaporation putting more H2O into the atmosphere actually provides a negative feedback to heat loss (H2O being a fairly efficient GHG).

Once we get a skin of ice, the loss becomes mostly radiative, with a small portion being conductive/convective.  The conductive portion diminishes rapidly as the ice thickens and slows down heat flow, thus permitting the ice/air interface to reach equilibrium at a lower temperature.

The presence of ice, even fairly thin ice also slows down transfer internally in the water column.  Without open water, we reduce turbulence and redistribution of heat from depth.  Without that, the water column thermo and haloclines will stabilize and reduce convective exchange with the surface.

[Andreas T]

As JD says the ice surface gets very cold because it radiates and has low heat transfer towards it from the water below or the atmosphere above, especially when it is covered by a layer of snow. This cold snow surface then cools air passing over it and this cold air can the in turn cool neighboring water surfaces. An additional effect of this is that the chilled air is dry in absolute humidity terms so it will take up water vapour as it warms over the open water. Air which is already humid over a larger water surface is less able to receive much water vapour which limits evaporative cooling.

[RoxTheGeologist]

 
I read that both Ice and Snow have lower IR emissivity than water, so the amount of radiation lost from the ocean surface should be reduced further. I am sure the remote sensing bods on the forum can expound on my thin knowledge!

[Andreas T]

There is data available for this, example  http://www.icess.ucsb.edu/modis/EMIS/html/em.html , can you say where you read this and how different you think emissivity is?

[RoxTheGeologist]

There is data available for this, example  http://www.icess.ucsb.edu/modis/EMIS/html/em.html , can you say where you read this and how different you think emissivity is?

It was an old paper, from 1992! It measured emissivity against angle, and concluded that snow and ice are lower at greater than 65% from the normal, but, also, that the radiation followed what would be expected from the Fresnal equation. I guess that would mean that ice would appear colder if observed from a high angle! I don't have the full paper unfortunately, I was simply traulling google scholar out of interest. For those that don't use google scholar, it's a great way to not appear completely stupid on this forum...)

https://scholar.google.is/

I'm sorry for the .is.. I'm in iceland and my laptop seems to think I want to read everything in Icelandic...

http://www.tandfonline.com/doi/abs/10.1080/01431169208904088