The 2015/2016 freezing season

[OldLeatherneck]

There has been a lot of chatter on several of the other  topics about whether the building El Nino or other factors might extend 2015's melt season, thereby delaying the start of the 2015/2016 re-freeze. Therefore, I decided to open a topic where we can start talking about the many factors that can either inhibit or facilitate the growth of sea ice during the dark months.

What will the impact of El Nino be?
Will the warm blob stay in the North Pacific and what impact might that have?
Will the fractured jet stream direct more or less heat from the continents?

I'm sure there are many commenters on this forum, who are smarter than me (the majority), that will pose better questions andor be able to provide good information as the fall and winter months approach.

[lifeblack]

Here's my speculation on the impact of el nino / warm blob on the refreeze season:

Because of more water vapor in the air from the warm waters, I think we're more likely than other years to have widespread snows across the arctic which would insulate the ice.  Thinner ice in winter this year => large melt season next year?

[OldLeatherneck]

Here's my speculation on the impact of el nino / warm blob on the refreeze season:

Because of more water vapor in the air from the warm waters, I think we're more likely than other years to have widespread snows across the arctic which would insulate the ice.  Thinner ice in winter this year => large melt season next year?

Welcome to the Forum lifeblack.

You are absolutely correct that increased water vapor increases the chance of snow, however, when and where the snow falls is entirely dependent on the weather.  I'd be interested to at what point of the re-freeze season does a heavy blanket of snow on the ice have the greatest impact on further ice growth.

Do we have any standard definition of when the re-freeze season actually begins??  Since our three standard metrics, area, extent and volume all reach their annual minimum at different times and then quite often bounce around the bottom for days or weeks before beginning to climb.  I would surmise that the true start of the re-freeze season is not until all three metrics have climbed by a certain amount (percent gain or absolute amount) or for a number of successive days.

If there is no current standard by which one can declare the start of the re-freeze season, I recommend that the members of this forum should develop that standard.

[seaicesailor]

OLN, thanks for inviting all to the new thread!

Very vague initial impressions for the refreezing season:

The NH is warmer than ever. El Niño started in 2014 give or take 0.1 C. It has already affected this season (my impression).
Since the planet is warmer and El Niño 2015 is very strong, the weather systems will very probably carry more warmth into the Arctic than ever. So the probabilities of having a mild Winter and a hot Summer in the Arctic are very high.
There may or may not have been a tipping point in the past, but the CO2 keeps increasing, and the planet keeps warming. The Arctic has no way out, the Arctic amplification effect is more than evident.

As I said, very vague, very little knowledge of Climate Science.

Thanks :--)

[Rick Aster]

I wonder about the potential for open water and wave action to melt ice in the early stages of freezing, especially if storms wander in from the Pacific.

[andy_t_roo]

Given the surface radiative imbalance, I think that extent of refreeze isn't going to do to much interesting, as irrespective of water temperature, the surface will always be able to freeze. (It's a non-linear process)
This however could mask a much bigger volume drop as the insulation effect of ice/show combined with slightly warmer water prevents thickness growth.

Are the end of the day, the energy required to melt/freeze all the ice is treated to volume, but we can only measure area (extent when melt ponds are around) with high accuracy. So I think we could set up for a season with unexpectedly high melting without realising it.

[Jim Hunt]

I wonder about the potential for open water and wave action to melt ice in the early stages of freezing, especially if storms wander in from the Pacific.

In which case keep your eyes peeled for the voyage of the Sikuliaq in late September:

http://forum.arctic-sea-ice.net/index.php/topic,1222.msg55728.html#msg55728

plus more info at:

http://www.onr.navy.mil/reports/FY13/agackley.pdf

Objective

Conduct a complete (the first) wave-ice interaction field experiment that adequately documents the relationship of a growing pancake ice cover with a time and space varying measured wave fields

[jdallen]

Here's my speculation on the impact of el nino / warm blob on the refreeze season:

Because of more water vapor in the air from the warm waters, I think we're more likely than other years to have widespread snows across the arctic which would insulate the ice.  Thinner ice in winter this year => large melt season next year?

Welcome to the Forum lifeblack.

You are absolutely correct that increased water vapor increases the chance of snow, however, when and where the snow falls is entirely dependent on the weather.  I'd be interested to at what point of the re-freeze season does a heavy blanket of snow on the ice have the greatest impact on further ice growth.

Do we have any standard definition of when the re-freeze season actually begins??  Since our three standard metrics, area, extent and volume all reach their annual minimum at different times and then quite often bounce around the bottom for days or weeks before beginning to climb.  I would surmise that the true start of the re-freeze season is not until all three metrics have climbed by a certain amount (percent gain or absolute amount) or for a number of successive days.

If there is no current standard by which one can declare the start of the re-freeze season, I recommend that the members of this forum should develop that standard.

I'll take a first stab...

Re freeze start is the first date when the five day running average for all three metrics - volume, area, extent has been positive on five consecutive days.

That said, I think the bumping along the bottom this year may continue into October...

[crandles]

I'll take a first stab...

Re freeze start is the first date when the five day running average for all three metrics - volume, area, extent has been positive on five consecutive days.

That said, I think the bumping along the bottom this year may continue into October...

If area and volume are going up but Extent is going down, then I would suggest freeze season has begun as more is being frozen than melting.

The first single day of area and volume increasing is too soon to recognise start of freeze season, it might just be a fluke.

Not sure I like positive on 5 consecutive days either: One year might have one day declining every 3 or 4 days for quite a long time meaning a very late start to freeze season when the reality is that freezing has commenced some time ago. Other years might have a fluke 5 consecutive day increase early on then have larger losses after this. Comparing such dates between years would then be rather misleading.

Perhaps we just want after 5 day average minimum volume? Or maybe different measures are needed for different purposes?

[jdallen]

I'll take a first stab...

Re freeze start is the first date when the five day running average for all three metrics - volume, area, extent has been positive on five consecutive days.

That said, I think the bumping along the bottom this year may continue into October...

If area and volume are going up but Extent is going down, then I would suggest freeze season has begun as more is being frozen than melting.

The first single day of area and volume increasing is too soon to recognise start of freeze season, it might just be a fluke.

Not sure I like positive on 5 consecutive days either: One year might have one day declining every 3 or 4 days for quite a long time meaning a very late start to freeze season when the reality is that freezing has commenced some time ago. Other years might have a fluke 5 consecutive day increase early on then have larger losses after this. Comparing such dates between years would then be rather misleading.

Perhaps we just want after 5 day average minimum volume? Or maybe different measures are needed for different purposes?

I think you might have passed over a detail.  Five days of the five day running average being positive is my signal.

It works equally well if you turn it around, I think.

I like using all three to make it unambiguous.  Similar to "slack water" when tides are changing, I'm comfortable having a block of time at the transition from one to the other which has mixed values, and including it with the previous regime.

[greatdying2]

It's difficult to get excited about the refreeze season when the next few weeks may be the most exciting of the melt season, but I suppose that's baked into the cake...

IMHO, what difference does it make whether it starts say the last week of Sept. or the first week of Oct.? The big question is, how will the many changes in this melt season and in the climate system affect the state of the ice?

For example, the question that has been raised about snowfall seems important. Are we certain that increased snowfall retards ice growth? Does it matter how early in the refreeze season it occurs, analogously to how "melt momentum" purportedly makes an impact? "Refreeze momentum"? If so, what might affect the snowfall amount at critical times? Is it random or can we make some predictions at this point?

Another question I have wondered about is, how does "compaction" affect refreeze? For instance, could disperse ice act as anchors and lead to more stable thin ice early in the season? If so, could this increase trapped ocean (less radiative loss) and also possibly catch early snowfalls? Or conversely could early ice lead to thicker ice in Spring?

Does anyone know of related papers?

[crandles]

Semtner 1975

see http://dosbat.blogspot.co.uk/2015/01/the-simplest-model-of-sea-ice-growth.html
(link to paper is early in above post)

[crandles]

An interesting feature of the 3-layer predictions of open water is that multi-year cycles of ice thickness can occur. These cycles repeat themselves exactly, and open water may appear only once in each multi-year period (see fig 8 ). The phenomenon is made possible by the differing conductivities of snow and ice. Following the appearance of open water in a given summer, the ocean does not freeze until after the heavy snowfall season. With a thinner cover of insulating snow, the ice grows rapidly during the subsequent winter to such a thickness that severa; years with usual snow cover must pass before open water reappears. This mechanism might be important in causing multi-year anomalies in the observed ice cover.

Are we certain that increased snowfall retards ice growth?

Increased snow thickness certainly does retard thermodynamic ice thickness growth. (Snow thickness can be different to snowfall as snow falling on open ocean will not stay as snow but suspect you realise this.)

Greater thickness of snow also retards melt though albedo effect.

Does it matter how early in the refreeze season it occurs, analogously to how "melt momentum" purportedly makes an impact?

Whether the snow falls on ocean or ice that can support it matters as indicated by multi-year cycles quote from Semtner above.

If the snow thickness is the same throughout Dec-March but it is built up earlier or later during the heavy snowfall season (~Sept-Nov) then I suggest this makes little difference. Early thick snow will retard growth of ice but then you get more rapid growth of ice in Dec Jan Feb which approaches an equilibrium ice thickness. So you end the freeze season with similar level of ice thickness.

Therefore I would suggest this is the opposite of melt momentum.

[crandles]

Semtner75iceMultiyearCycles

Trying to think about the above, I imagine this applying to regions that rarely get down to zero ice.

One region might be ice free one year and then is not ice free for several years. Different regions are unlikely to be in sync with this because weather that is good for melting in one region is likely to be bad for melting in another region. So you could still get a fairly steady decline in total ice despite having such cycles taking place because different regions will tend to be at different places in such cycles.

Adjacent years do seem to have different regions where the ice retreats more than usual. However I am not sure this is really evidence for such cycles. I suspect transport of ice makes what is going on a whole lot more chaotic than simple cycles that 'repeat themselves exactly'.

[OldLeatherneck]

Semtner75iceMultiyearCycles

Trying to think about the above, I imagine this applying to regions that rarely get down to zero ice.

One region might be ice free one year and then is not ice free for several years. Different regions are unlikely to be in sync with this because weather that is good for melting in one region is likely to be bad for melting in another region. So you could still get a fairly steady decline in total ice despite having such cycles taking place because different regions will tend to be at different places in such cycles.

Adjacent years do seem to have different regions where the ice retreats more than usual. However I am not sure this is really evidence for such cycles. I suspect transport of ice makes what is going on a whole lot more chaotic than simple cycles that 'repeat themselves exactly'.

Crandles,

Thanks for posting all of this  important information.  This is the kind of information I was looking for when I started this thread. 

Someone posed the question about why I started this thread in the midst of all of the excitement with the dramatic events currently occurring in Arctic. As I have read the various threads about this year's melting, I have read many comments about setting the stage for massive losses in 2016 and varying thoughts on whether the current events might delay the start of the 2015/2016 re-freeze.  So I decided to start this thread  now to start capturing some of the thoughts and the rationale behind  them.  I realize that attention should be remained focused on current events, however within 3-4 weeks (??) we will see the end of this year's melting.  I, for one, would like to have a better understanding of everything that impacts the re-growth of ice and what that may mean for the following year's melt season.

[Steven]

Semtner 1975

see http://dosbat.blogspot.co.uk/2015/01/the-simplest-model-of-sea-ice-growth.html
(link to paper is early in above post)

To illustrate the insulating effect of snow:

I made the graph below, using the simplest model for thermodynamic growth of sea ice in the appendix of the Semtner paper.

The graph shows sea ice thickness as a function of time, from day 1 to day 200, assuming that:

1) initial ice thickness (on day 1) is 50cm
2) air temperature is constant -20°C
3) ocean heat flux is negligible
4) there are no dynamic processes such as compaction/ridging, and
5) snow depth S is constant (i.e. no extra snowfall during the freezing season): S=0 for the blue curve, S=10cm for the red curve, and S=20cm for the green curve.

(Clearly most of the above assumptions are unrealistic.)

[OldLeatherneck]

Semtner 1975

see http://dosbat.blogspot.co.uk/2015/01/the-simplest-model-of-sea-ice-growth.html
(link to paper is early in above post)

To illustrate the insulating effect of snow:

I made the graph below, using the simplest model for thermodynamic growth of sea ice in the appendix of the Semtner paper.

The graph shows sea ice thickness as a function of time, from day 1 to day 200, assuming that:

1) initial ice thickness (on day 1) is 50cm
2) air temperature is constant -20°C
3) ocean heat flux is negligible
4) there are no dynamic processes such as compaction/ridging, and
5) snow depth S is constant (i.e. no extra snowfall during the freezing season): S=0 for the blue curve, S=10cm for the red curve, and S=20cm for the green curve.

(Clearly most of the above assumptions are unrealistic.)

Thanks Steven.  That certainly implies that an early and heavy blanket of snow can have a significant limiting on the effect eventual thickness of the ice pack in spring.  Another fact we have to consider going forward in the future is rapid warming of the Arctic in the dark months.  Last month Neven had an very fascinating post on the ASIB entitled "A Wetter and Warmer Arctic". http://neven1.typepad.com/blog/2015/08/a-wetter-and-warmer-arctic.html

The most alarming and detrimental to ice growth is the decadal increase in November to April temperatures from 2003 to 2013, with November being the month with dramatic annual gains. See quote below with my highlights:

The Arctic became warmer and wetter since the beginning of the 21st century, a self-reinforcing trend likely to continue because it is linked to sea-ice melt and more persistent open-water conditions in the world’s northern ocean, a newly published study concludes.

Data from NASA shows that average surface temperatures across the Arctic Ocean increased an average of 0.16 degrees Celsius per year from 2003 to 2013, and air temperatures rose 0.09 degrees Celsius annually over the same period, says the study, published online in Geophysical Research Letters.

The changes weren't evenly distributed, though. They were dominated by large increases in the November-to-April period,during which Arctic-wide surface temperatures rose 2.5 degrees Celsius and air temperatures rose 1.5 degrees Celsius from 2003 to 2013.

November saw the biggest increases in “skin temperature” (defined as temperature at the Earth’s surface), and air temperature, with an average annual rise of 0.42 degrees Celsius on the surface and 0.32 degrees Celsius in the air, said the study, by Linette Boisvert of the NASA-affiliated Earth System Science Interdisciplinary Center at the University of Maryland and Julienne Stroeve, a research scientist with the National Snow and Ice Data Center in Colorado.

[seaicesailor]

Chances of a mild Winter are increased given El Niño 2014-2015, and therefore land/ice snow cover could be low starting Spring 2016. Low snow cover was enabler of strong melting in July despite relatively weak surface melting in May-June** .

I am getting convinced (by the fact of how much extra energy it takes to melt snow, its high albedo and by its insulating properties, and by Frivolousz posts :--) that no snow cover (not only land but also ice) is necessary for strong melting.

** Nevertheless, I considered at the time that melt ponding was being strong in May where it mattered: Beaufort / Chukchi, and this for one was right

[seaicesailor]

Semtner75iceMultiyearCycles

Trying to think about the above, I imagine this applying to regions that rarely get down to zero ice.

One region might be ice free one year and then is not ice free for several years. Different regions are unlikely to be in sync with this because weather that is good for melting in one region is likely to be bad for melting in another region. So you could still get a fairly steady decline in total ice despite having such cycles taking place because different regions will tend to be at different places in such cycles.

Adjacent years do seem to have different regions where the ice retreats more than usual. However I am not sure this is really evidence for such cycles. I suspect transport of ice makes what is going on a whole lot more chaotic than simple cycles that 'repeat themselves exactly'.

Spot on.

[DoomInTheUK]

My gut feeling is that how the developing El Nino influences the regional weather of the Arctic will be critical.
It may transport warmth and moisture, or it may produce a blocking pattern that allows the Arctic to cool.
It would seem that the number and size of the variables affecting the Arctic are changing and about all we can say for certain is that they all point to increased melt.

One year fairly soon all those ducks will be in a line and then who knows - it might even make a small segment of the nightly news, for a day or so.

[OldLeatherneck]

My gut feeling is that how the developing El Nino influences the regional weather of the Arctic will be critical.
It may transport warmth and moisture, or it may produce a blocking pattern that allows the Arctic to cool.
It would seem that the number and size of the variables affecting the Arctic are changing and about all we can say for certain is that they all point to increased melt.

One year fairly soon all those ducks will be in a line and then who knows - it might even make a small segment of the nightly news, for a day or so.

It will be interesting to see watch the DMI Temps North of 80².  The deviations up from the mean are clear indication that warm is is entering from the continents.  I'd be interested to know in which months elevated temps have the greatest impact on any of our standard metrics, area, extent and volume.

[Paddy]

Lots of speculation here, but surely we should look at how refreezing went in previous El Niño years?

[TerryM]

It will be interesting to see watch the DMI Temps North of 80².  The deviations up from the mean are clear indication that warm is is entering from the continents.  I'd be interested to know in which months elevated temps have the greatest impact on any of our standard metrics, area, extent and volume.

I believe some of the high air temps we're seeing might, in part be attributed to high water temps left after the cyclones have past.
Tor is reporting that Healy at ~2.5 degrees from the Pole, is experiencing 0C degree air & -1.5C water.
Could our season be a long one?


Terry

[TerryM]

It will be interesting to see watch the DMI Temps North of 80².  The deviations up from the mean are clear indication that warm is is entering from the continents.  I'd be interested to know in which months elevated temps have the greatest impact on any of our standard metrics, area, extent and volume.

I believe some of the high air temps we're seeing might be attributed to high water temps left after the cyclones have past.
Tor is reporting that Healy at ~2.5 degrees from the Pole, is experiencing 0C degree air & -1.5C water.
Could our season be a long one?


Terry

[Frivolousz21]

Here's my speculation on the impact of el nino / warm blob on the refreeze season:

Because of more water vapor in the air from the warm waters, I think we're more likely than other years to have widespread snows across the arctic which would insulate the ice.  Thinner ice in winter this year => large melt season next year?

Extra snow is higher albedo for longer.

Slowing melt rates And retarding surface heating a lot versus bare ice

2014 is a great example of snow albedo as a negative feedback while 2015 is the stark opposite.

[sofouuk]

does  the 'multiyear cycle based on depth of insulating snow' model partly explain the apparent cycle since 2007 (2007 lots of open water, 2008-09 'recovery', 2010-11 gradual decline, 2012 crash with open water, 13-14 recovery, 15 gradual decline, and so presumably on)? don't remember reading that specific argument before, presumably it has been made and I missed it?

[seaicesailor]

2015F buoy (located at CAB, 81N 170E, North of the ESS "protuberance") shows much slower bottom melt and already 12 cm of snow. It seems the buoy will survive until next Summer :--)

[Bruce Steele]

Projecting out till next spring and the ice extent I might expect for the Bering sea. PDO phase has a direct correlation to Bering Sea ice extent and PDO should still be positive next spring. PDO may weaken a bit if the current batch of Pacific hurricanes/typhoons break down the RRR and the blob cools but I would expect it to stay positive at least until we re-enter La Nina sometime later next year. So spring ice in the Bering will likely be weak and with some favorable wind conditions I would expect early melt conditions for the Waters north of the Bering Strait, like this year.

58
Arctic sea ice / Re: The 2015 melting season
« on: June 28, 2015, 05:03:39 PM »
Schumacher and Stabeno make the argument that the PDO phase has large effects upon Bering Sea Ice extent.

http://www.pmel.noaa.gov/pubs/outstand/stab2529/features.shtml

From a Jan. dosbat posting I borrowed a chart showing Bering Sea ice anomalies 79-2015. This provides info to test the Schumacher paper PDO / Bering Ice correlation . Thanks Chris

 http://dosbat.blogspot.com/2015/01/where-is-ice-in-bering-sea.html

The PDO flipped back to positive last year and the Bering Ice has has responded as predicted. It will be interesting to see what happens if PDO phase changes back negative with the associated La Nina that usually follows a strong El Nino( which is presumably still in the works )

The above was a repost from earlier this season.

On the Atlantic side I am less informed . The appearance of the cold water anomaly south of
Greenland may or may not have something to do with Greenland ice melt but it's persistence into next year and the slowdown in the gulf stream will IMO decrease chances of advanced ice melt conditions we saw in the north Atlantic before it appeared. If however the gulf stream were to strengthen and move warm water north then we might see strong early melt conditions on both the Atlantic and Pacific fronts. It isn't just the warm water pushing north but the atmospheric heat that warm water can transfer beyond the reach of the water itself that contributes to early melt and the chances of
another advanced melt season for the Arctic.
 On the Pacific side I would argue for a repeat of this year , on the Atlantic i don't know.

[Paddy]

Lots of speculation here, but surely we should look at how refreezing went in previous El Niño years?

I'm going to take this back.  Looking at it, the last really big El Nino was in 1998-1999, when the starting conditions were so different (in terms of the amount of ice present etc.) that it's just not comparable.

Going to be an interesting year.

[Bruce Steele]

I don't know why the Dosbat link failed in my post above failed but here it is again. In the chart the years prior to 1999 tend towards negative ice anomalies and the years after positive anomalies until we again transitioned in late 2014 and into 2015.

http://dosbat.blogspot.com/2015/01/where-is-ice-in-bering-sea.html

The PDO reference in the Stabeno paper is in the second to last paragraph.

[Tor Bejnar]

Is this the temperature spike the accompanies the start of sea ice freezing?

[kingbum]

In the Arctic clearly without a doubt it will be below freezing....so what causes less ice growth? An early snow prevents volume growth by insolating the pack and not allowing the heat just below the water surface say at 100M deep to escape...late season snows however preserve whatever ice has developed during the season...the evidence? Look how long this year it took for the massive snow piles where the cities of Buffalo and Boston dumped its snow...In Boston it lasted until Mid-July and Buffalo well into August. The snow/ice was actually being insolated by city garbage believe it or not. Snow and wind are your two variables...so are we expecting a dry fall and a wet spring? Those are the questions we need to answer to get an idea about the ice recovery in a refreeze

[iceman]

This post from The 2015 melting season thread is also relevant to re-freeze, so replying here.

http://forum.arctic-sea-ice.net/index.php/topic,1149.4000.html

that's very interesting. So it appears the Arctic Winter snow is usually denser than fresh dry snow and so a less good insulator.
   ....

Yes, freshly fallen dry snow has a density of less than 0.1 g/cm³, whereas the typical density of the snow layers on the arctic sea ice in winter seems to be about 0.3 g/cm³.

Melt-refreeze cycles in early autumn seem to play a role in compacting the snow.  And thick snow layer may be compacted somewhat by gravity.

The graph below shows the average snow density for each month as measured at drifting stations on Arctic sea ice.  Source: Warren et al. 1999.  The average snow density in the freezing season is about 0.3 g/cm³ (= 300 kg/m³), consistent with the numbers mentioned before.  It gradually increases from about 0.25 g/cm³ in September, to 0.32 g/cm³ in April and May.

re:  "Melt-refreeze cycles in early autumn seem to play a role in compacting the snow."
This is noteworthy because it could increase variability in ice cover year-over-year (though it might reduce seasonal variability).  The causality goes in both directions:
    If there's a low (or late) minimum in ice area/extent,
       snow that accumulates early will be at higher latitudes and fall at lower temperatures (on average)
       so will be less subject to melt/refreeze cycles
       and will retain more of its low density.
    If the new snow compacts less,
       it will reduce ocean heat loss through the ice below it
       and slow the rate of thickness regrowth during the freeze season.
The latter effect might be hard to spot, because it acts opposite to the more dominant influence of heat loss from open water.  However, ice re-growth from that heat loss appears to accelerate early in the freeze season (variable but centered on November, according to PIOMAS), whereas the retarding effect of less-compact snow might persist longer.  To test this idea, I compared a group of recent years with low extent minima (2007, 2008, 2011, 2012) to those with higher minima (2006, 2009, 2010, 2013, 2014).  There is some indication of a slowdown in volume increase around December-January.  Caveats:
       a)  This is just from eyeballing the charts; too few years for a robust statistical analysis
       b)  There could be other causes for the slowdown than less-compact snow
       c)  Volume anomaly is anyway a loose proxy: ideally what you'd like to measure is thickness over time of ice covered by snow that fell early in the freeze season
   
My sense is this causal chain is a positive feedback on an interannual basis.  It's probably minor compared to weather and decreasing stability of the overall Arctic system, but might well contribute to increasing variability.

[Andreas T]

I have read on various occasions (and it can be seen on buoy footage) that snow on the ice is driven by wind, i.e. moved around once it has fallen. This would explain the higher density to some extent, compacting snow flakes into more rounded grains. It also means that snow thickness varies a lot, forming drifts especially behind ridges. Not sure what consequences this has for freeze and melt, but one should not think of it as a even layer of fallen snow.

[johnm33]

In the UK dry snow is extremely rare, so maybe you need extreme cold and or altitude to generate it, also whenever cracks form you'll get mist rising from the sea before it freezes, and if the snow blows through the mist it'll pick up moisture. When the sea freezes in the cracks drifts will immediately begin to fill them, the mist rolling over the surface will also condense as it passes forming crusts between layers, the crusts giving a false impression of coherence, but coherent enough in some cases for the drift fill to be dragged below sea level and become waterlogged in sub zero seawater.
I've been following SSS from http://www7320.nrlssc.navy.mil/hycomARC/arctic.html and watching the the fresher arctic water drain through the CAA into Hudson and Baffin/Labrador to an even greater extent than last year, how much this will affect the refreeze is anyones guess. Mine is that Hudson will freeze earlier and more completely than last year and be more persistent into next years melt. The less saline water will continue to drain from the arctic arriving in Labrador full of energy leading to plenty of evaporation thereabouts. It'll continue down the east coast hugging that coast possibly raising sea levels a little more locally. When it reaches the Gulf stream their opposite inertial energies will cause drastic mixing, slowing the NAD even more leading to an even cooler North Atlantic leaving Europe a hostage to fortune weatherwise but probably a very cool spring. With the warm blob feeding into the arctic and the possibilty of a shallow refreeze against the CAA we could be watching the beginning of the draining of the surface waters of the arctic through the NWP complex.

[ktonine]

Extra snow is higher albedo for longer.

Slowing melt rates And retarding surface heating a lot versus bare ice

2014 is a great example of snow albedo as a negative feedback while 2015 is the stark opposite.

Friv, this is true for spring. It is not true for fall/winter.  In the fall and winter snow helps to retard heat loss and reduce ice thickening.  Making for a warmer arctic winter *below* the snow.  Air temperatures can actually get colder because little of the ocean heat is being transferred to the air.

In the spring, early snow melt is a significant factor in 'pre-conditioning' the ice for large losses.  It creates early melt ponds, reduces albedo, and let's insolation get to work directly on the ice.

[Shared Humanity]

I have always found the refreeze (winter) as interesting as the melt season and, with the new lower minimums, the winter is as unpredictable as our melt season. With vastly more open water in the Arctic, snowfall across the northern hemisphere has increased as the cold air picks up moisture from the now open seas. In the peripheral seas, the refreeze is a question of how soon and how fast.

The CAA is of real interest as the refreeze in 2014 commenced in earnest at the end of September and the minimum this year is considerably lower than last year. The Beaufort is also going to be fun to watch. The ESS as well. With all of the heat on the Pacific side and CAA through this melt season, will melt be later and/or slower than last year?

And what of the parade of cyclones in the North Pacific? I was never concerned about a significant impact in the Arctic basin but could this retard the refreeze in the Bering Sea and/or the Sea of Okhotsk?

I know that one thing I will be watching closely is evidence of stuck weather patterns. Alaska had an absurdly warm winter last year as the RRR drove warm air masses all the way into the Beaufort. Will we see a similar pattern or will some other region of the northern hemisphere be affected this year? This, more than anything, is why I find stuck patterns and regional weather so fascinating. I believe the warm winter last year contributed a great deal to the wide open basin on the Pacific side this melt season.

[Shared Humanity]

The very warm weather in the Pacific last year resulted in the lowest maximums in the Sea of Okhotsk and Bering Sea since 1979.

[Peter Ellis]

Friv, this is true for spring. It is not true for fall/winter.  In the fall and winter snow helps to retard heat loss and reduce ice thickening.  Making for a warmer arctic winter *below* the snow.  Air temperatures can actually get colder because little of the ocean heat is being transferred to the air.

Well yes, but snow that falls during autumn/winter will still be there in spring, short of a magical army of elves turning up with brooms.  More snow in autumn/winter means the ice grows slower, but also means it has a thicker reflective/protective blanket in spring, which will slow the onset of melt.

Buoy 2013F was a perfect example of this.  It's not travelled far, so it's a reasonably fair geographic comparison.

In 2013-14, a thick blanket of snow fell early in autumn, and there was correspondingly less thickness accumulation over winter due to the insulating effect of the snow.  However, the snow then never fully melted back, and there was no surface loss the following summer. In 2014-15, the snow was much thinner, and so it was rapidly ablated in spring, allowing the onset of surface melt - and indeed complete melt-through after only a few weeks).
http://imb.erdc.dren.mil/irid_data/2013F_thick.png

I don't think it's possible to make blanket pronouncements about whether winter snowfall is good or bad for ice retention, when there are so many variables involved. Probably an area that would benefit from some more detailed modelling by people who actually know about such things >.<

[jplotinus]

Yes, the melting season has been officially ended. NSDIC-9/11 @ 4.41million, confirmed by Neven. ;-)

IJIS reached a little lower, 9/14 @ 4.25(7)million.

Of the two, NSDIC/IJIS, which is considered the more accurate for baseline purposes for the re-freeze? :-0

[Neven]

IJIS/JAXA uses a satellite sensor with higher resolution AMSR-2, but other than that you can't say one is more accurate than the other, as it's not really useful to compare datasets with each other, when differences are this small. As long as they keep doing what they do consistently every year, they're all equally useful.

[Tor Bejnar]

The freezing season has started, for sure, but the melting season hasn't ended!  As long as sea surface temperatures under the ice are above the freezing point (-1.8ºC, we're told), and the air isn't too cold, a little melting will continue. 

Where the Healy was yesterday(N 85°06'): air temp -8.3ºC  water temp -1.7 ºC.

[pauldry600]

Minimum surely reached now

I predicted 3.172  min on other cryosphere thread
Did it go near that?

If you take 10 from the cryosphere max you more or less have the min.

Wonder how long before iceless summer

2030 at latest

[Neven]

Minimum surely reached now

I predicted 3.172  min on other cryosphere thread
Did it go near that?

CT SIA has reached its lowest daily number on September 8th at 3.093 million km2.

[AmbiValent]

So my extent guess was too high and my area guess too low...

[kingbum]

My guess and its just a guess but for best ice retention and gains what needs to happen is a dry autumn and early winter dominated by high pressure and is storm free but cold....Then a snow blitz insolating the pack in February and March...so that by mid April everything needed has been there...prolonged exposure to the cold helping with the refreeze then a blanket of snow to preserve the winter's work...I seen this process first hand on the Great Lakes a couple of years ago and it got to 95% coverage

[jdallen]

My guess and its just a guess but for best ice retention and gains what needs to happen is a dry autumn and early winter dominated by high pressure and is storm free but cold....Then a snow blitz insolating the pack in February and March...so that by mid April everything needed has been there...prolonged exposure to the cold helping with the refreeze then a blanket of snow to preserve the winter's work...I seen this process first hand on the Great Lakes a couple of years ago and it got to 95% coverage

Unfortunately with the increased moisture, it looks to me like the snow may be happening in tandem with the refreeze.

[Lord M Vader]

Jdallen: agree with that!

Cyclonic weather seems to dominate the Arctic during the next week or so. The Arctic Oscillation (AO) is currently weakly positive. Since the middle of June, the AO has mainly been negative. Continued cyclonic weather from now and until november followed by a switch to the negative phase would IMO be a really bad outcome as snow will insulate the ice and prevent it from growing. But if too much snow falls on the ice pack it will likely help to preserve the ice. What did the situation look like in 2006 and 2009? In 2009, I know the AO went berserk by late fall but was the AO positive before it went berserk down to the bottom?

And is there any correlation between AO and El Niño during winters?

//LMV

[seaicesailor]

Jdallen: agree with that!

Cyclonic weather seems to dominate the Arctic during the next week or so. The Arctic Oscillation (AO) is currently weakly positive. Since the middle of June, the AO has mainly been negative. Continued cyclonic weather from now and until november followed by a switch to the negative phase would IMO be a really bad outcome as snow will insulate the ice and prevent it from growing. But if too much snow falls on the ice pack it will likely help to preserve the ice. What did the situation look like in 2006 and 2009? In 2009, I know the AO went berserk by late fall but was the AO positive before it went berserk down to the bottom?

And is there any correlation between AO and El Niño during winters?

//LMV

Anybody knows what happened in Winter 2011-2012 in terms of snow? Are there records of that?

[Neven]

You can flip through monthly anomaly charts here.