2020/21 Freezing Season Predictions

[PragmaticAntithesis]

It's that time of year again, almost time for the least exciting but most important part of the ice cycle: freezing season!

I think this year's freezing season is going to be interesting, as the Great Arctic Anti-Cyclone has put a lot of heat in the peripheral waters, particularly on the Siberian side. Might we see a particularly low maximum next year as warm water fails to re-freeze?

Starting this thread before freezing season actually starts, as some people may want to predict things in advance.

[Phil42]

Not a very bold prediction given the state of the ice, but I predict a record or near-record late refreeze of the Kara and Laptev Sea. My reasoning behind this is because those two seas...:

a) had a very early complete melt-out this year. The Laptev Sea broke the record by about 2 weeks, the Kara had (eyeballing) a tied record-early melt-out with some other years. This can be seen in Attachment 1 (graph from here).

b) were exposed to very high temperatures compared to normal. Throughout the whole summer (and also a long time in spring) these areas recorded very high temperature anomalies. Because of the early melt-out, this additionaly means that during a long time period it was open water exposed to those temperatures. Here are the temperature anomalies for May, June, July and August.

c) are expected to keep being exposed to really high temperatures in the coming days, as can be seen in the 7-day forecast.

I think all these factors put together make a record-late refreeze possible or even likely. Of course cold, clear weather with little wind in October could change this outlook.

[Paddy]

The linear trend would suggest an NSIDC March monthly average of about 14.4 m square km, based on eyeballing the graph here :http://nsidc.org/arcticseaicenews/2020/04/

The only firm prediction I'd make is that it'll be somewhere between 14 and 15 million square km, but if I had to go for one of our standard polling bins I'd say 14.25 to 14.75 square km. Which covers a range of outcomes from lowest ever to 7th lowest, so it seems fairly safe.

[Niall Dollard]

Not a very bold prediction given the state of the ice, but I predict a record or near-record late refreeze of the Kara and Laptev Sea. ...
 Of course cold, clear weather with little wind in October could change this outlook.

I agree advancement of the pack towards the Russian side will be very slow.

On the other hand not too sure about advancement from the Russian coast. We have seen some wild temperature swings in Siberia of late and you could not rule out a cold October. The shallow coastal edge of the Laptev and ESS could see a rapid temperature drop. But there is a limit to how far this will extend and it is likely it still will be a long time before the pack joins up with ice coming out from the land.   

The remnant Beaufort ice will assist advancement here and subsequently north of Alaska. But Chukchi freeze up is bound to be slow and will be very dependent on synoptic situation in Nov/Dec.

Barents and much of Kara away from favoured coastal spots is bound to be very slow.

Another thing to wonder, is what will happen in the Nares Strait ? Last year I was surprised to see an early freeze up and back in 2012 after the record low extent, we did see an early November formation of the arch. North of Greenland much of the ice is thin - but there are some big blocks so who knows what could happen if a block arrives down at an opportune time (when combination of weather/wind/tides are favourable) ?

[oren]

I think salinity can play a bigger role than water temperatures in determining difficulty of freeze-up. Seas that have been ice free for while, especially if they've been hit by repeated storms and wind events, will be well mixed and late to freeze. This would include all the Siberian seas as well as the Chukchi. OTOH the CAB, Beaufort and CAA should have an easier time.
As for the extent max, it is set in very peripheral seas, mainly the four Bs - Baffin, Barents, Bering and (B)Okhotsk, and isn't very interesting. However max volume in May 2021 could be relatively low if Arctic Basin refreeze is delayed. I think there is good potential for 2nd lowest max behind 2017.

[Tor Bejnar]

A-Team just posted a photo in the MOSAiC thread showing some melt ponds frozen over and others not.  I bet salinity is the main difference.

[Comradez]

Will we see a full freeze-up of the Kara sea this winter?  I remember last year I was doubtful that we would see much of any freeze-up in the Bering Sea because 2019 saw an extremely early melt-out of the Bering and Chukchi...but then the Bering Sea froze right up and attained a fairly large extent in the spring of 2020.  So I guess the Kara will freeze-up like normal this year too, odds are.  But perhaps the "bight" of open water that usually reaches from the Barents Sea to the north of the Kara Sea will be especially pronounced next spring...and perhaps the Barents Sea will not witness much of an ice advance past Franz Josef Land and Spitsbergen this winter....

The Laptev Sea will freeze up for sure.  There's too much soon-to-be cold land surrounding it. 

If we ever get a real "Blue Ocean Event" one of these summers in the near future, we will probably see a "donut" of ice in the subsequent fall all around the landmasses ranging from north of Greenland, to north of the CAA, to the Beaufort, Chukchi, ESS, and Laptev, but with the CAB actually mostly open water or extremely slushy ice that never truly freezes into a solid pack that next winter, and probably the Kara too.

[PragmaticAntithesis]

Knowing how the season has started, I think we'll likely see a record low maximum, but not by much.

Somewhere around 13.7±0.3sqMm extent seems like a good ball park estimate for me.

[jdallen]

Knowing how the season has started, I think we'll likely see a record low maximum, but not by much.

Somewhere around 13.7±0.3sqMm extent seems like a good ball park estimate for me.

A hair low, but close I think.

My prediction is three fold:
High probability      - 14.0-14.5, shaded towards the low end.
Medium probability - 13.75-14.25, shaded towards the middle.
Very high probability 13.75-14.5, of course.

It depends if the refreeze is more like 2019/2020 or 2017/2018.

Looking at the NOAA data there seems to be a natural limit driven by geography which I suspect lies between 14.0 and 14.5 million km2, which may be the range our winter max falls in for a while.

Annual minimum will continue to decline along with volume, and be driven more by how soon the melt season starts and how hot Siberia gets, and how early.

There also I'll predict, volume at max for sometime is going to plateau at around 20,000-23,000 km3 similarly, drifting down slowly to the high teens.  This year I'll place my bet on 21,000km3 +/- 250.

However, if we get the inflow from cyclones I'm half expecting, my estimate could be high by over 1000km3.

[Simon]

I predict volume to be 20,000km3 at maximum.

[PragmaticAntithesis]

A bit of a short term prediction, but I think we'll see some record low days around the end of next week covering the 2019/2012 crossover.

[Paddy]

A bit of a short term prediction, but I think we'll see some record low days around the end of next week covering the 2019/2012 crossover.

Seems reasonable.

[RikW]

I guess we will finish around/ a little below the last 10 year average as maximum; thus we will have high/ above average refreeze numbers; There is more sea refreeze, because a lot has melted; The ice will probably be thin though;

[gerontocrat]

JAXA Sea Ice Extent data

If remaining extent gain in October is at or below the previous 10 years average, the Ocober monthly average will be a record low at 5.46 million km2, below 2019 at 5.48 million km2, and 2012 at 5.63 million km2. It would also be the first time for two back to back record October lows since 1984 & 1985.

[jdallen]

I predict volume to be 20,000km3 at maximum.

On reflection, and seeing how the refreeze is starting, I think I need to revise my estimate down as well. 

I think even at 20,000, we may be high this season.

[PragmaticAntithesis]

A bit of a short term prediction, but I think we'll see some record low days around the end of next week covering the 2019/2012 crossover.

Horrah, I actually got something right for once!

[gerontocrat]

JAXA Arctic EXTENT

The projection of the maximum using the last 10 years average extent gain is 13.04 million km2, 0.84 million km2 below the record low of 13.88 million km2 in March 2017.

The source of this massive drop is the Sept 2020 2nd lowest minimum and very low extent gains from minimum to date. In 2007, 2012 and 2019 record low minima were followed by a strong rebound in October. After those extreme gains, extent gains were pretty much average for the remainder of the freezing season.

However, in 2016 this October rebound did not happen - extent gains were pretty much at average for the whole freezing season - hence the record low maximum in March 2017.

So my speculation that belongs to me is that extent gains in the remainder of October will be the major influence on the March 2021 maximum. But maybe the rebound will be delayed to November?

At the moment my guess is for a record low maximum, though by much less than the projection.

[Positive retroaction]

it's interesting
this year will it do like last year or 2012 with a rapid gain, but shifted?
or 2016/2017 with a classic gain?
because of the very strong anomalies on kara / ESS / laptev, rather like 2017 concerning me
so I bet on a record maximum, between 13 million and 13.5 million, for me 13.2 million is a good probability, with a hudge deficit particulary on barent and west kara seas

[gerontocrat]

I was fascinated by A-team's analysis of freezing in the High Arctic being "a tale of two halves", which if you haven't read it is at....

https://forum.arctic-sea-ice.net/index.php/topic,3299.msg290364.html#msg290364

So I've been looking at the graphs from NSIDC data. My first observation is that the very fast Beaufort freeze will end soon - because in a few days it will probably be complete. The question is then - will the Chukchi freeze take up the slack? It is freezing fast now, but in A-tem's lower half. Will high SST's at the Paciific end slow Chukchi re-freeze down?

I attach the Beaufort and Chukchi sea ice extent graphs, plus SST anomalies maps from DMI for October 18 2020 and October 10 2019 (the white areas being ice covered).

[gerontocrat]

A-Team's Tale of Two halves" - observations (continued)

As the freezing of the Beaufort completes, and even if the Chukchi freeze continues to be fast, then any further delay to freezing of the High Arctic seas from the ESS to the Barents is going to push Arctic sea ice extent and area further into uncharted territory.

The SST anomalies map above show that there is plenty of above average surface ocean heat, and A-team plus others convince me, at any rate, a compromised Halocline makes ocean heat at greater depth available to resist surface freeze.

Certainly the extent graphs for the Barents, Kara and Laptev show zero extent increases, while for the ESS extent growth is slow.

[gerontocrat]

A-Team's Tale of Two halves" - observations (continued)

I attach the sea ice extent and area graphs of the Central Arctic region - 3.224 million km2. Extent is 150k less than 2012, and 200k less than the 2010's average. Area is 2nd lowest, 2016 lowest.
The questions are - how slow to freeze and will freezing be complete?

And a prediction, NSIDC October average extent will be at least a record low by 250k, at leat 750k km2 (9 years) lower than the linear trend value.

Will November be the catch-up month? (or December, or like 2016, not at all?)

[Stephan]

gerontocrat - thank you for the analyses. Very interesting and informative indeed.

[PragmaticAntithesis]

Today's ice gain (tomorrows posts) will have a century break in JAXA extent.

In any other late October this would be the safest bet ever...

Update: wow, that was easy