Support the Arctic Sea Ice Forum and Blog

Author Topic: The 2020/2021 freezing season  (Read 100141 times)

Freegrass

  • Nilas ice
  • Posts: 1641
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: The 2020/2021 freezing season
« Reply #450 on: October 22, 2020, 04:01:48 PM »
The flute playing mouse is getting its ass sucked into the Beaufort Sea...
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

wili

  • Young ice
  • Posts: 3275
    • View Profile
  • Liked: 590
  • Likes Given: 391
Re: The 2020/2021 freezing season
« Reply #451 on: October 22, 2020, 04:40:50 PM »
(That's not a sentence that has ever been written before, I'm guessing! :) )
"A force de chercher de bonnes raisons, on en trouve; on les dit; et après on y tient, non pas tant parce qu'elles sont bonnes que pour ne pas se démentir." Choderlos de Laclos "You struggle to come up with some valid reasons, then cling to them, not because they're good, but just to not back down."

Freegrass

  • Nilas ice
  • Posts: 1641
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: The 2020/2021 freezing season
« Reply #452 on: October 22, 2020, 04:59:21 PM »
(That's not a sentence that has ever been written before, I'm guessing! :) )
;D ;D ;D
And a good thing I used "into" instead of "in"...  ::)
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

Glen Koehler

  • Frazil ice
  • Posts: 382
    • View Profile
  • Liked: 381
  • Likes Given: 811
Re: The 2020/2021 freezing season
« Reply #453 on: October 22, 2020, 05:24:34 PM »
Once the water is back to to freezing, the ice will form at the same sort of rates, but it'll be a few days later than in the past...
I'd say it will be a faster refreeze as the surrounding air and the continents will be much colder by then than during previous refreezes. Late but fast refreeze. We shall see
      The Guardian article mentions a possible ecological impact on nutrient transfer from the delayed Laptev Sea refreeze.  It seems likely that once refreeze begins it will be more rapid than "normal" because it will be occurring at a later date.  That makes me wonder if the rate of Arctic Ocean refreeze has important but little-discussed impacts.  If the ice pack edge advances many more miles per day than normal, how does that affect the microscopic and macroscopic organism communities that interact with the water/ice environment?

       Ice vs. water is a major habitat change, and the rate at which that habitat shift occurs could have consequences.  While I doubt that ice-edge advance is going to be so fast as to outrun the ability of air-breathing marine mammals that need access to open water to relocate, that's an extreme (though I think implausible) example of the kind of scenario that comes to mind.  What seems more likely is some effect on the colonization, population growth rate, and niche partitioning of microflora/fauna on newly formed ice.  For example, it may make a difference to community structure if there only 3 versus 30 days between initial colonization and the date when discriminating environmental conditions occur. 

       The relative timing of such events may have trivial consequences, or it may not.  Small differences over such a large scale can have a large impact, especially in a tightly linked system where each domino affects all the following dominoes.  Even though Extent / Area / Thickness / Volume will probably return to closer to the normal range quickly once refreezing begins, even getting back to a closer match with "normal" values after a late refreeze start and rapid rebound may bring with it subtle but significant qualitative physical, chemical, biological and/or behavioral differences that are not apparent from the quantitative Ex / Ar / Th / Vol measurements. 

       The fact that the Russians could not really test their new ice breaker on a run to the North Pole because the ice was too thin and broken is not an Earth-shaking consequence, but it exemplifies how changes ripple through a system in unforeseen ways.  Everything is connected.  I do not expect obvious or catastrophic impacts, but the potential effects of refreeze timing and rate do seem worth noting.  I wonder if/how Arctic scientists are tracking such potential qualitative impacts.
« Last Edit: October 23, 2020, 04:13:39 PM by Glen Koehler »

BornFromTheVoid

  • Nilas ice
  • Posts: 1253
    • View Profile
  • Liked: 579
  • Likes Given: 204
Re: The 2020/2021 freezing season
« Reply #454 on: October 22, 2020, 05:32:43 PM »
Animation of the extent and distribution of ice for October 21st.
(Click to play)
I recently joined the twitter thing, where I post more analysis, pics and animations: @Icy_Samuel

mabarnes

  • New ice
  • Posts: 55
    • View Profile
  • Liked: 23
  • Likes Given: 50
Re: The 2020/2021 freezing season
« Reply #455 on: October 22, 2020, 07:07:20 PM »
I know nothing, but I'd say that low pressure = clouds = less heat out into space. Aint it so?

Hey thank you for the response.  I was wondering about that - but I read something about tropical storms in the Pacific and convection transferring warmer (moist) air up to where it can radiate out from there ... I would figure the Arctic may be different as it's so much colder (Scoop! LOL) and maybe the clouds formed block IR from the sea surface ... geez Louise this weather thing can get complicated....

I'm really curious about this stuff so again, thank you.


UCMiami

  • Frazil ice
  • Posts: 129
    • View Profile
  • Liked: 94
  • Likes Given: 112
Re: The 2020/2021 freezing season
« Reply #456 on: October 22, 2020, 07:37:32 PM »
I don't know that there is any data on when the change over in 'net' bottom melt to 'net' bottom growth of ice has occurred historically because it is hidden from easy observation and is likely to be extremely dynamic on a very local basis, but extent growth (rapid or slow) in itself is not as important as growth in thickness to the long term state of arctic ice.

What is clear is that in order for ice to thicken there has to be ice to start with (duh!)

What I am trying to bring up is that extent by itself is only an indication of when ice thickening can possibly begin, and with the current extreme delay in the Asian sea ice extent, the ice growth in thickness is being delayed. While historically some areas of the Asian seas have had slow growth in extent, 2020 is the first year on record (as far as I know) that will likely still have an ice free NE passage on Nov 1. (And not just dodging ice, but wide a open sea lane.) The fear is that if this continues much longer even with eventual universal 100% extent on the Asian side, that first year ice will not have a chance for a 'normal' gain in thickness. Instead of >1M ice, much of the Asian sea ice could end the freezing season in a very fragile state leading to much earlier breakup and melt in 2021. Already basically the whole of the Asian side has lost a month of thickness growth, where in previous years a fairly large percentage of those seas had already started that growth.

And the fact that salinity/warmth incursion of Atlantic waters may eventually be counteracted by night and cold weather that allows the 'normal ice extent' in these seas, the effects of increased warmth and salinity will probably continue to retard the 'normal' sea ice thickening of that late extent.

Shared Humanity

  • Frazil ice
  • Posts: 245
    • View Profile
  • Liked: 149
  • Likes Given: 9
Re: The 2020/2021 freezing season
« Reply #457 on: October 22, 2020, 08:16:27 PM »
I know nothing, but I'd say that low pressure = clouds = less heat out into space. Aint it so?

Hey thank you for the response.  I was wondering about that - but I read something about tropical storms in the Pacific and convection transferring warmer (moist) air up to where it can radiate out from there ... I would figure the Arctic may be different as it's so much colder (Scoop! LOL) and maybe the clouds formed block IR from the sea surface ... geez Louise this weather thing can get complicated....

I'm really curious about this stuff so again, thank you.

H2O is a powerful greenhouse gas...there is a reason deserts cool off so rapidly at night.

HapHazard

  • Frazil ice
  • Posts: 379
    • View Profile
  • Liked: 152
  • Likes Given: 2792
Re: The 2020/2021 freezing season
« Reply #458 on: October 22, 2020, 09:12:29 PM »
Freezing season 2020/21

Siberian Seas "Ice Thickening Days" running total = 0

Pagophilus

  • Grease ice
  • Posts: 535
    • View Profile
  • Liked: 300
  • Likes Given: 443
Re: The 2020/2021 freezing season
« Reply #459 on: October 22, 2020, 09:13:19 PM »
The flute playing mouse is getting its ass sucked into the Beaufort Sea...
I shall never look at the Arctic in quite the same way again.

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #460 on: October 22, 2020, 09:16:22 PM »
Quote
runaway feedbacK? much of Siberian side sea ice could end the freezing season in a very fragile state leading to earlier breakup and melt in 2021. A few days weeks late then catch-up with really fast refreeze (even though graphs of past years don't show this behavior)?
The SST will be nowhere near freezing by the end of the month; indeed air temperatures at -3 aren't cold enough to even move the needle. It takes much deeper longer cold to lower water temperature from +1ºC enough to set the stage at -1.8º for a flash freeze especially if the upper 10m of water is in play (not talking here about a micron at the surface).

The weather system over the next 120 hrs will be bringing in a fair amount of TCW (total cloud water). This again is not conducive to radiative cooling.

I'm wondering if the TransPolar Drift will set up late or perhaps hardly at all. This amount and location of open water heat may feed back on the atmospheric pressure pattern to some extent, undercutting what is needed for persistent TPD winds. If so, export out the Fram would be diminished, mostly to what ice is in the intake funnel now.

Alternatively, since Fram export stopped back in mid-May along with no garlic press and only a quarter turn Beaufort Gyre, something at a much larger atmospheric scale has changed, with late open water only having a secondary effect.

Where is nullschool getting its SST and TCW values from or rather, are there other sources that might be better or at least independent? Hard to say: multiple sources are listed for SST; none for TCW! The latter may be derived from Band 7 (2.1 μm) on Modis where it is called Cloud Water Path at WorldView
« Last Edit: October 22, 2020, 09:49:23 PM by A-Team »

Tor Bejnar

  • Young ice
  • Posts: 3737
    • View Profile
  • Liked: 654
  • Likes Given: 444
Re: The 2020/2021 freezing season
« Reply #461 on: October 22, 2020, 09:24:06 PM »
Does anybody follow the freezing-degree-days data?  (by region would be useful) 
Arctic ice is healthy for children and other living things.

pearscot

  • Frazil ice
  • Posts: 341
    • View Profile
  • Liked: 157
  • Likes Given: 60
Re: The 2020/2021 freezing season
« Reply #462 on: October 22, 2020, 09:58:39 PM »
It's so weird to me seeing both this wave action and zero ice (but with snow on the ground) this late in October in Barrow (so glad the camera is up and running again).

I'm just so curious as to what weather/ice formation effects the Siberian side will have. Ice will form, but I think it's going to be more fragile/mobile than ever.

pls!

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #463 on: October 22, 2020, 10:26:06 PM »
Quote
freezing-degree-days? by region
Would it be easier and more informative just to query nullschool very three hours for the next five days at a handful of representative locations? Or drop a time series of screenshots onto a contouring tool? FDD is not likely to have the correct freezing temperature for the given surface salinity nor current open water as a region nor go forward in time. It's the kind of thing you would see with associated with the 80ºN DMI graph. Why, when 2m ave temperature maps are readily available?

http://psc.apl.washington.edu/nonwp_projects/landfast_ice/freezing.php
« Last Edit: October 23, 2020, 04:05:47 AM by A-Team »

Glen Koehler

  • Frazil ice
  • Posts: 382
    • View Profile
  • Liked: 381
  • Likes Given: 811
Re: The 2020/2021 freezing season
« Reply #464 on: October 22, 2020, 10:27:48 PM »
Climate reanalyzer does not foresee 2m temperatures dropping too much over the next ten days. Because some colors are dithered (for print!?!), the correspondence with the color bar is poor. Accurate tick marks cannot be put on the color bar because its pixel width is not an integral multiple of degrees. These errors degrade expensively acquired data so need to be fixed.
     The brain and 10 nimble fingers that singlehandedly (correction: 10 fingers = 2 hands) operates Climate Reanalyzer is aware of your critique.  The reason for dithering the color scales is because for the weather forecast animations it reduces file download size by a factor of 6X.  The file sizes are not so large as to matter for folks on an unlimited-data high-speed connection, but for people on a slower (or data-metered) connection (DSL was mentioned as a slower connection, and smartphone internet is an example of a metered connection), the file size does matter. 

    That person also mentioned that for stand-alone images, dithering the color scale may not be necessary so a possible change will be investigated.  Suggestions are well received, just remember that while CR may look like some well-funded institutionalized juggernaut, it really is a part-time operation by one person with a vision, programming skill, and committment who built something nobody else (including well-funded institutionalized juggernauts) had gotten around to doing, and done while juggling multiple other responsibilities and deadlines, including a recurring requirement for periods away from the keyboard to eat, sleep and other aspects of life.
« Last Edit: October 23, 2020, 04:00:43 PM by Glen Koehler »

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #465 on: October 22, 2020, 11:28:57 PM »
Climate Reanalyzer is very popular and deservedly so! Indeed I thought it was a large institutional project with bloated govt grants and large indifferent staff.

In terms of colors, keeping to small file size, faithful gif animation frames vs vastly smaller mp4, and importantly allowing values from the map to be read off on the palette (respectively selecting all map colors for a given value range), viewer expectations have been going up as they see examples elsewhere where someone got it all going. The wheel has been invented, let's get it rolling.

What I would recommend here is just copying the methods of L Kaleschke at AWI in producing what we've been discussing as AMSR2_AWI. Here the number of discrete colors is strictly clamped to 102 though 8-bit gif has room for 256. That is plenty. There is never a need for the 16.7 million colors of 24-bit color except with photography. Here we want binned colors. It is actually quite hard to find a set of even 20 related but distinguishable colors (for maximally unrelated, glassby).

With the AMSR2 there are no lossy jpg compressions, dithering for print, sq root two rotations of 45º to standard 'Greenland down' orientation of satellite data. The file size is merely 567 kB for 2432 x 3584 pixels. Plus it is internally geo-referenced so converts to netCDF/Panoply which people want for numeric array work. Plus viewers can easily highlight selected portions of the palette or easily replace it with the thousands of LUTs at ImageJ or even one they specify from scratch.

Three more easy suggestions, seldom seen together: use the standard 3413 EPSG WGS-84 stereographic projection for the Arctic, specify the scale (pixel count out to special latitude of 70ºM), and provide the land mask. The isolationism we have now is an incredible waste of everyone's time; climate change above all needs data integration from disparate sources. Maps aren't pictures, they're working data.

While inclusivity (rotary dial-up, pay per MB) is laudable, it is just not the present much less the future for the 99% who might have an interest in climate reanalysis. At some point, it is not possible to cater to the very lowest devices nor assume everyone is on campus but better to find a middle ground lest the more typical users be constantly punished.

ftp://ftp.awi.de/sea_ice/product/amsr2/v103/nh/2020/10/
« Last Edit: October 23, 2020, 03:54:13 AM by A-Team »

Glen Koehler

  • Frazil ice
  • Posts: 382
    • View Profile
  • Liked: 381
  • Likes Given: 811
Re: The 2020/2021 freezing season
« Reply #466 on: October 22, 2020, 11:50:23 PM »
     I'll pass those comments along.  Out of my league to make any comment.

Niall Dollard

  • Grease ice
  • Posts: 772
    • View Profile
  • Liked: 299
  • Likes Given: 75
Re: The 2020/2021 freezing season
« Reply #467 on: October 23, 2020, 12:14:00 AM »
Does anybody follow the freezing-degree-days data?  (by region would be useful)

Alaska FDD data here : https://www.weather.gov/aprfc/FreezingDegreeDays

Mostly below normal, but early days. Some interior parts (eg Fairbanks) are above normal

And Nico Sun's FDD charts using the DMI 2m 80N+ data  :

https://sites.google.com/site/cryospherecomputing/fdd


Tor Bejnar

  • Young ice
  • Posts: 3737
    • View Profile
  • Liked: 654
  • Likes Given: 444
Re: The 2020/2021 freezing season
« Reply #468 on: October 23, 2020, 12:28:46 AM »
Thanks, Niall.  I was looking at the DMI 80N chart (shame on me...) and it looked like it might be showing record 'minimized cooling'.  I was hoping somebody knew more than I.
Arctic ice is healthy for children and other living things.

aslan

  • Frazil ice
  • Posts: 214
    • View Profile
  • Liked: 168
  • Likes Given: 0
Re: The 2020/2021 freezing season
« Reply #469 on: October 23, 2020, 08:41:20 AM »
What I am trying to bring up is that extent by itself is only an indication of when ice thickening can possibly begin, and with the current extreme delay in the Asian sea ice extent, the ice growth in thickness is being delayed. While historically some areas of the Asian seas have had slow growth in extent, 2020 is the first year on record (as far as I know) that will likely still have an ice free NE passage on Nov 1. (And not just dodging ice, but wide a open sea lane.) The fear is that if this continues much longer even with eventual universal 100% extent on the Asian side, that first year ice will not have a chance for a 'normal' gain in thickness. Instead of >1M ice, much of the Asian sea ice could end the freezing season in a very fragile state leading to much earlier breakup and melt in 2021. Already basically the whole of the Asian side has lost a month of thickness growth, where in previous years a fairly large percentage of those seas had already started that growth.

It is a rough estimate, but if you use the correlation of Freezing Degrees Day (FDD) with ice thickness, you need ~5500 FDD to go to 2m first year ice, and ~3500 FDD to go to 1.5 meters. October is usually worth ~300 to ~400 FDD in the Arctic, so it can make a significant dent into the ice growth.
To give some more numbers, for Ostrov kotel'nyj for example. Mean temperature from 1st of October to 30th of April over the last 10 years (2010-2019) was -20.8°C, which is about 4400 - 4500 FDD. If you count from the 1st of November, this leads to 4100 - 4200 FDD. And if you ignore November and start the ice thickening the 1st of December, it makes only 3800 - 3900 FDD. This is ignoring the risk that oceanic heat flux could be strong enough this winter to weaken this correlation. If ice growth does not start in a hurry on the Siberian side, the winter would probably not be able to fully erase the memory of this melting season. Which is a great peril, as up to the last years, winter was always cold and long enough to at least bring Arctic back to some kind of a "2m FYI" state, helping to stabilize the system.

Sepp

  • New ice
  • Posts: 14
    • View Profile
  • Liked: 12
  • Likes Given: 38
Re: The 2020/2021 freezing season
« Reply #470 on: October 23, 2020, 09:09:05 AM »
It is a rough estimate, but if you use the correlation of Freezing Degrees Day (FDD) with ice thickness, you need ~5500 FDD to go to 2m first year ice, and ~3500 FDD to go to 1.5 meters. October is usually worth ~300 to ~400 FDD in the Arctic, so it can make a significant dent into the ice growth. [...]

Wow, thank you, this was pretty much exactly the information I was wondering if it existed in such condensed form. :)

blu_ice

  • New ice
  • Posts: 38
    • View Profile
  • Liked: 16
  • Likes Given: 43
Re: The 2020/2021 freezing season
« Reply #471 on: October 23, 2020, 09:59:06 AM »
[ADS NIPR VISHOP (JAXA)] Arctic Sea Ice Extent.

On the following days, 2019 is going to have several increases of more than 200K km2. As a result, on October 27th, the year 2016 starts having the lead as the lowest on record.

For 2020 to have a difference of less than one million km2 versus 2016, 2020 needs to have an average increase of more than 112.5K km2 until August 27.

Will 2020 have this average increase? If it does not have it, then on August 27th, the year 2020 will be more than one million km2 lower than any other year on record.
My guess is that on Oct 27th 2020 will have it's biggest lead. If not, this will get very interesting indeed.

gandul

  • Grease ice
  • Posts: 769
    • View Profile
  • Liked: 184
  • Likes Given: 10
Re: The 2020/2021 freezing season
« Reply #472 on: October 23, 2020, 11:12:33 AM »
It is a rough estimate, but if you use the correlation of Freezing Degrees Day (FDD) with ice thickness, you need ~5500 FDD to go to 2m first year ice, and ~3500 FDD to go to 1.5 meters. October is usually worth ~300 to ~400 FDD in the Arctic, so it can make a significant dent into the ice growth.
To give some more numbers, for Ostrov kotel'nyj for example. Mean temperature from 1st of October to 30th of April over the last 10 years (2010-2019) was -20.8°C, which is about 4400 - 4500 FDD. If you count from the 1st of November, this leads to 4100 - 4200 FDD. And if you ignore November and start the ice thickening the 1st of December, it makes only 3800 - 3900 FDD. This is ignoring the risk that oceanic heat flux could be strong enough this winter to weaken this correlation. If ice growth does not start in a hurry on the Siberian side, the winter would probably not be able to fully erase the memory of this melting season. Which is a great peril, as up to the last years, winter was always cold and long enough to at least bring Arctic back to some kind of a "2m FYI" state, helping to stabilize the system.

Have you seen the snow extent growth over Siberia and Western Canada? Anything to comment or you only rejoice on the bad feedbacks?

https://www.ccin.ca/index.php/ccw/snow/current

Apart from this, the weather is about to become propitious for a quick rebound of sea ice next week.
« Last Edit: October 23, 2020, 12:27:14 PM by gandul »

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #473 on: October 23, 2020, 11:28:25 AM »
Quote
up to the last years, winter was always cold and long enough to at least bring Arctic back to some kind of a "2m FYI" state
Good way of putting it. The 'last years' might be worth an extra look given what the Polarstern and its buoys observed for ice thickness growth. Here the FDD kelvin air temperature and oceanic heat flux have to drift thousands of km with the floe under TPD rather than stay at Kotelny Island. The TPD has yet to take shape this fall; there's no Laptev ice yet to push towards the Fram.

The rate of ice thickening is initially fast but it falls off in proportion to its attained thickness per the heat equation as it approaches 2m. This means it can largely make up for a late start and catch up to "2m FYI" but this ice differs greatly in the extent of maturation (mechanical strength, brine exclusion, ridging, snow accumulation pattern, melt pond susceptibility) from ice that begins forming in early October.

There's been no indication in recent days of the ice pack front advancing towards the Siberian side nor spot ocean temperatures there going negative much less approaching the -1.8ºC freezing point. The lobe approaching the ESS/eastern Laptev will be the first to freeze as it is -0.3ºC now and following historical precedent. The Chukchi is currently at +3.5ºC so some will probably stay open into mid-January.
« Last Edit: October 23, 2020, 11:42:00 AM by A-Team »

SimonF92

  • Frazil ice
  • Posts: 331
    • View Profile
  • Liked: 142
  • Likes Given: 69
Re: The 2020/2021 freezing season
« Reply #474 on: October 23, 2020, 11:34:19 AM »
A-Team, that visualisation is much more detailed than other SST resources and hence very useful (so thank you) but the discrete colour bins don't feel intuitive- if it isn't too much effort I think a Temps legend would be useful. Its hard to determine the range etc.
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #475 on: October 23, 2020, 12:04:03 PM »
Quote
temperature labels would be useful on the above graphic
The purple is -1.8ºC under the ice pack and it goes up from there in increments of 0.4 or so. This was taken from nullschool Ocean Currents @ Surface for 2020-10-22 12:00 UTC; there is no forecast capacity. I don't know how accurate that data is nor if it improves on or agrees with other online SST product sources. All of them show the fringing around the ice pack periphery, the disconnect of the Chukchi-Bering inflow, and overall commonality (with mild pocketing) of the Laptev-ESS.

It's quite difficult to work with the low contrasts in the dark temperature palette nullschool provides; the image is best converted to '8-bit color' in ImageJ which allows palette replacement, done above with 'glasbey inverted' which by design makes consecutive colors as different as possible. Those bands could be color-picked and replaced with a conventional color gradient palette, worthwhile on a monthly time series.

It is perhaps better for people to just mouse around on that nullschool page, a click puts a green circle down and provides temperature data at that lat/lon. CMEMS, linked above, has better tools that allow temperature transects along lines and averages in polygonal box to be displayed. It may be more accurate; alternatively nullschool has just assimilated it (ie Mercator Ocean).

If we could locate the original infrared satellite image, there would still be the problem of  calibration algorithm. It's not clear how many independent determinations of SST exist. Forecasting products imply additional use of a model of which there are several. It's an interesting question how rapidly SST can change given the thermal inertia of sea water and how that is affected by wind and cold air temperature.

In the past, we've mostly just followed year-on-year SST atlantification at various depths at Mercator Ocean. It's worth delving in deeper to daily SST details this fall because of the extraordinary delay in Siberian refreeze.
« Last Edit: October 23, 2020, 04:21:03 PM by A-Team »

BornFromTheVoid

  • Nilas ice
  • Posts: 1253
    • View Profile
  • Liked: 579
  • Likes Given: 204
Re: The 2020/2021 freezing season
« Reply #476 on: October 23, 2020, 12:36:04 PM »
I recently joined the twitter thing, where I post more analysis, pics and animations: @Icy_Samuel

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #477 on: October 23, 2020, 01:03:51 PM »
Can that be set for a narrower palette gamut (called squeeze at WorldView vs outlier clamp at CMEMS)? The bins are too large to resolve temperatures in the -2 to 0 and 0 to 2 ranges on the last frame of Nov 7th. The shoreline color seems off, not clear what temperature it is. Overall it does not seem that unreasonable: very little advance in the ice pack but with the stage still not set for rapid closure with so much open water between 0 to 2ºC.
« Last Edit: October 23, 2020, 01:15:13 PM by A-Team »

BornFromTheVoid

  • Nilas ice
  • Posts: 1253
    • View Profile
  • Liked: 579
  • Likes Given: 204
Re: The 2020/2021 freezing season
« Reply #478 on: October 23, 2020, 01:34:39 PM »
Can that be set for a narrower palette gamut (called squeeze at WorldView vs outlier clamp at CMEMS)? The bins are too large to resolve temperatures in the -2 to 0 and 0 to 2 ranges on the last frame of Nov 7th. The shoreline color seems off, not clear what temperature it is. Overall it does not seem that unreasonable: very little advance in the ice pack but with the stage still not set for rapid closure with so much open water between 0 to 2ºC.

It doesn't offer any other colour maps. It's easier to see in the full res in the link, but the discoloured shoreline is showing ice formation. Also, the dashed lines act as the 1ºC isotherms, providing slightly better resolution than the colour legend alone.
The website provides SSTAs at 0.5ºC intervals, which may help too. Not sure why they couldn't do that for the absolute SSTs...
I recently joined the twitter thing, where I post more analysis, pics and animations: @Icy_Samuel

Shared Humanity

  • Frazil ice
  • Posts: 245
    • View Profile
  • Liked: 149
  • Likes Given: 9
Re: The 2020/2021 freezing season
« Reply #479 on: October 23, 2020, 01:55:10 PM »
Have you seen the snow extent growth over Siberia and Western Canada? Anything to comment or you only rejoice on the bad feedbacks?

This is the freezing season thread so it is appropriate to focus on it.


Apart from this, the weather is about to become propitious for a quick rebound of sea ice next week.

I sure hope so.

uniquorn

  • Young ice
  • Posts: 2844
    • View Profile
  • Liked: 1296
  • Likes Given: 258
Re: The 2020/2021 freezing season
« Reply #480 on: October 23, 2020, 02:35:23 PM »
Mercator's take on SST's.
amsr2, awi dev v103 overlaid onto mercator 0m ocean temperature, sep4-oct22  (7MB)
« Last Edit: October 23, 2020, 02:46:31 PM by uniquorn »

gandul

  • Grease ice
  • Posts: 769
    • View Profile
  • Liked: 184
  • Likes Given: 10
Re: The 2020/2021 freezing season
« Reply #481 on: October 23, 2020, 03:14:30 PM »
Mercator's take on SST's.
amsr2, awi dev v103 overlaid onto mercator 0m ocean temperature, sep4-oct22  (7MB)
Ok, from your animation is it correct to say all the regions susceptible to warming from Atlantic Waters are already covered by ice or about to be covered? (exception Barents and Kara but we know those two seas are already lost to climate change)
It seems to me the anomalously warm Laptev and ESS areas are basically over the shallow shelf, which will have a record heat release prior to refreeze (consistent with the record heat income from GAAC and the season in general). So no Atlantic warmth here to speak about.

In other words, 2021 and 2022 could be rebound years as probable as big melt years. I don’t see the doom scenario here (more in line with the scientific consensus of 2040+)

aslan

  • Frazil ice
  • Posts: 214
    • View Profile
  • Liked: 168
  • Likes Given: 0
Re: The 2020/2021 freezing season
« Reply #482 on: October 23, 2020, 03:15:23 PM »

Have you seen the snow extent growth over Siberia and Western Canada? Anything to comment or you only rejoice on the bad feedbacks?

https://www.ccin.ca/index.php/ccw/snow/current

Apart from this, the weather is about to become propitious for a quick rebound of sea ice next week.

Speaking of a great peril does not sound too much like being merry and joyfull... And for the snow extent, it is no news that late Autumn snow extent is not showing a significant decrease lately. It is even more the case this year. But this fact has not prevented ongoing anomalous warmth up to now. Which is still this case this year, as temperatures are still above normal overall in Siberia, no matter the snow extent. And for the quick rebound in ice extent we will see.

Can that be set for a narrower palette gamut (called squeeze at WorldView vs outlier clamp at CMEMS)? The bins are too large to resolve temperatures in the -2 to 0 and 0 to 2 ranges on the last frame of Nov 7th. The shoreline color seems off, not clear what temperature it is. Overall it does not seem that unreasonable: very little advance in the ice pack but with the stage still not set for rapid closure with so much open water between 0 to 2ºC.

There is also the SSTs from the coral watch of the NESDIS :

https://www.ospo.noaa.gov/data/sst/contour/global_small.fc.gif

Same picture, Laptev and even more Chukchi sea are not going to freeze tomorow...

Top level directory of the files are here :

ftp://ftp.star.nesdis.noaa.gov/pub/socd2/coastwatch/sst_blended/sst5km/

And the direct link for this year :

ftp://ftp.star.nesdis.noaa.gov/pub/socd2/coastwatch/sst_blended/sst5km/night/ghrsst_ospo/2020/

aslan

  • Frazil ice
  • Posts: 214
    • View Profile
  • Liked: 168
  • Likes Given: 0
Re: The 2020/2021 freezing season
« Reply #483 on: October 23, 2020, 03:18:35 PM »
Mercator's take on SST's.
amsr2, awi dev v103 overlaid onto mercator 0m ocean temperature, sep4-oct22  (7MB)
Ok, from your animation is it correct to say all the regions susceptible to warming from Atlantic Waters are already covered by ice or about to be covered? (exception Barents and Kara but we know those two seas are already lost to climate change)
It seems to me the anomalously warm Laptev and ESS areas are basically over the shallow shelf, which will have a record heat release (consistent with the record heat income from GAAC and the season in general). So no Atlantic warmth here to speak about.

The Laptev sea extent beyond the shelf (and so the heat anomaly, by the way). Intrusion of Atlantic waters are discernible on the salinity maps of the mercator. (P.S. : And acknoledging that the heat and salinity extent beyond the Laptev into the central bassin, even under the sea ice).
« Last Edit: October 23, 2020, 03:26:31 PM by aslan »

Sublime_Rime

  • New ice
  • Posts: 41
    • View Profile
  • Liked: 37
  • Likes Given: 8
Re: The 2020/2021 freezing season
« Reply #484 on: October 23, 2020, 03:20:30 PM »
My apologies if this is an ignorant or rookie question, but has anyone looked at the integral of sea ice extent, area and volume across a given year, or year-to-date, in order to quantitatively compare them with regard to sea ice coverage?

I think that would highlight 2020's prominance more clearly over past years, even 2012, with regard to the amount of time its been in the lead and by how much. This comparison would give equal measure to all time points (as I think they 'deserve') rather than having bias as to the minimum and maximum. Such an analysis may help to see an overall trend over the years, relative to looking at particular time-points.

I think volume may be the most sensitive metric to use, as sea ice area and extent seem to have much less variation during the winter, and those large numbers may blunt changes occurring from year-to-year.
Max
Know thyself
Here to learn and connect in these wondrous and quickly changing times.

uniquorn

  • Young ice
  • Posts: 2844
    • View Profile
  • Liked: 1296
  • Likes Given: 258
Re: The 2020/2021 freezing season
« Reply #485 on: October 23, 2020, 03:40:46 PM »
Mercator's take on SST's.
amsr2, awi dev v103 overlaid onto mercator 0m ocean temperature, sep4-oct22  (7MB)
Ok, from your animation is it correct to say all the regions susceptible to warming from Atlantic Waters are already covered by ice or about to be covered? (exception Barents and Kara but we know those two seas are already lost to climate change)
It seems to me the anomalously warm Laptev and ESS areas are basically over the shallow shelf, which will have a record heat release (consistent with the record heat income from GAAC and the season in general). So no Atlantic warmth here to speak about.

The Laptev sea extent beyond the shelf (and so the heat anomaly, by the way). Intrusion of Atlantic waters are discernible on the salinity maps of the mercator. (P.S. : And acknoledging that the heat and salinity extent beyond the Laptev into the central bassin, even under the sea ice).
Yes, SST is more useful if we also show SSS
amsr2, awi dev v103 overlaid at 75% transparent onto mercator 0m ocean salinity, sep4-oct22  (7MB)
mercator label is just visible.
edit: It has often been proposed that the Atlantic waters would meet the Pacific incoming at some point. The delay in refreeze and perhaps some mixing from recent strong winds would seem to make that prospect more likely (according to the model). The Siberian shelf and the Chukchi plateau help to prevent it. (Shown on the 92m salinity map upthread)
« Last Edit: October 23, 2020, 04:45:35 PM by uniquorn »

kassy

  • Young ice
  • Posts: 2614
    • View Profile
  • Liked: 1207
  • Likes Given: 1075
Re: The 2020/2021 freezing season
« Reply #486 on: October 23, 2020, 04:53:09 PM »
Have you seen the snow extent growth over Siberia and Western Canada? Anything to comment or you only rejoice on the bad feedbacks?
...
apart from this, the weather is about to become propitious for a quick rebound of sea ice next week.

This seems like an emotional type response, especially as a reply to such an informative post which did not actually feature any rejoicing...

The previous posts were an actual attempt at figuring out some of the extent of the damage done.

A quick rebound next week would already be late but both recent comments in this thread and in the SIA&E thread hint that the it might not be quick.

You are thinking too much about the area/extent (so 2D) while ignoring the 3D problems like the stall in TPD.

Quote
In other words, 2021 and 2022 could be rebound years as probable as big melt years. I don’t see the doom scenario here (more in line with the scientific consensus of 2040+)

In case you missed it we are discussing what we see. Oh and i think that scientific consensus might have shifted a bit...
Þetta minnismerki er til vitnis um að við vitum hvað er að gerast og hvað þarf að gera. Aðeins þú veist hvort við gerðum eitthvað.

Shared Humanity

  • Frazil ice
  • Posts: 245
    • View Profile
  • Liked: 149
  • Likes Given: 9
Re: The 2020/2021 freezing season
« Reply #487 on: October 23, 2020, 05:04:29 PM »
I prefer looking at 2m temperatures instead of anomalies when trying to determine when freezing might occur. Much of the Laptev and ESS are no more than -4C. With large areas of water at 1C, we're going to need to wait to see dramatic freeze.

Shared Humanity

  • Frazil ice
  • Posts: 245
    • View Profile
  • Liked: 149
  • Likes Given: 9
Re: The 2020/2021 freezing season
« Reply #488 on: October 23, 2020, 05:12:24 PM »
The entire Arctic Ocean is cloud covered. Is this slowing down the release of heat?

SimonF92

  • Frazil ice
  • Posts: 331
    • View Profile
  • Liked: 142
  • Likes Given: 69
Re: The 2020/2021 freezing season
« Reply #489 on: October 23, 2020, 05:24:11 PM »
My apologies if this is an ignorant or rookie question, but has anyone looked at the integral of sea ice extent, area and volume across a given year, or year-to-date, in order to quantitatively compare them with regard to sea ice coverage?

I think that would highlight 2020's prominance more clearly over past years, even 2012, with regard to the amount of time its been in the lead and by how much. This comparison would give equal measure to all time points (as I think they 'deserve') rather than having bias as to the minimum and maximum. Such an analysis may help to see an overall trend over the years, relative to looking at particular time-points.

I think volume may be the most sensitive metric to use, as sea ice area and extent seem to have much less variation during the winter, and those large numbers may blunt changes occurring from year-to-year.

I think this is a really good idea, would you be willing to have a go at it? Otherwise I can try to pick it up
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic

Freegrass

  • Nilas ice
  • Posts: 1641
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: The 2020/2021 freezing season
« Reply #490 on: October 23, 2020, 05:35:08 PM »
Latest Five Day Forecast
Wind + Temp @ Surface
Large GiF!

More heat and moisture coming in with a new storm from Europe this time...
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

A-Team

  • Young ice
  • Posts: 2740
    • View Profile
  • Liked: 685
  • Likes Given: 35
Re: The 2020/2021 freezing season
« Reply #491 on: October 23, 2020, 05:43:08 PM »
Quote
weather is about to become propitious for a quick rebound of sea ice next week.
The time series below looks at the rate of change of SST over the last 16 weeks at 5 day intervals at a single central Laptev site, lat 77.77º. The open water surface temperature peaked at an astonishing 7.3ºC on Aug 17th.

There do not seem to have been any profiler buoys in the area so no data on how this temperature changed with depth to shelf. For shallow water and stratification lost to atlantification, the water may be fairly well mixed and the temperature profile flat.

It would have been better to take a 'reading' every day rather than every fifth, then repeating at say a dozen different locations by loading a list of manufactured urls as browser tabs, then turning the data over to Geronto for graphing with some kind of multi-regressional best fit. 

Another way to go is with a quantitative CMEMS grayscale for a 2D time series of the whole Siberian side. These can literally be subtracted to get at numerical temperature change between any two dates, put into a heat map palette, and averaged over any polygonal region. However CMEMS is very taken with politically correct (perceptually uniform) palettes; the grayscale may or may not be affected.

The series shown went from to 6.9ºC to 1.0ºC in 65 days, cooling roughly 0.1º per day. This could be quite misleading from a thermodynamic standpoint if vertical turbulence or horizontal currents shifted water masses about. Really severe and persistent cold air could also set in, contrary to forecast and past climatology.

However, if the nominal trend continues, the water at this lat/lon would reach -1.8º in the last week of November.
« Last Edit: October 23, 2020, 07:28:05 PM by A-Team »

FishOutofWater

  • Grease ice
  • Posts: 875
    • View Profile
  • Liked: 501
  • Likes Given: 179
Re: The 2020/2021 freezing season
« Reply #492 on: October 23, 2020, 05:52:57 PM »
Warm water in the Barents and Kara seas is known to enhance Urals ridging, inducing a downstream trough in central Siberia. Judah Cohen discussed this issue in his latest blog post. This pattern causes snowfall extent to grow rapidly in October in Siberia and may cause a warm Arctic cold Siberia pattern to develop as fall progresses into winter. Moreover, a warm Arctic cold continents pattern may develop where a cold pool develops in central Canada mirroring the cold in central Siberia. That appears to be what's beginning now.

However, as winter progresses, this pattern enhances upward wave activity into the stratosphere. The weak polar vortex that develops in mid-winter may suddenly break down in the stratosphere causing a major stratospheric warming. Sudden stratospheric warmings may lead to subsidence and high pressure over the pole in February or March, which could be favorable to ice thickening. However, it could also lead to a rapid transition to spring warmth in the Arctic. These are tings we have seen in the past decade that could be exaggerated this winter by the warmth that will persist on the Atlantic side of the Arctic because of the high heat content of the Barents sea waters.

It's going to be an interesting winter.

Freegrass

  • Nilas ice
  • Posts: 1641
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: The 2020/2021 freezing season
« Reply #493 on: October 23, 2020, 05:53:33 PM »
Quote
temperature labels would be useful on the above graphic
The purple is -1.8ºC under the ice pack and it goes up from there in increments of 0.4 or so. This was taken from nullschool Ocean Currents @ Surface for 2020-10-22 12:00 UTC; there is no forecast capacity. I don't know how accurate that data is nor if it improves on or agrees with other online SST product sources. All of them show the fringing around the ice pack periphery, the disconnect of the Chukchi-Bering inflow, and overall commonality (with mild pocketing) of the Laptev-ESS.

It's quite difficult to work with the low contrasts in the dark temperature palette nullschool provides; the image is best converted to '8-bit color' in ImageJ which allows palette replacement, done above with 'glasbey inverted' which by design makes consecutive colors as different as possible. Those bands could be color-picked and replaced with a conventional color gradient palette, worthwhile on a monthly time series.

It is perhaps better for people to just mouse around on that nullschool page, a click puts a green circle down and provides temperature data at that lat/lon. CMEMS, linked above, has better tools that allow temperature transects along lines and averages in polygonal box to be displayed. It may be more accurate; alternatively nullschool has just assimilated it (ie Mercator Ocean).

If we could locate the original infrared satellite image, there would still be the problem of  calibration algorithm. It's not clear how many independent determinations of SST exist. Forecasting products imply additional use of a model of which there are several. It's an interesting question how rapidly SST can change given the thermal inertia of sea water and how that is affected by wind and cold air temperature.

In the past, we've mostly just followed year-on-year SST atlantification at various depths at Mercator Ocean. It's worth delving in deeper to daily SST details this fall because of the extraordinary delay in Siberian refreeze.
I think Nullschool is using publicly available data to render these maps, so wouldn't it be easier to cut out Nullschool completely and make something yourself directly from the available data?
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

Shared Humanity

  • Frazil ice
  • Posts: 245
    • View Profile
  • Liked: 149
  • Likes Given: 9
Re: The 2020/2021 freezing season
« Reply #494 on: October 23, 2020, 05:54:03 PM »
If the nominal trend continues, the water at this lat/lon would reach -1.8º in the last week of November.

Shit. :-[

SimonF92

  • Frazil ice
  • Posts: 331
    • View Profile
  • Liked: 142
  • Likes Given: 69
Re: The 2020/2021 freezing season
« Reply #495 on: October 23, 2020, 06:33:43 PM »
Quote
temperature labels would be useful on the above graphic
The purple is -1.8ºC under the ice pack and it goes up from there in increments of 0.4 or so. This was taken from nullschool Ocean Currents @ Surface for 2020-10-22 12:00 UTC; there is no forecast capacity. I don't know how accurate that data is nor if it improves on or agrees with other online SST product sources. All of them show the fringing around the ice pack periphery, the disconnect of the Chukchi-Bering inflow, and overall commonality (with mild pocketing) of the Laptev-ESS.

It's quite difficult to work with the low contrasts in the dark temperature palette nullschool provides; the image is best converted to '8-bit color' in ImageJ which allows palette replacement, done above with 'glasbey inverted' which by design makes consecutive colors as different as possible. Those bands could be color-picked and replaced with a conventional color gradient palette, worthwhile on a monthly time series.

It is perhaps better for people to just mouse around on that nullschool page, a click puts a green circle down and provides temperature data at that lat/lon. CMEMS, linked above, has better tools that allow temperature transects along lines and averages in polygonal box to be displayed. It may be more accurate; alternatively nullschool has just assimilated it (ie Mercator Ocean).

If we could locate the original infrared satellite image, there would still be the problem of  calibration algorithm. It's not clear how many independent determinations of SST exist. Forecasting products imply additional use of a model of which there are several. It's an interesting question how rapidly SST can change given the thermal inertia of sea water and how that is affected by wind and cold air temperature.

In the past, we've mostly just followed year-on-year SST atlantification at various depths at Mercator Ocean. It's worth delving in deeper to daily SST details this fall because of the extraordinary delay in Siberian refreeze.
I think Nullschool is using publicly available data to render these maps, so wouldn't it be easier to cut out Nullschool completely and make something yourself directly from the available data?

The data is here

https://resources.marine.copernicus.eu/?option=com_csw&view=details&product_id=GLOBAL_ANALYSIS_FORECAST_PHY_001_024

and is in NetCDF format.

Maybe better to continue this is the Dev corner
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic

oren

  • Moderator
  • First-year ice
  • Posts: 6452
    • View Profile
  • Liked: 2394
  • Likes Given: 2052
Re: The 2020/2021 freezing season
« Reply #496 on: October 23, 2020, 06:43:19 PM »
Yes, SST is more useful if we also show SSS
amsr2, awi dev v103 overlaid at 75% transparent onto mercator 0m ocean salinity, sep4-oct22 
Thanks for this uniquorn.
Looking at the animation, it would appear SSS is even more predictive of where refreeze might hit.

In general, I want to commend the very high level of discussion and the collection of excellent posts by several contributors. I can barely keep up... but this exceptional "freezing" season deserves all the analysis it can get.

oren

  • Moderator
  • First-year ice
  • Posts: 6452
    • View Profile
  • Liked: 2394
  • Likes Given: 2052
Re: The 2020/2021 freezing season
« Reply #497 on: October 23, 2020, 06:47:04 PM »
My apologies if this is an ignorant or rookie question, but has anyone looked at the integral of sea ice extent, area and volume across a given year, or year-to-date, in order to quantitatively compare them with regard to sea ice coverage?

I think that would highlight 2020's prominance more clearly over past years, even 2012, with regard to the amount of time its been in the lead and by how much. This comparison would give equal measure to all time points (as I think they 'deserve') rather than having bias as to the minimum and maximum. Such an analysis may help to see an overall trend over the years, relative to looking at particular time-points.

I think volume may be the most sensitive metric to use, as sea ice area and extent seem to have much less variation during the winter, and those large numbers may blunt changes occurring from year-to-year.

I think this is a really good idea, would you be willing to have a go at it? Otherwise I can try to pick it up
If I understand correctly, this would partly be covered by Gero's 365-day extent average, posted regularly in its own thread. As for volume I seem to recall he also made such a chart, and there's also my own modest chart detailing the number of daily records held by each year. For all measures, 2020 is still behind a combination of 2016-17, however this could change quite quickly as the "Siberian Seas problem" continues to develop and considering the low minima achieved this year.

gandul

  • Grease ice
  • Posts: 769
    • View Profile
  • Liked: 184
  • Likes Given: 10
Re: The 2020/2021 freezing season
« Reply #498 on: October 23, 2020, 06:49:30 PM »
If the nominal trend continues, the water at this lat/lon would reach -1.8º in the last week of November.

Shit. :-[

Good, we’ll see.
I am more worried about the Chukchi sea refreeze, I believe the Bering strait may have another “blue” Christmas in what is becoming more frequent later refreeze.

gandul

  • Grease ice
  • Posts: 769
    • View Profile
  • Liked: 184
  • Likes Given: 10
Re: The 2020/2021 freezing season
« Reply #499 on: October 23, 2020, 06:57:23 PM »
Have you seen the snow extent growth over Siberia and Western Canada? Anything to comment or you only rejoice on the bad feedbacks?
...
apart from this, the weather is about to become propitious for a quick rebound of sea ice next week.

This seems like an emotional type response, especially as a reply to such an informative post which did not actually feature any rejoicing...

The previous posts were an actual attempt at figuring out some of the extent of the damage done.

A quick rebound next week would already be late but both recent comments in this thread and in the SIA&E thread hint that the it might not be quick.

You are thinking too much about the area/extent (so 2D) while ignoring the 3D problems like the stall in TPD.

Quote
In other words, 2021 and 2022 could be rebound years as probable as big melt years. I don’t see the doom scenario here (more in line with the scientific consensus of 2040+)

In case you missed it we are discussing what we see. Oh and i think that scientific consensus might have shifted a bit...
Yes Kassy it’s an emotional response because I get the feeling people choose to not talk about that, there may be other mechanisms that kick in and explain why the Arctic didn’t completely tank 8 years ago.

Rebound years (08/09, 13/14, 17/18) are surrounded by certain mystery or silence.

Ok I’ll try to open that discussion in another thread.
In any case the posts here lately are excellent, I don’t deny people are well informed, much better than me in any case.