The 2020 melting season

[Rubikscube]

How did the 2012 GAC look on false color as a comparison?  I am a relative newby and did not start following the melt season till a couple of years later.  Is that available to look up on the Bremen site?

You can find the live coverage of the GAC 2012 in the archives of Neven's blog. There are several Uni-Bremen false color gifs for comparison there. Ice concentration was notably lower in areas affected by that storm, which I believe make a lot of difference to the outcome. Powerful August storms in 2013, 2014 impacted high ice concentration areas and had much the opposite effect of GAC-2012. It is interesting to see what the impact will be of this storm.

[icy voyeur2]

Is it just me or do others get the shudders when they see that Will elf? Sorry, OT.

I'm irrelevant but I'm afraid of the consequences of blowing so much rubble into warm seas. Don't understand the roles of elves beyond their generally chaotic nature.

[Gizmo]

I don't think any of us is irrelevant.  And your concern is well founded.   

lots of very thin ice, much of it just rubble according to our eyes in the sky.  And, depending on which thickness map you look at the thick multi-year ice is pretty much gone.

Lots of warm water.

And August doesn't start until tomorrow.

I'm new here but I've been following the ice...pretty much since 2012, when it really got my attention. (thanks Neven)   And I'm starting to see why Friv' is so.... passionate.

[Pagophilus]

An update on the storm and concentration comparison. Higher res version on twitter.
https://twitter.com/Icy_Samuel/status/1288708042951929857

(click to play)

Startling, really, BftV.  Thanks for bringing this here.  I think it is worth adding the term 'apparent' concentration, as Samuel Hayes did.   Apart from the dramatic 'storm fingerprint' phenomenon, for me it illustrates how ice wetness/other conditions can produce seeming changes in concentration, only for that 'concentration variation' to vanish almost instantly. 

Often the transient 'concentration changes' get over-interpreted.  This helps me get better at reading this stuff.

[aslan]

It is also worth underligning that IFS is trying to deepen again the low runs over runs, with a new minima developing within the front and warm air over canadian archipelago. As a consequence, IFS go down to 991 hPa at 90 hours (sunday evening UTC), and with a high above 1025 hPa for the siberian coast, this means a continuous wild ride for sea ice. GFS is not following, but probable answer is that GFS is lost in the wildnerss of fantasy land, as usual. Even more as others models also are trying something within this front, even tough they don't go as low as the euro guy. And again, follow the trend  IFS is loosing hPa by hPa run over run, and it is quite likely at this point that the trend will continue and the low will end in the 980 something hPa. And again, the difference with 2012 is the strong high over siberian coast. Even if we only bottom out to 985 hPa Sunday evening (UTC), gradient will be ~ 50 hPa beetwen Beaufort Sea and siberian coast. At the height of the GAC 2012, it was only ~ 55 hPa over about the same distance...

[Pagophilus]

Some updates:
* OSI SAF ice drift shows CAB ice racing towards the now empty Laptev and ESS. Some of the Beaufort ice is pushed back into the CAB, and some is dispersed south.
All this moving around can't be good for the ice and surely increases bottom melt.
* UH AMSR2 area of the four regions expected to participate in the minimum. CAB and CAA are tracking low, Beaufort tracking high but a cyclone-related blip is worrying. Greenland Sea tracking high but with export halted, area should drop sharply soon.
* CAA max temps as measured in various ground stations have gone down a notch since a few days ago, but are still warm enough for continued vigorous melt.

Thank you for the extensive expert summary, oren.  Very useful to this amateur.

[uniquorn]

A quick view, after the worst part of the storm...

uni-hamburg version, amsr2-uhh, beaufort-chukchi, jul18-29 
click

[oren]

As it appears (to my non-expert eyes) that clouds cause artifacts in the concentration products, the big question in my mind is whether they cause high-concentration streaks, or whether they cause low concentration streaks... Can some kind soul superimpose (and synchronize) the cyclone and the concentration images and prove it one way or the other?
p.s. Of course it's also possible the storm causes actual concentration changes that appear to be in its shape. Would love an expert opinion.

[Niall Dollard]

How did the 2012 GAC look on false color as a comparison?  I am a relative newby and did not start following the melt season till a couple of years later.  Is that available to look up on the Bremen site?

You can find the live coverage of the GAC 2012 in the archives of Neven's blog. There are several Uni-Bremen false color gifs for comparison there. Ice concentration was notably lower in areas affected by that storm, which I believe make a lot of difference to the outcome. Powerful August storms in 2013, 2014 impacted high ice concentration areas and had much the opposite effect of GAC-2012. It is interesting to see what the impact will be of this storm.

Here is the direct link to Bremen 2012 archives for August 2012.

https://seaice.uni-bremen.de/data/oldversions/amsr2/asi_daygrid_swath/n6250/2012/aug/Arctic/


Sample image attached for august 4th 2012. 

[Killian]

Prediction:

2020 crosses 2012 by August 10th, most likely between Aug. 6th and Aug 8th.

2020 crosses 2019 sometime between Aug 9th and Aug 13th, then crosses back into 2nd lowest territory between Aug 13th and Aug 21st.

Rationale: The compacting of the ice makes it difficult for extent to change other than at or slower than averages, though area may fall significantly, particularly if extent increases. However, 2019's curve flattens over those time periods, also.

We knew the low extent numbers would allow weird things to happen. The ice-saving compaction is one of them: Put stuff in a gyre and it accumulates to the center, creating a negative feedback for melt rates.

Nothing you all don't know, just me playing with trends and patterns and putting myself on the line because, why not?

Anyone care to join in?

[uniquorn]

superimpose (and synchronize) the cyclone and the concentration images

I think amsr2 is representing the state of the ice quite accurately in this case.
click

[JayW]

As it appears (to my non-expert eyes) that clouds cause artifacts in the concentration products, the big question in my mind is whether they cause high-concentration streaks, or whether they cause low concentration streaks... Can some kind soul superimpose (and synchronize) the cyclone and the concentration images and prove it one way or the other?
p.s. Of course it's also possible the storm causes actual concentration changes that appear to be in its shape. Would love an expert opinion.

I'm not an expert, but I have seen enough artifacts caused by clouds and rain to agree, they do confuse the sensors.  I'd also add that like worldview, these concentration/area products are stitching together multiple satellite passes over the course of 24 hours.  When sea ice is moving around, I wonder how much this motion also plays into the final product. In my humble opinion, these figures should include error bounds to describe the inherent uncertainty in the data being displayed.   

[blumenkraft]

I'm not an expert, but I have seen enough artifacts caused by clouds and rain to agree, they do confuse the sensors.

+1

I've seen it too often by now to believe in a coincidence.

[oren]

superimpose (and synchronize) the cyclone and the concentration images

I think amsr2 is representing the state of the ice quite accurately in this case.
click

Thanks for this uniquorn, amazing imaging capabilities as usual.
My eyes insist that under the thick cloud bands (in Worldview) AMSR2 concentration is higher than elsewhere, thus my conclusion/suspicion that if anything, current concentration might be jacked up by such artifacts, rather than jacked down.

[bluice]

Prediction:

2020 crosses 2012 by August 10th, most likely between Aug. 6th and Aug 8th.

2020 crosses 2019 sometime between Aug 9th and Aug 13th, then crosses back into 2nd lowest territory between Aug 13th and Aug 21st.

Rationale: The compacting of the ice makes it difficult for extent to change other than at or slower than averages, though area may fall significantly, particularly if extent increases. However, 2019's curve flattens over those time periods, also.

We knew the low extent numbers would allow weird things to happen. The ice-saving compaction is one of them: Put stuff in a gyre and it accumulates to the center, creating a negative feedback for melt rates.

Nothing you all don't know, just me playing with trends and patterns and putting myself on the line because, why not?

Anyone care to join in?

My 2 cents: maybe the compacting movement didn’t save the ice, but instead acted as a cover for the melt within the CAB during the GAAC?

As the pack is now dispersing we get below average extent loss. But we also get very rapid area loss which hints a lot of ice has been lost. This will likely show on the end of July Piomas.

I’m possibly/probably wrong about the above but one thing I’m certain of: August and September extent numbers will be decided by the weather during the rest of the season.

[JayW]

A look at ice motion under the center of the storm.
Latitude line is 77.5°N
Needs click.

[Thawing Thunder]

Fascinating Jay. I've seen the ice being scattered around many times in the last years, but this is one of the worst events. But this broken up ice has proven to be resilient. There are few weeks left for the ice to melt under normal conditions, and we might end up with average extent and area numbers. Volumen is another thing, though. Or it could all go poof in the very last moment. But I think that is not very probable, I rather expect some damaged areas.

But whatever the outcome might be, the gravest damage is already partly avoided: The open waters would not serve as giant collectors of solar energy, as strong insolation is already coming to an end in August. The real damage will happen once there is too much open water in July to revert the energy input in the winter and in the following year. Then all hell might break loose.

[BornFromTheVoid]

As it appears (to my non-expert eyes) that clouds cause artifacts in the concentration products, the big question in my mind is whether they cause high-concentration streaks, or whether they cause low concentration streaks... Can some kind soul superimpose (and synchronize) the cyclone and the concentration images and prove it one way or the other?
p.s. Of course it's also possible the storm causes actual concentration changes that appear to be in its shape. Would love an expert opinion.

Another version, to add to uniquorns great work.

(Large file warning!)

[binntho]

But whatever the outcome might be, the gravest damage is already partly avoided: The open waters would not serve as giant collectors of solar energy, as strong insolation is already coming to an end in August.

I think you are wrong - the storm now is probably the worst that could happen, and if more storms arrive in the next weeks the situation will be even worse.

Insolation is falling very rapidly, albeit still strong, but the current cloudiness in the Arctic is probably working as an amplifier - the thin, practically see-through, clouds we have now are great for reflecting infrared radiation back to the surface, while letting a surprising amount of sunlight through.

Dispersion of the current icepack through repeated storms with light clouds inbetween is probablyby far the most efficient method of melting in the second half of the melting season. And one should avoid the trap of judging the health of the ice from above. We have no way of "seeing" thickness, and much of the dispersed ice is probably only half a meter or less. With melt from above and particularly bottom melt easily ranging in several cms per day, the rest of the ice could very well melt.

Which is not to say that a new minmin will be reached - 2012 had strong compaction towards the end, which can easily subtract 100k or more from extent.

[uniquorn]

My eyes insist that under the thick cloud bands (in Worldview) AMSR2 concentration is higher than elsewhere, thus my conclusion/suspicion that if anything, current concentration might be jacked up by such artifacts, rather than jacked down.

Agreed. amsr2 concentration is often represented as higher under cloud, so higher concentration streaks. It's the low radials that are the worry.
Nice images BFTV.

[Thawing Thunder]

Binntho, I agree with you. The ice looks muddy and dispersed up to the pole and what you describe could well happen. If one zooms in on World View, Ice looks really bad and one can easily judge by this a certain lack of thickness.

I also stick to my prediction of a second-place behind 2012, due to similar conclusions like yours.

But what I was referring to in my anterior post was not the actual melt, but the long term effect if the waters would open up much earlier (and yes, 2020 was the worst ice in July ever seen). The water under the ice is capable to store solar energy for a long time, some of it won't be released into space in the dark winter months. So the energy balance would be entirely altered on a global scale with an early open ocean at or near the pole.

[Freegrass]

Latest Five Day Forecast
Wind @ Surface
Wind + Temp @ 850hPa
Wind @ 250hPa
Large GiFS!

I like this combination. You can use all the levels to see how that new storm is being created picked up by the jetstream while the old one moves south quickly.

Edit: Can you tell us what's happening Friv? And have you seen the long term GFS? It looks like we could be in for another big one.

[glennbuck]

Arctic sea ice area for July 29th,  3,796.435 km^2. NSIDC Daily Area.

lowests minimum: 2.241 (2012), 2.477 (2016)

Graph by Nico Sun, https://cryospherecomputing.tk/

This will be posted in the off topic thread in future as people seem to get upset about this posted in the melt season thread daily.

[D-Penguin]

An update on the storm and concentration comparison. Higher res version on twitter.
https://twitter.com/Icy_Samuel/status/1288708042951929857

(click to play)

Thanks for the posting BFTV. The graphic shows very clearly how, by the 29 July, the storm had driven the higher ice concentration of the CAB into that crescent shaped area of lower ice concentration that had stretched across the entire CAB for the previous two days.

I do not know if my logic on this is right but it suggests to me that the arc of lower concentration would have consisted of 'broken/weak' ice, although now more compact; could it have created a 'fault line' traversing the CAB ice? Should the appropriate weather conditions prevail could the CAB ice separate with the 'broken/weak' ice melting out?

It would obviously be an extraordinary new feature of the 'melt' season with the CAB melting from the inside out!

Perhaps an idea that is too fanciful

[ArcticMelt2]

According maps a lot of ice drifts from the Central Arctic to the Laptev and East Siberian. But the Extent is still decreasing there.

[bbr2315]

If DMI is correct it looks like everything east of the prime meridian and west of the international dateline is imminently going to collapse into open floes & residual pressure ridges.

You can also see the hunk of MYI on the ATL front now bifurcating from the main pack as it is heading for export / destruction. Wonder if any bits of it will survive.

[Pagophilus]

But whatever the outcome might be, the gravest damage is already partly avoided: The open waters would not serve as giant collectors of solar energy, as strong insolation is already coming to an end in August.

I think you are wrong - the storm now is probably the worst that could happen, and if more storms arrive in the next weeks the situation will be even worse.

Insolation is falling very rapidly, albeit still strong, but the current cloudiness in the Arctic is probably working as an amplifier - the thin, practically see-through, clouds we have now are great for reflecting infrared radiation back to the surface, while letting a surprising amount of sunlight through.

Dispersion of the current icepack through repeated storms with light clouds inbetween is probablyby far the most efficient method of melting in the second half of the melting season. And one should avoid the trap of judging the health of the ice from above. We have no way of "seeing" thickness, and much of the dispersed ice is probably only half a meter or less. With melt from above and particularly bottom melt easily ranging in several cms per day, the rest of the ice could very well melt.

Which is not to say that a new minmin will be reached - 2012 had strong compaction towards the end, which can easily subtract 100k or more from extent.

Strongly agree, binntho.  So much heat is in the surrounding waters.  Dispersion of ice can only bring ice in contact with that heat.  And now is the time of bottom melt.  I would add that the situation is not like a 'normal' year, as the Laptev has become extraordinarily warm.  The thin clouds reminder is useful to me.

I did not know that the ice had been compacted at very the end of the 2012 season, bringing down extent further.  In a way, it makes 2012 a bit less extraordinary in terms of melt to me.

[weatherdude88]

High resolution extent is no longer the lowest on record for the date. 2020 is headed in the direction with the rest of the pack.

[Pagophilus]

NSIDC Total Area as at 29-Jul-2020 (5 day trailing average) 3,999,723 KM2         

Total Area           3,999,723    km2       -674,231    km2   <   2010's average. -371,569    km2   <   2019 -1,481,182    km2   <   2000's average.           Total Change   -106    k   loss Peripheral Seas   -17    k   loss Central Seas___   -90    k   loss           Peripheral Seas          Okhotsk______    0    k   gain Bering _______    0    k   gain Hudson Bay___   -2    k   loss Baffin  Bay____   -2    k   loss St Lawrence___    0    k   gain Greenland____   -12    k   loss Barents ______   -1    k   loss           Central Arctic  Ocean Seas          Chukchi______   -12    k   loss Beaufort_____   -8    k   loss CAA_________   -13    k   loss East Siberian__   -7    k   loss Central Arctic_   -50    k   loss Laptev_______    2    k   gain Kara_________   -2    k   loss

         
Sea ice area loss on this day 106 k, 59 k more than the 2010's average loss of 47 k         
         
- 2020 area is at position #1 in the satellite record.         
- 2020 Area is 674 k less than the 2010's average         
- 2020 Area is 1,481 k less than the 2000's average         
- 2020 Area is 480 k less than 2016         
- 2020 Area is 372 k less than 2019          
- 2020 Area is 289 k less than 2012         

Daily sea ice Area losses have been above 100k for 4 days in a row.
Total Area has been lowest in the satellite record for 7 days in a row.
Can't just be the sensors getting fooled, especially as the data is 5 day trailing average, which tends to smooth out oddities.

Posting here so blumenkraft doesn't (justifiably) yell at me on the area and extent thread.   
Thanks Gerontocrat, as ever.  (my red bolding, btw, G'crat is not that vulgar)
Remarkable area losses over the past few days.
That huge area loss from the Central Arctic -- is that mainly the result of losses on the icepack's boundary with the Atlantic?  There is a lot of melting there, I know.  Is the ice also being pushed back by the recent constant southerly winds on that Atlantic side? 

[weatherdude88]

Comparing the last 2 days of AMSR2 sea ice concentrations, we can already see the start of a correction in the Central Arctic and southwestern Beaufort.

I suspect we will continue to see corrections from sensor errors, that are currently showing faux reductions of sea ice concentrations. This will be obvious in the Central Arctic and Beaufort over the next several days.

[ArcticMelt2]

Records for maximum temperature in the Canadian Arctic have made the region the third worst in the Arctic.

https://twitter.com/AlaskaWx/status/1288512376363339777

July 28 regional ranks of #seaice extent around the #Arctic Basin in @NSIDC data: 1=lowest since 1979, 42=highest since 1979. Biggest change past two weeks in the Canadian Archipelago (now lowest extent past 42 years) and Central Arctic (only 2013 lower).

[Tigertown]

Enlarged view of the Beaufort+
Broke up into two shots to see better detail
Click to enlarge

[ajouis]

Comparing the last 2 days of AMSR2 sea ice concentrations, we can already see the start of a correction in the Central Arctic and southwestern Beaufort.

I suspect we will continue to see corrections from sensor errors, that are currently showing faux reductions of sea ice concentrations. This will be obvious in the Central Arctic and Beaufort over the next several days.

It is not faux reduction, it is a known effect subsequent to meltponding, similarly the concentration is artificially inflated where the low is compared to reality due to clouds, sensor error makes it seem like a one time thing, which is being adjusted back, it is only the long acknowledged response to certain phenomenons.
Edit added high res amsr2 on the Beaufort you can see the preliminary results of the storm and I think even the clouds that blocked concentration in the real colour image

[ajouis]

I think given that the boundaries still have high pressure (chuchki and eastern cab), that there will be some rain in the low pressures, that the temperatures at the surface are globally still positive and therefore conducing to top melt, especially with the waa north of Greenland and Fram, with continued pronounced positive temperature anomalies in the Siberia to Greenland front, all for the next five days at least, it is not unreasonable to predict that the cab area nosedive will probably continue at the present pace (50 k) for at least a week, putting us in a very good position to beat the record low and affecting both concentration and ice pack integrity (compaction is already back to very low levels), and probably the piomas numbers too as they take the area numbers into account. The run to the pole before refreeze is on

[BornFromTheVoid]

Some of the concentration reductions are very real, some aren't. How much is real won't be apparent until we get more clear views.
Por ejemplo...

[blumenkraft]

Sea surface temperature anomalies. Holy macaroni...

[Paul]

From what you can see through the clouds,  situated on the Beaufort Chukchi border really does look very diffused and it days does look more limited. The ice nearer to the Alaskan coastline looks less diffused and may last a bit longer. Traditionally not alot of ice survives in the Beaufort sea so I'll be surprised if this year is an exception.

In one way its good the ice has slowed down but not slowed down toko quickly as that may suggest ice spread is not too widespread but the forecast still favours ice spreading out. Still odds on for extent to be under 4 million though but I'm still not yet convinced on record lows unless something extraordinary happens during August.

[gerontocrat]

Beaufort and Central Arctic Sea

With 47 days to go to mid-September  my money is still on "where area leads extent will follow" even if by not so much, certainly in the Central Arctic Sea..

Extent & Area Graphs attached (NSIDC 5-day trailing average)

[Freegrass]

Latest Five Day Forecast
Wind @ Surface + Total Precipitable Water
Large GiF!

[D-Penguin]

Some of the concentration reductions are very real, some aren't. How much is real won't be apparent until we get more clear views.
Por ejemplo...

These comparative graphics are interesting and informative, particularly at this time with fast moving events.
+1

[pearscot]

I still don't know how to feel about this melt season...there's too much at play but the storm has really dispersed the ice. I did a few quick measurements and the gap between the ice and land in northern Greenland is almost 30 miles(!) in some sports. I'm interested to see if the Atlantic will affect it or if the gyre rotation will really begin to melt a lot of ice in the Laptev/Eastern Siberian Sea.

Anyways, the recent storm must have caused some fairly significant beach erosion...just look at all of the new/replacement white bags used to bolster the dirt sea wall...

[D-Penguin]

Latest Five Day Forecast
Wind @ Surface + Total Precipitable Water
Large GiF!

Another great graphic. Thank you Freegrass.

The CAA looks like it will be very active from 02 Aug!
+1

[Pagophilus]

Sea surface temperature anomalies. Holy macaroni...

I went to Nullschool to find the map of today's SSTAs there to back up your better map above, blumenthal.   

I got this Eye of Sauron type image, which is maybe apropos, given the circumstances.
 
Not the usual color spectrum scale -- yellow and light yellow are the most anomalously warm.

[KenB]

Latest Five Day Forecast
Wind @ Surface + Total Precipitable Water

Snow over Greenland and the CAB, I'm guessing.  Rain over the CAA, though?

[uniquorn]

A casual look at the Lincoln Sea on worldview doesn't show much sign of melt but there is some kind of attrition going on. https://go.nasa.gov/33993H1, jul23-30

[jdallen]

Prediction:

2020 crosses 2012 by August 10th, most likely between Aug. 6th and Aug 8th.

2020 crosses 2019 sometime between Aug 9th and Aug 13th, then crosses back into 2nd lowest territory between Aug 13th and Aug 21st.

<snippage>

Extent *may* cross over 2012.  Not convinced it will pass back over 2019.

The area losses are continuing un-abated, and I suspect will continue in much the same vein, as the CAA and Greenland seas start being hit hard, and as ice exported into the Atlantic "killing zones"  disappears.

Extent will start failing  as a result of that, as dispersed ice suddenly drops under the concentration thresholds over wide areas.

*Everything* is melting fast, and as a lot of that is "in situ", I think it will artificially prop up extent numbers.

Looking at the Bremen concentration map and comparing it to that posted on Aug 4th of 2012, 2020's concentration really doesn't look a lot better, with vast areas of the Pacific side - at lower latitudes - dropping under 70% concentration.

Meanwhile, the Laptev/ESS margins will reverse their retreat and return to their implacable march towards the pole, with the Barents and Kara margins closing on the flanks of the CAB as well.

[jdallen]

A casual look at the Lincoln Sea on worldview doesn't show much sign of melt but there is some kind of attrition going on. https://go.nasa.gov/33993H1, jul23-30

Disagree! 

I think it's showing a lot of melt.  I'd say lots of thinner "lead" ice is melting out exposing blocks of thicker MYI, which themselves being weakened, are starting to break up.

[UCMiami]

A casual look at the Lincoln Sea on worldview doesn't show much sign of melt but there is some kind of attrition going on. https://go.nasa.gov/33993H1, jul23-30

It's like putting a bunch of friable rocks in a bag and shaking it - they grind each other down into smaller and smaller sizes until you have a bag of sand. Same is happening all through the pack, especially under the current cyclone. A few big flows may resist, but each collision between chunks of ice breaks off a bit more and the smaller bits just disappear without visible signs of melting.

[uniquorn]

I get that, but can you actually see it? or measure it? I picked one floe that was easy. Normally we see 'goodbye waves' but I assume some other process happens in open water in the 'crack'.

edit: ok. It was easy to fine 'glue ice' melting too.

[Michael Hauber]

2020 coming 3rd is definitely still on the table.

Compare 2020 to 2019.  The main area where 2019 has more ice is in the ESS, and this ice area is thinly spread and partially detached from the main area of ice.  For a good while 2019 melts at a rapid rate and keeps within striking distance of 2012.  But then slows down and the gap opens up quickly later in the season.  This happens at about the same time that the thinly spread partially detached area finishes its melt out.