Latest PIOMAS Volume update

[oren]

Looking at the surrounding seas that traditionally participate in the minimum, trends have continued with relatively little changes.
* The CAA continues to slide but is still lagging behind the leading years. With all the fast ice now broken up and after the high temps all summer, I expect a finish between 2019 and 2012, at around 100 km3.
* The Beaufort is still tracking high, but appears to be on the verge of breaking down. With all the mess apparent on Worldview and AMSR2, I don't have much hope for volume losses to slow. I assume a minimum of 100 km3 is reasonable.
* The Greenland Sea is also tracking high but with mounting losses towards the end of the period. With the winds reversed and export halted, it is a miracle that so much ice still remains. Best guess for the minimum is around 150 km3, but variability is very high.

The table lists previous minima with a breakdown of regional details as they were on the day of each respective minimum. An extrapolation of current trends with some extra losses in the CAB will bring 2020 to a 2nd place finish, as shown. An August area crash and accompanying volume crash may be able to clinch a new record low, which I still think is possible, maybe even probable assuming PIOMAS hasn't taken into account the full enormity of the GAAC event. A sharp August slowdown could bring about a 3rd place finish.

[oren]

Completing the picture, the four Inner Basin seas expected to reach zero for this year's minimum. As usual, click all charts and tables for enlarged versions.

[interstitial]

Oren:
Where do you get Piomas data? All I seem to find is references to asif thread.

[oren]

Some people have use for the updated regional data files:

monthly:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional-monthly.txt.gz
daily:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional.txt.gz

This is how I get the data, thanks to the incessant efforts of Wipneus. Otherwise it is possible to download and process the raw PIOMAS files themselves, but this requires software processing and specialized knowledge which I unfortunately lack.

[Wipneus]

PIOMAS gridded thickness data was updated to day 228 (15/16 August). Calculated volume on that day is 5.14 [1000km³], which means a third lowest place before 2012 and 2019.

Here is the animation for August thus far.

[Wipneus]

Updated volume and volume-anomaly graphs.

[Wipneus]

Updated Fram volume export graph.

[Wipneus]

Some people have good use for the updated regional data files.

daily:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional.txt.gz

[Sepp]

Those numbers seem to be quite strange. The modeled thickness assumes partly more than 1m when there is actually open water.

I know PIOMAS relies on extent measurements, but such a divergence from what is even identifiable via Worldview is strange (Beaufort/Chukchi pretty thick, Svalbard still connected to ice, quite thick ice at north coast of Greenland).

[oren]

Remember PIOMAS gives average thickness, not typical thickness, and that it relies on a coarse grid. But still looking at the animation I get the feeling myself that PIOMAS has some overestimates, for example in the Lincoln Sea where the thick ice has moved west towards and around Ellesmere.
Be that as it may, the numbers tell their own story, as that is the only way to compare with years past (where overestimates may have happened as well in some regions).

[Lord M Vader]

Wipneus, what are the numbers for 2012 and 2019?

Sepp: I don't think that number is too strange given the strong polar vortex we had last winter that bottled up cold air in the Arctic. Beside that, check Wipneus maps for July 31 where you can see the difference between the different years for that date.

[Wipneus]

Wipneus, what are the numbers for 2012 and 2019?

4.75  4.88  5.14  5,31 (2012, 2019, 2020, 2017)

[oren]

Thanks again Wipneus for the regional data.
The string of record breaking losses in the CAB has finally come to a halt. CAB loss was 863 km3, 199 km3 less than 2012 and 85 km3 less than 2019.
The Beaufort lost a large amount of volume but is having a very hard time catching up (despite near-record losses in the 2007+ era), and it seems only a massive area loss can hasten this trend in time for the minimum.
The CAA has also seen a slowdown of sorts, due to the refusal of the ice in the main channels to drift away or melt out, except in the PGAS where a garlic press of sorts is underway.
The only region that crashed was the Greenland Sea, as expected with no Fram export to feed it.
All other regions have dropped from 156 km3 in total to 54 km3 in total, and are out of the game.
In general a rather ho-hum update, but the potential is still there IMHO, seeing the bad visible state of the ice, and knowing the amount of energy that went into the system.

[BornFromTheVoid]

Some of my own graphs.
Thanks for your work on this, Wipneus.

[marcel_g]

Those numbers seem to be quite strange. The modeled thickness assumes partly more than 1m when there is actually open water.

I know PIOMAS relies on extent measurements, but such a divergence from what is even identifiable via Worldview is strange (Beaufort/Chukchi pretty thick, Svalbard still connected to ice, quite thick ice at north coast of Greenland).

I'd add that considering how fast the PolarStern sailed straight to the NP, and according to the Piomass map they would've sailed through some 1.5m+ ice, but I'm pretty sure ice that thick would've slowed them down a lot.

[interstitial]

Those numbers seem to be quite strange. The modeled thickness assumes partly more than 1m when there is actually open water.

I know PIOMAS relies on extent measurements, but such a divergence from what is even identifiable via Worldview is strange (Beaufort/Chukchi pretty thick, Svalbard still connected to ice, quite thick ice at north coast of Greenland).

I'd add that considering how fast the PolarStern sailed straight to the NP, and according to the Piomass map they would've sailed through some 1.5m+ ice, but I'm pretty sure ice that thick would've slowed them down a lot.

Hycom shows no more than 0.8 m for that trip.

[Aluminium]

Changes from July 31.

Laptev ... -100%
Barents ... -100%
Hudson ... -100%
Kara ... -91%
Chukchi ... -72%
ESS ... -63%
Greenland ... -61%
Beaufort ... -37%
CAA ... -28%
Baffin ... -22%
CAB ... -16%

[interstitial]

Seeing the relative sizes of what is left is a helpful reminder this time of year. Thanks Aluminium.

[Juan C. García]

Changes from July 31.

Laptev ... -100%
Barents ... -100%
Hudson ... -100%
Kara ... -91%
Chukchi ... -72%
ESS ... -63%
Greenland ... -61%
Beaufort ... -37%
CAA ... -28%
Baffin ... -22%
CAB ... -16%

Great representation of the Arctic, Aluminium!
A question: What is the meaning of the red circle?

[gregcharles]

Changes from July 31.

Laptev ... -100%
Barents ... -100%
Hudson ... -100%
Kara ... -91%
Chukchi ... -72%
ESS ... -63%
Greenland ... -61%
Beaufort ... -37%
CAA ... -28%
Baffin ... -22%
CAB ... -16%

Great representation of the Arctic, Aluminium!
A question: What is the meaning of the red circle?

If I'm reading it right, the red circles are the amount of ice on July 31, and the blue circles on top of the red ones are the amount left now. Kudos to Aluminium! It's a clever way of conveying that data.

[Aluminium]

"Red ice" has melted. "Blue ice" is alive.

[gerontocrat]

Some what late, an analysis of volume in my standard JAXA format.

Volume reduced by above average in the second half of July, the in the first 10 days or so of August daily volume loss was well below average. In the last few days daily volume loss has increased.
This is very much in line with sea ice area losses in this period. It seems a major input into PIOMAS is the NSIDC area data.
Attachment Vol-2.

In mid-August volume was 5.14 thousand km3, third lowest in the satellite record behind 2012 and 2019.
Attachment Vol-2.

Average remaining volume loss results in a minimum of 4.054 thousand Km3, almost a dead heat with 2019 and equal 2nd lowest with 2019. Note that 2019 volume loss was extremely low from mid-August onwards.
Attachment Vol-1.

This is also shown as a plume. Even at this late stage in the season variations in remaining ice volume loss are large.

[gerontocrat]

Some Thickness Graphs = PIOMAS Volume / NSIDC Area

I attach graphs of average thickness for
- all Arctic Seas,
- the 7 seas of the High Arctic,
- the 7 Peripheral seas.

These show
- the spectacular reduction in thickness since the 2000's.
- that in the 2nd week of August average thickness has increased for the Arctic as a whole and in the High Arctic, but reduced in the Peripheral seas.

To note also is that average thickness continues to reduce well after the date of minimum volume. I presume this is because as refreeze starts large areas of thin ice form, and arithmetic dictates that as area increases quickly but volume slowly, average thickness must decline.

I also attach the thickness graph for the Greenland Sea. What happens here is as much to do with  ice imported from the Fram as with melt.

ps:
I am confident in the arithmetic used, and that the data is properly summarised.
But as volume reduces over the years, and the Arctic ice cap decays into a rotten broken up mess, can one have as much confidence in the data?

[mjb]

Hi All,
my first post here.

I think it would be great for many of those nice line graph diagrams
to use the color scheme to express some meaning and not use it just randomly.

My idea is to use a gradiant coding between 2 colors e.g. blue to red ;-)
or any easily visible scheme - to show the time scale.

Like dark blue oldest year .... dark red for last/actual year.
Then the wobbeling lines could even transport some meaning.
Like showing how the trend is moving.

Other option could be rainbow colors ...

[oren]

Welcome mjb.
Chart color scheme is a painful issue. When there are many lines it is very hard to tell which is which. Having a gradual color scheme often makes it worse, because years that are close in time often have similar characteristics, due to the long-term trend of sea ice loss. In my own charts I finally came to the conclusion I had to break the original default color scheme and try to choose distinct colors that are different from their surroundings. I also gave certain years (lowest years, current year) thicker lines.
In some charts where the long term trend is more important than individual years, your suggestion works very well. I've seen some fine examples for both temps and sea ice though I can't recall them specifically right now.

[gerontocrat]

Hi All,
my first post here.

My idea is to use a gradiant coding between 2 colors e.g. blue to red ;-)
or any easily visible scheme - to show the time scale.

Like dark blue oldest year .... dark red for last/actual year.
Then the wobbeling lines could even transport some meaning.
Like showing how the trend is moving.

Other option could be rainbow colors ...

Welcome mjb,

You highlight a problem that vexes us all from time to time. There is no one-size-fits-all fix for this.

Thye problem is that when you use colours with only a small change in shading it is diffult to distinguish between the years, especially when there are many lines on the graph.

At the moment I am developing a single convention for all my graphs - e.g. 2000's average should always be green and decadal averages always solid lines. The current year is always red and solid.

I am also starting to use dotted lines to highlight years of interest. e.g. 2012 always red and should be fine dotted.

But with several hundred graphs on extent, area and volume, I'm afraid my PIOMAS graphs are a work-in-progress.

Gero.

[mjb]

Welcome mjb.
Chart color scheme is a painful issue. When there are many lines it is very hard to tell which is which. Having a gradual color scheme often makes it worse, because years that are close in time often have similar characteristics, due to the long-term trend of sea ice loss. In my own charts I finally came to the conclusion I had to break the original default color scheme and try to choose distinct colors that are different from their surroundings. I also gave certain years (lowest years, current year) thicker lines.
In some charts where the long term trend is more important than individual years, your suggestion works very well. I've seen some fine examples for both temps and sea ice though I can't recall them specifically right now.

my suggestion was not thought of as a replacement, but rather an additional option which would really show the trend more than individual years.

[Wipneus]

PIOMAS has updated the gridded thickness data to day 244 (31 August or first of September). Volume calculated from thickness was 4.31 [1000 km³] which is third lowest for that day (behind 2012 and 2019).

Here is the August animation.

[Wipneus]

Volume and volume-anomaly graphs.

[Wipneus]

Fram volume export graph.

[Wipneus]

Latest thickness map compared with recent years and their differences.

[Wipneus]

Some people have use for the updated regional data files:

monthly:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional-monthly.txt.gz
daily:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional.txt.gz

[Aluminium]

Changes from August 15.

Kara ... -100%
ESS ... -100%
Chukchi ... -90%
Beaufort ... -22%
CAA ... -21%
CAB ... -15%
Greenland ... -5%
Baffin ... +28%

[grixm]

Baffin ... +28%

wut

[gerontocrat]

& here is the Volume data analysis as in my JAXA extent postings (as at 31 Aug or 1 Sep  -see previous posts for expanation)

Overall volume loss in August was pretty much average.

Average remaining volume loss wouldm give a minimum of 4.084 thousand km3, 3rd lowest.
This is 0.034 thousand km3 above the 2019 minimum.

In only one year in the last ten does remaining volume loss give a projection of less than 4 thousand km3.

[gerontocrat]

PIOMAS Data... a bit more

the PLUME shows a final result of a minimum below 4 thousand km3 is still a possibility.

The graph of August monthly averages shows 2020 is at the moment very much on the linear trend

The graphs of High Arctic & the Peripheral seas volume emphasise that by August the story is almost purely about the central arctic seas.

[gandul]

If PIOMAS was able to model all the bottom melt that must be going on correctly, especially in the Pacific side, I think this should become an easy second place.

[interstitial]

piomas shows 3 meter ice for a good part of polersterns journey to the north pole.  even after adjusting for concentration that is way to thick. This model does not really show the current arctic. It shows how the arctic used to behave.

[FishOutofWater]

PIOMAS is missing whatever happened north of Greenland to the pole and it's also not getting the obvious low ice areas on the north coast of Ellesmere Island. To misquote a famous songwriter, "Something's happening and we don't know what it is do we Mr. Jones."

I'm not sure what's more important than the unprecedented melt region the Polarstern blasted through between Greenland and the pole, but I hope that they are studying that more important thing.

I find this PIOMAS volume map disquieting. The results do not appear to fit what we have observed.

[oren]

Time for some regional analysis - many thanks to Wipneus for providing the data in usable format.

Let me just mention I agree with the sentiment echoed above. It seems PIOMAS is overestimating thickness in the region above Greenland, even considering that PIOMAS models ridges that skew the average thickness upwards. However, with no data available for current thickness of typical floes in the vicinity, despite the Polarstern passing that way recently, this criticism remains in the realm of intuition.
Onto the data.

* CAB loss in H2Aug was a respectable 693 km3, passing the losses of 2012 and 2007 by a smidgeon, though a far second to 2016's 888 km3 loss. Remaining CAB volume for the date passed 2019 and is now 2nd lowest. Highest past losses from now until the minimum were 328 km3 (in 2008) and 258 km3 (in 2012).
* CAA losses slowed sharply, both because of continuous import from the CAB through the PGAS "garlic press", and the refusal of the ice in the main channel to disappear. It seems current volume will remain rather stable going into the minimum.
* Beaufort losses were also far too slow considering the magnitude of remaining volume. Unless the Beaufort arm disappears very soon, anomalous volume will remain through the minimum. Loss from now to minimum should be 50-70 km3, up to 100 km3 in an aggressive scenario.
* Greenland Sea volume had already crashed in H1Aug, and remained stable in this period. The remaining volume is at the lowest level normal for the minimum, except for 2017 and 2018 which were highly anomalous in the Greenland Sea.
* Remaining volume in other seas is a meager 25 km3.

Summing expectations moving forward, with total remaining volume at 4311 km3:
* Passing 2011's 4302 km3 is guaranteed, thus a 3rd place finish is secured.
* It's easily plausible to lose a further 259 km3 and pass 2019's 4052 km3 for a 2nd place finish.
* It is reasonable with a late minimum scenario to break the 4000 km3 barrier.
* However passing 2012's 3673 km3 appears to be impossible, excepting some melting disaster/miracle.

Click to enlarge.

[oren]

Time for the prediction of volume for the September minimum. Please zoom the regional charts above to follow the analysis.
Of the seas traditionally participating in the minimum, the Laptev is nearly zero now. The ESS is record low. The Kara is very low. The Barents is on the low side. Chukchi is rather low as well. All are assumed to be zero in September - for some of these seas that is a very common achievement, for some a rather rare one.

We remain with the four unknowns:
* Greenland Sea - has been running high since spring. I expect it to crash at some point due to melting and lack of imports, especially now that most of the thick ice at the export region has disappeared. My bet is somewhere above 2019 but far below 2012. 125 km3.
* CAA - had a very slow start, but a fast decline. Temps have been running very high for a month. I expect a result above 2012/2011 but probably below 2019. 100 km3.
* Beaufort - the most anomalous. Has been running high and getting higher, due to extra imports and lower exports. I can't see it heading towards near-zero, and foresee a high volume, probably somewhere below 2013's 200 km3, but above 2018's 80 km3. 125 km3.
* CAB - naturally the most variable in magnitude. has been running high in winter and spring, only to crash in the last month back into the fold. From where we are now it is quite common to reach 4000-4200 km3. Following the high losses of 2007, 2012, 2016, we might get to 3600 km3, still above 2012's minimum of 3400 km3. And who knows? We might break loss records, and perhaps even reach a CAB record in September.

Summing everything, we get 350 km3 outside the CAB. With a reasonable/aggressive scenario of 3600 km3 in the CAB, we get a total of 3950 km3, a respectable 2nd, lower than 2019's 4050 km3, but still much higher than 2012's 3670 km3.
However, my money is still on a record low volume. I am betting the impact of the GAAC was partially missed by the model, and will be incorporated later when ice area crashes. In addition, early open water on the Siberian side might eat deeper into the CAB than usual. If a CAB record can be achieved, total volume will probably also reach a record low.

Examining my mid-July prediction, things appear to be roughly on track, but "my money" is going down the drain.

* Greenland Sea - I now expect a 100 km3 finish, compared to prediction of 125 km3.
* CAA - I now expect a 150 km3 finish, compared to prediction of 100 km3.
* Beaufort - I now expect a 140 km3 finish, compared to prediction of 125 km3.
* CAB - I now expect a 3550-3600 km3 finish, compared to main scenario prediction of 3600 km3.
* Total - I now expect a 3950-4000 km3 finish, compared to main scenario prediction of 3950 km3.

My hunch that PIOMAS will somehow catch up to the GAAC effects proved false (for now, for now).

The minimum should be with us in the next PIOMAS update - stay tuned. With the non-stop winds over the basin, predicted major storm, and lots of uncertainty about timing of the refreeze, it's not going to be boring.

[Frivolousz21]

PIOMAS is missing whatever happened north of Greenland to the pole and it's also not getting the obvious low ice areas on the north coast of Ellesmere Island. To misquote a famous songwriter, "Something's happening and we don't know what it is do we Mr. Jones."

I'm not sure what's more important than the unprecedented melt region the Polarstern blasted through between Greenland and the pole, but I hope that they are studying that more important thing.

I find this PIOMAS volume map disquieting. The results do not appear to fit what we have observed.

We wait fof cryosat and icesat 2 in October and get real volumes.

Combine cryosat and smos for thickness and ice sat  for snow depth and we have a great product

[oren]

Monthly update from the Polar Science Center:

August 2020 Monthly Update

Average Arctic sea ice volume in August 2020 was 5,300 km3. This value is  only 300 km3 above the  record minimum  value of 5,000 km3 set in 2012. This makes 2020 the third lowest on record for August.   Monthly  ice volume was 71% below the maximum in 1979 and 55% below the mean value for 1979-2019. Average August  2020 ice volume falls on the trend line. August 2020 volume anomaly falls on the lower end of the spectrum of recent years (Fig 4.)  Average ice thickness is in the middle of the pack for the more recent years  (Fig 5). Ice thickness anomalies for August 2020 relative to 2011-2018 (Fig 6) continue the pattern that has emerged over the winter, spring and shows relatively thin ice along the Russian Coast and thicker than normal in the Eastern Beaufort and the along the Canadian Archipelago. Thicker than normal ice in the Barents Sea that was present in previous months is nearly gone. (  Those thickness anomalies have expanded to cover a large part of the Arctic now. A strong negative ice thickness anomaly has developed north of Greenland and corresponds to low ice concentrations apparent in passive microwave data. This is noteworthy as this area typically features some of the thickest ice in the Arctic and is termed the Last Ice Area (see here).  The overall  thickness anomaly pattern is likely a combination of the recent warm temperatures along the Siberian Coast  and the sea ice drift pattern that pushed  sea ice away from the Siberian Coast and towards North American and Greenland Coast that has persistent since January (Fig 8.).  The mean circulation pattern for the first half of 2020 shows a very small Beaufort Gyre and transpolar drift stream located closer to North America than normal. The drift anomaly shows a counter clock wise pattern compared to the normal counter clockwise pattern of sea ice drift. 

[anaphylaxia]

Monthly update from the Polar Science Center:

August 2020 Monthly Update

 The drift anomaly shows a counter clock wise pattern compared to the normal counter clockwise pattern of sea ice drift. 

So the ice spins the way it should?

[oren]

I guess the normal pattern is clockwise.

[Tor Bejnar]

I read that your way, anaphylaxia.

[oren]

Fig 8. PIOMAS Mean Sea Ice Motion (left) and anomaly relative to 2011-2018 for January through June 2020.

[Tor Bejnar]

Did you contact them about the apparent typo?  (It is still there.)

[Lord M Vader]

September 18 today. And it looks very likely that the minimum occurred by September 15. Anyone who has checker PIOMAS data to fond out where 2020 is compared to other years? By August 31 we where third lowest behind 2012 and 2019.

Wipneus?

[oren]

Data should be released today or tomorrow.