Latest PIOMAS Volume update

[Richard Rathbone]

.

The volume and volume anomaly graphs, official data with the June data calculated from grid. Will 2017 stay at the low anomaly level that it held for months?

I think it'll take really bad weather for melt in July to close the gap to 2012 before August. The shallower anomaly troughs do tend to be broader ones and there's still a fair amount of the Svalbard blob to go, so I reckon 2017 has a rather better than evens chance of extending its lead during July before falling back again in the remainder of the melt season.

[VeliAlbertKallio]

The rate of sea ice pulverization and vertical mixing of ocean water is very difficult to estimate, given that many areas remain cloud covered and to assess how much overturning and mixing is there is hard to see. When ice is sufficiently broken down, the winds do not need to be as strong as in the past for large transfers of heat between ocean and ice to occur. The more pulverized and well mixed sea water becomes more stable the temperatures may appear. This means that the temperature tend to be suppressed the more scattered and broken sea ice becomes. Sea ice is most broken at the rear where it is bathing in sunlight and heat, but also exceedingly broken on the Atlantic front meaning that much more heat can be taken up despite seemingly cold air. There are pictures showing some areas very battered ice. The ice does not only erode by surface and bottom melt, but by the ice floes pounding against eace other when spatial viscosity lowers.

Volume on day 181 is about 12.163 [1000km3], that is 0.18 [1000km3] below 2012's volume. IOW the volume gap (barely) survived.

It could have been worse, but it's bad enough that volume and extent are even close to 2012 numbers, not to mention, slightly below.

In terms of total volume, the decline was average. Very slightly above my "black dot". I am very curious about the regional distribution.
Looking at all that yellow color draining from the pacific side of the CAB, and the vulnerable white spots next to Svalbard and Greenland, I have a feeling this is not over yet.

[DavidR]

The volume and volume anomaly graphs, official data with the June data calculated from grid. Will 2017 stay at the low anomaly level that it held for months?

These graphs show how extreme the volume loss was in 2012, while emphasizing that 2017 has maintained a large difference with 2016.  Compared to  2016  the ice is still much thinner and the only reasonable expectation from that is that  the amount of area / extent loss will be much  greater from now to the end of the season than in 2016, 2017 is much  more likely  to  follow the trajectory  of 2012 than 2016.

[A-Team]

Here is a comparison to current hycom followed by a progression to thinner ice with Atlantic Water corridor ice removed, without re-dispersion, up to the end of September. Recall last fall that the melt season overlapped considerably with the re-freeze season, depending on location and date.

Note how Piomas has returned from its long eccentric odyssey to again resemble hycom and sea ice age along the CAA.

Only the Inner Basin gap relative to other years has any interest; for example ice along the greenland coast has racheted out of relevance and should not be used in inter-year volume comparisons. It never returns to the Arctic Ocean once the 80th parallel has been cleared, nor do its melting floes cool northbound Atlantic Water (which is well to the east at 300m depth).

Another effect distorting comparisons this year has been pile-up of ice along the Svalbard-FJL-SZ line. Unless persistent winds somehow reverse the Transpolar Drift, better guidance will result from deleting all ice volume south of the ~1000 km Barents Bathymetry Break before figuring gaps relative to prior years. This masking step is shown in the next to last frame below but so far has not surfaced in our line graphs or grid cell eliminations.

It would be great if people who lump/gap the whole Northern Hemisphere instead of the Inner Basin (as modified) out to D90 could explain their thought processes. I wouldn't bet the farm on a D1 for the western US.

The real question though is whether inter-year inner-basin volume comparisons (say 2015-17 to earlier) still make sense. At one time, subtraction (ie relative volume change) took out shared error in the algorithm but in a 'New Arctic' increasingly dominated by dodgy first-year ice, that is a dubious proposition.

[magnamentis]

Here is a comparison to current hycom followed by thinner ice and Atlantic Water corridor ice removed, without re-dispersion, up to the end of September.

thats good for a best guess, had something similar imagined great animation, thanks

[FishOutofWater]

Excellent figure that shows how dangerously thin the ice is now, A-Team. If I were to bet I would put in a slightly higher extent than the figure shows because compaction will help preserve some of the thin ice.

[Lord M Vader]

I think the pic A-Team is presenting is somewhat bullish and I don't see a new record minimum ahead of us.

[oren]

While waiting for the more detailed and regional data, here's my stab at total volume extrapolated to the end of July, using the "typical-low melt" figures, which basically are the same numbers as 2012 (no volume cliff came about for 2012 in July) and unsurprisingly keeping 2017 with the same slight lead. Other numbers from the top melt years would carry 2017 further into the lead.
The first black dot is Wipneus' calculation of the current number, the next two are extrapolations for mid- and end-July.

[werther]

PIOMAS update is available. Just under 2012. Extent just under '12 too. Ice quality might be considered far worse, though. Next four weeks of weather will be decisive.

[Wipneus]

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

[Wipneus]

Fram export was low last month. Monthly and daily graphs attached.

[Wipneus]

Area per thickness categories . None of the categories on 30th June is a good proxy for the "final" area. That will be different for the mid-July update .

[Wipneus]

The thickness graph for 30st June, compared with previous years.

[Wipneus]

The 2012-2017 volume gap.

[Neven]

Very interesting to have the gap with 2012 close, but at the same time see this huge discrepancy, an almost diametrical opposition, on the comparison map. Sure, 2012 will catch up with the dark blue in the Chukchi and ESS where the ice has disappeared quickly this year, but the same goes for 2017 with all that red on the Atlantic side, especially that big blotch north of Svalbard and of course Baffin Bay. And then there's that last zone of red, in the CAA, that has received plenty of sunshine during the past few weeks.

It implies that it should be more than possible for 2017 to keep pace with 2012.

[oren]

Very interesting race indeed. I will be away from my computer for some days but have made some calculations analyzing the "inner basin" volume gap. Pray excuse rounding errors and no chart. Note: all my "averages" include only 2010, 2011, 2012 and 2016. Numbers are in km3.
First, 2017 had a slightly above-average inner volume loss, 2460, reaching 10,656 on day 181, compared to 2012 at 11,480 and 2011 at 11,333.
Comparing by region vs 2012:
Sea - 2017 lead day 166 - 2017 lead day 181
Beauf   -73   -16
Chukc   496   477
ESS       293   422
Laptv   -310   -152
CAB      598    210   big change
CAA      -29 -149
Total inner  ~970     ~820

Looking at losses to day 196 (mid-July) with a clustered range of 2539-2691, and assuming low-typical 2550, extrapolated 2017 will get to inner basin volume of ~8,100 and lead of ~680. Race not over. (But of course August was the real freak).

Looking at regional distribution and the comparison maps, it's mostly a tale of Pacific side vs. Atlantic side. As the Atlantic side is much more vulnerable, I would say it's one more  advantage to 2017.
On the other hand, 2012 had ice left only in CAB and CAA at minimum (besides Greenland sea) and the lead in these two regions has almost vanished.
Interesting indeed.

[RikW]

The most striking to me is the positive difference on the atlantic side with almost all of the past 10 years. The model implies that we have a 'record-high' amount of ice on the atlantic. It won't surprise me if that difference will disappear in the coming weeks, since large parts of that area melt out completely the last decade

[Jim Pettit]

...t should be more than possible for 2017 to keep pace with 2012.

That seems almost a certainty at this point. Only 2014 had a small enough decrease from this date onward that a repeat would fail to set a new record; a repetition of the behavior of any and every other year on the record would see a new one. (FWIW, the average July through minimum melt of the last ten seasons would render a 2017 September minimum of 3.18 thousand km3, or nearly 500 km3 below the 2012 record.)

July and August behavior are--very obviously--going to make a huge difference.


[url=http://iwantsomeproof.com/extimg/siv_projections_from_current_date.png]http://iwantsomeproof.com/extimg/siv_projections_from_current_date.png
[/url]

[Bill Fothergill]

...t should be more than possible for 2017 to keep pace with 2012.

That seems almost a certainty at this point. Only 2014 had a small enough decrease from this date onward that a repeat would fail to set a new record; a repetition of the behavior of any and every other year on the record would see a new one. (FWIW, the average July through minimum melt of the last ten seasons would render a 2017 September minimum of 3.18 thousand km3, or nearly 500 km3 below the 2012 record.) ...

Thanks Jim - I'll be sure to sleep better tonight.
 
(Although the beer at the local quiz will probably have a part to play as well.)

[Blizzard92]

My PIOMAS graphics have also been updated for June at http://sites.uci.edu/zlabe/arctic-sea-ice-volumethickness/.

[Tigertown]

I think that it might be worth pointing out that the monthly value for June came out to 15.4 km³ and was 500 km³ less than the previous record for June which was set in 2012. Also. June of this year saw a value that came in under 2016 by about 1000 km³. There is a sizable difference in the monthly numbers as opposed to the slighter difference in the end of the month volume numbers. I would think this to carry more weight in comparing years than the day of the year numbers.

[Neven]

Latest PIOMAS update has been posted on the ASIB.

[Rob Dekker]

...t should be more than possible for 2017 to keep pace with 2012.

That seems almost a certainty at this point. Only 2014 had a small enough decrease from this date onward that a repeat would fail to set a new record; a repetition of the behavior of any and every other year on the record would see a new one. (FWIW, the average July through minimum melt of the last ten seasons would render a 2017 September minimum of 3.18 thousand km3, or nearly 500 km3 below the 2012 record.)

Jim, this does not make sense.
Since 2017 and 2012 are in a virtual tie for volume right now, your remarks suggest that 2012 had the smallest volume loss of any year on the record (from July to September), except for 2014. Is that true ?

[seaicesailor]

Truth be told, if one looks at the volume anomaly from July onwards, 2012 is not outstanding, (not in its value itself which is outstanding, but in its evolution to becoming less anomalous):
https://forum.arctic-sea-ice.net/index.php/topic,119.msg119288.html#msg119288

Don't ask me why though

[Clenchie]

Truth be told, if one looks at the volume anomaly from July onwards, 2012 is not outstanding, (not in its value itself which is outstanding, but in its evolution to becoming less anomalous):
https://forum.arctic-sea-ice.net/index.php/topic,119.msg119288.html#msg119288

Don't ask me why though

I am no expert but if what you are saying is that from July onwards, 2012 lost a lot of extent but not an anomalous amount of volume then my take would be that it latterly lost a large area of very thin ice, leaving the thicker core relatively normally affected.  That way the extent declined heavily but the volume declined normally.  Does that make sense?

Other than that it could just be inaccurate volume measurements.

[seaicesailor]

Truth be told, if one looks at the volume anomaly from July onwards, 2012 is not outstanding, (not in its value itself which is outstanding, but in its evolution to becoming less anomalous):
https://forum.arctic-sea-ice.net/index.php/topic,119.msg119288.html#msg119288

Don't ask me why though

I am no expert but if what you are saying is that from July onwards, 2012 lost a lot of extent but not an anomalous amount of volume then my take would be that it latterly lost a large area of very thin ice, leaving the thicker core relatively normally affected.  That way the extent declined heavily but the volume declined normally.  Does that make sense?

Other than that it could just be inaccurate volume measurements.

Yes makes sense to me that much of the ice that was gone in September 2012 had already thinned considerably by July. Given the crazy May/June 2012 ...

[Avalonian]

Yes makes sense to me that much of the ice that was gone in September 2012 had already thinned considerably by July. Given the crazy May/June 2012 ...

Yes. Looking at A-Team's recent graphics on the 2017 melting thread (p. 52, based on Hycom data), I'm not sure whether we shouldn't be expecting the same this year. The lack of thick ice, and the rapid thinning from the Siberian side, looks very like the start of a 2012-type extent crash.

 If I understand it right, a significant proportion of the thicker ice that's left this year (e.g. N of Svalbard) is practically doomed; so on balance, it looks to me as though we should expect a sustained extent drop for the remainder of the season, AND a large volume drop. A lot of people have been thinking of 2012 as an extreme outlier that we're unlikely to match for the rest of the season. What I'm seeing (and some others here, I know) suggests that this year can still be even more of an outlier.

[Jim Pettit]

Jim, this does not make sense.
Since 2017 and 2012 are in a virtual tie for volume right now, your remarks suggest that 2012 had the smallest volume loss of any year on the record (from July to September), except for 2014. Is that true ?

Close to it. Here are the volume losses (in 1000 km3) from the end of June each year through that year's respective minimum (note that 2011 and 2012 were virtually tied):

2016: -9.020
2015: -9.599
2014: -7.837
2013: -8.648
2012: -8.622
2011: -8.615
2010: -8.876
2009: -9.724
2008: -10.370
2007: -9.135
2006: -8.936
2005: -9.405
2004: -10.034
2003: -9.830
2002: -9.878
2001: -8.895
2000: -10.088

So, again, a repeat of any of those years but 2014 would leave us with a new record.

[Richard Rathbone]

...
Since 2017 and 2012 are in a virtual tie for volume right now, your remarks suggest that 2012 had the smallest volume loss of any year on the record (from July to September), except for 2014. Is that true ?

The normal volume pattern post 2007 is big to very big losses in June and small to average losses in July/August. 2012 had a very big June, a small July and an average August.

During the last decade, melting out exceptional amounts of ice in June has been followed by finding the remaining ice a bit harder to melt than normal. The volume in April 2013 was pretty much back where it was in April 2012, and part of that was slow July 2012 and the rest was high freeze rates.

The melt given up in July is rather less than the amount gained in June, so a big June still means June+July is big, but the record anomaly happens at the end of June.

[VeliAlbertKallio]

It is obvious that the early seasons in the past have recorded large losses by the end of June. This is mainly as the increased melt rates hit the easy-to-melt peripheral areas first where volume losses occur. The small 'core ice' region of the Central Arctic around the North Pole is small in areawise. It is always this ice that melts at the end of the season. If winter sea ice area growth is low in the peripheral seas, it depresses the easy-to-melt volume in June like in 2012.
So one cannot have a huge volume loss in June if the vast outspanning areas melted earlier in the season or did not form ice in the first place due to a warm winter. In the past (when the core area of ice hardly melted at all) ice like the Lincoln Sea was several meters thick! 2017 sees fractured and pulverized ice even in CAB, this was not the case in the past.

...
Since 2017 and 2012 are in a virtual tie for volume right now, your remarks suggest that 2012 had the smallest volume loss of any year on the record (from July to September), except for 2014. Is that true ?

The normal volume pattern post 2007 is big to very big losses in June and small to average losses in July/August. 2012 had a very big June, a small July and an average August.

During the last decade, melting out exceptional amounts of ice in June has been followed by finding the remaining ice a bit harder to melt than normal. The volume in April 2013 was pretty much back where it was in April 2012, and part of that was slow July 2012 and the rest was high freeze rates.

The melt given up in July is rather less than the amount gained in June, so a big June still means June+July is big, but the record anomaly happens at the end of June.

[Thawing Thunder]

So, again, a repeat of any of those years but 2014 would leave us with a new record.

And a slow 2014 melt is already improbable, given the vast open areas close to the CAB, which means the sea in many places has become a solar radiation collector. 2014, if memory serves, still had vast – or at least well above average – area and extent in June.

[Neven]

And a slow 2014 melt is already improbable, given the vast open areas close to the CAB, which means the sea in many places has become a solar radiation collector. 2014, if memory serves, still had vast – or at least well above average – area and extent in June.

But a huge hole in the Laptev area as well:



It's going to take weather to slow down volume loss sufficiently for a new record to be prevented. I can't remember exactly what killed it in 2014, but it was probably very uneventful weather (what I call neither fish nor flesh), no high highs or low lows. Just clouds and little wind.

[magnamentis]

just been reading through neven's blog and i think we should once again talk about naked numbers compared to percentages, i gonna use examples to save me the search for all the numbers because this is about a concept and/or how to look at things from different angles, not about specific numbers:

even though the gap seems to be small, the gap in volume to second lowest in percent is not that as small as it seems, it's a gap of around 6-7%

once the volume will be down to let's say 1000km3 the very same SMALL gap of 131km3 we have now will be a gap of 13.1% and once we are down to 131km3 the same gap will be the ultimate gap of 100%, i hope that without further words it's clear where i'm heading to

nota bene that the same, applies to most if not all the measurements and statistics we are dealing with here. like jesus said, the one who has 2 monetary units and shares one gives more than the rich who shares 10. OT but the same way to look at things LOL

as i proposed (and predict) earlier, we should start to pay more attention to percent difference than to differences in numbers the closer we get to the "null" line.

similarly i vote to put into account from where we are coming from. if we start at 3000 of a given measurement we can't expect to loose 2000 thousand in one months. so the lower we start into the melting seasons the less can we compare the rate of loss with earlier years because if we do we shall sooner or later expect impossible, i.e. sub zero numbers and since that can't happen hear things like the decline has come to a halt or the rate of decline has slowed down, while in reality it has not.

with predict i mean that sooner or later everyone will recognize that the old way of comparing things and looking at what happens will not serve the purpose anymore.

<snip; let's leave out the strawmen and ego stuff; N.>

i'm open to any open and honest exchange in any place where it's not OT and i always consider the possibility that i overlooked something or could be wrong, this just for the usual suspects, i.e. AT<AT or BY LOL

[Tigertown]

If or when the mid-month numbers come in, I think the report might be a shocker. Just looking at the JAXA volume chart by Wipneus, and it seems to have corrected itself from all the pseudo super-thick areas that were previously showing on the JAXA thickness chart. Although it shows volume at below 7,000 km³, I don't  expect PIOMAS to be quite that low. It would not surprise me if it has dropped more than expected, though.

[DavidR]

...t should be more than possible for 2017 to keep pace with 2012.

That seems almost a certainty at this point. Only 2014 had a small enough decrease from this date onward that a repeat would fail to set a new record; a repetition of the behavior of any and every other year on the record would see a new one. (FWIW, the average July through minimum melt of the last ten seasons would render a 2017 September minimum of 3.18 thousand km3, or nearly 500 km3 below the 2012 record.)

Jim, this does not make sense.
Since 2017 and 2012 are in a virtual tie for volume right now, your remarks suggest that 2012 had the smallest volume loss of any year on the record (from July to September), except for 2014. Is that true ?

2012 has the third lowest volume loss (10.96)  in the record after 1996 and 2014.  However it  is worth noting that the trend for volume loss after 30 June is decreasing!.   The graph shows the 30th June volume, and total volume loss along with  pre and post 30th June loss figures. 

[oren]

2012 has the third lowest volume loss (10.96)  in the record after 1996 and 2014.

Thanks for the interesting chart DavidR. Here's a partial explanation why 2012 lost less volume post-June than other years. It had very little peripheral volume to lose.

[oren]

Waiting for the mid-month update, here's the season "inner basin" chart with my usual extrapolation method, looking only at the top years and choosing the lowest loss (rounded) in each half-month period. My personal expectation is that results should come in at these black dots or below them but not above. Note there is a strong bias here, as I'm only using years that finished in the top 4. Worked til now though...

[Cid_Yama]

It's like watching a ship sink.  The air inside the hull maintains buoyancy while it slowly bleeds out until the buoyancy threshold is reached at which point the ship sinks suddenly.

[Tor Bejnar]

... The graph shows the 30th June volume, and total volume loss along with  pre and post 30th June loss figures.

Projecting (by eye) the June 30 volume curve and the post June 30 loss curve (or line), they seem to intersect about 2019. [Click on the bold type quote link at the top to see David's graph.][later edit after David's post below:  David corrected my 'eyeball' that didn't notice the years were 'every-other', therefore my approx. halving the projected ice-free date.]

[DavidR]

... The graph shows the 30th June volume, and total volume loss along with  pre and post 30th June loss figures.

Projecting (by eye) the June 30 volume curve and the post June 30 loss curve (or line), they seem to intersect about 2019. [Click on the bold type quote link at the top to see David's graph.]

It's more like 2022 which fits with the projection for an ice free September.

[Wipneus]

Mid month update of gridded daily data has arrived.

The volume gap with 2012 is still there, it is still small (0.15 [1000 km³]).

Here is the animation, more info later today.

[Tigertown]

Many people keep quoting extent alone and say there is not much melting, but watching the thick ice melt like ice cream is a wake up call. By thick I mean the proxy ice we have for thick now. Curious to see how this loss will translate into numbers.

[greatdying2]

The "island" of purported "thick" ice north of Svalbard shows up as low concentration in recent area maps, which seems to be borne out by satellite imagery.

Is this just normal PIOMAS modelling error (potentially), or is PIOMAS having trouble this year?

[Andreas T]

I have been watching this area and the discrepancy there is striking. PIOMAS has been piling ice against Svalbard for months. That movement towards the islands was visible on the tracking I did in AMSR images but later when it relaxed and slightly reversed in July water opened north of Spitsbergen but ice has still compacted against Nordaustlandet. What the satellite images don't tell us is how thick the floes are, but the dispersion should show as thinning on PIOMAS.
It seems possible to me that PIOMAS has underestimated the thinning which took place from bottom melt in that area (the strip of shelf on the north coast of Spitsbergen), it may also have got some of the movement wrong (more against Nordaustlandet than Spitsbergen)

[oren]

I think the high-concentration ice next to Nordauslandet is what PIOMAS "means" when showing that blob, with somewhat wrong location. By end-July this whole blob might be gone.

[Wipneus]

Updated daily volume and daily volume anomaly graphs.

[Wipneus]

My estimated total volume on July 1-15:

[EDIT: last day was missing (spotted by oren)]
array([ 11.953,  11.734,  11.52 ,  11.309,  11.102,  10.916,  10.686,
        10.437,  10.239,  10.068,   9.894,   9.709,   9.527,   9.28 ,
         9.049])

Updated egional daily data file is here:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional.txt.gz

[oren]

My estimated total volume on July 1-15:

array([ 11.953,  11.734,  11.52 ,  11.309,  11.102,  10.916,  10.686,
        10.437,  10.239,  10.068,   9.894,   9.709,   9.527,   9.28 ])

Updated egional daily data file is here:
https://sites.google.com/site/arctischepinguin/home/piomas/data/PIOMAS-regional.txt.gz

As usual, thank you for all the updates. Note one day is missing at the end, ~9.05.

An update to the Inner Basin volume chart. The gap to 2012/2011 closed by ~300km³, which is ~200km³ more than expected by my "method" and a sign of lowering probabilities for records. My next extrapolations were shifted upwards accordingly. In my next post I will attempt to find out which region contributed the most to this.

[oren]

Turns out the 2017-2012 gap closed by about ~150 on the pacific side, and by ~150 in the CAB, while in the periphery the 2012 advantage shrank by ~300 on the atlantic side, keeping the same total volume difference.
          2017 lead   2017 lead   2017 lead
Region   Day 181   Day 196    Fell by
Beauf         16        -27           43
Chukc        477       378          99
ESS           422       401          21
Laptv        -152      -162         10
KaraS       -230      -85          -145
Baren       -91        -44          -47
GrnLS      -217       -92          -125
CAB          210        28           182
CAA         -149      -117          -32
Baffn        -196      -126         -70
Hudsn        42         -4             46
Pacific        915       752          163
Atlantic     -538      -221         -317
Inner Basin   824      501         323
Periph. Seas -692   -351         -341
(all numbers in km³)
Graphically, 2017 seems to be losing steam in the CAB, while 2012 has the GAC visibly waiting. Time will tell if the remaining pacific advantage can be translated to an acceleration in the CAB.

[seaicesailor]

Nice. Thank you Oren!
The weather vs thinness "game" ain't over till it's over.