The 2015 melting season

[DavidR]

That's a tough call, and I'm absolutely no expert on sea ice, just looking at the numbers of it.

But if ocean heat related and wider than the 2011 previous record plateau, we could at least lose the same amount of extent as 2011 did from its max to min. That would take us from 13942060 max to 13942060-9858530 = 4.08 million km2 min. 3rd.

Code: [Select]

2007: 14209677 30 days (-269894)
2008: 14774776 27 days (494895)
2009: 14657047 27 days (-30429)
2010: 14688540 33 days (-219301)
2011: 14127729 42 days (-277725)
2012: 14709086 30 days (-86637)
2013: 14523635 26 days (459327)
2014: 14448416 10 days (-94278)
2015: 13942060 42 days (?)

Can you  provide the loss each year (Max - Min) so I can do  the comparison.
Thanks

[viddaloo]

Code: [Select]

2007: 14209677 30 days (from 46 to 75) (max–min: 10143938)
2008: 14774776 27 days (from 56 to 82) (max–min: 10274153)
2009: 14657047 27 days (from 60 to 86) (max–min: 9602992)
2010: 14688540 33 days (from 65 to 97) (max–min: 10066448)
2011: 14127729 42 days (from 52 to 93) (max–min: 9858530)
2012: 14709086 30 days (from 63 to 92) (max–min: 11531631)
2013: 14523635 26 days (from 57 to 82) (max–min: 9714347)
2014: 14448416 10 days (from 74 to 83) (max–min: 9564296)
2015: 13942060 43 days (from 42 to 84) (max–min: ?)
Also: 105% more melt expected this week compared to 2011!

[plg]

I had a look at the NH, IJIS and CRYO datasets, and noted the following for 2015:

• Cryophere had the earliest maximum so far (Feb 17). By today only 2012 has a later max date (Mar 29), and since the area was 320k below the maximum on the 24th, it appears reasonably safe to call the February maximum.
• IJIS/JAXA also had the earliest maximum so far (Feb 15). By today only 2010 has a later max date (Mar 31). However, since the extent now (24th) is only about 88k below the maximum, a new maximum is not inconceivable.
• NOAA had the fourth earliest maximum so far (Feb 22), by one day (1987, 1994 and 1996 all peaked on Feb 21). By today only 1999 and 2010 has a later max dates (Mar 30, 31). The extent on the 23rd was 89k below maximum, and on the 24th 227k below. Will probably not exceed maximum, but...

Although the early maxima will probably hold, I find it interesting that there is a undisregardable (is that a word?) probability that the earliest maximum suddenly becomes the latest maximum, at least for extent.

Caveat: this is statistics (numerology?), no regard for current weather or forecasts.

[DavidR]

Code: [Select]

2007: 14209677 30 days (from 46 to 75) (max–min: 10143938)
2008: 14774776 27 days (from 56 to 82) (max–min: 10274153)
2009: 14657047 27 days (from 60 to 86) (max–min: 9602992)
2010: 14688540 33 days (from 65 to 97) (max–min: 10066448)
2011: 14127729 42 days (from 52 to 93) (max–min: 9858530)
2012: 14709086 30 days (from 63 to 92) (max–min: 11531631)
2013: 14523635 26 days (from 57 to 82) (max–min: 9714347)
2014: 14448416 10 days (from 74 to 83) (max–min: 9564296)
2015: 13942060 43 days (from 42 to 84) (max–min: ?)

Thanks Vid, 
The estimate  using this data is an IJIS extent minimum in 2015 around 400K km^2 above 2012 and 500 below 2007.
NSIDC by  comparison is  predicting a result much  closer than that to 2012 and we are still adding days to 2015 values in both sets.

[Neven]

I had a look at the NH, IJIS and CRYO datasets, and noted the following for 2015:

• Cryophere had the earliest maximum so far (Feb 17). By today only 2012 has a later max date (Mar 29), and since the area was 320k below the maximum on the 24th, it appears reasonably safe to call the February maximum.
• IJIS/JAXA also had the earliest maximum so far (Feb 15). By today only 2010 has a later max date (Mar 31). However, since the extent now (24th) is only about 88k below the maximum, a new maximum is not inconceivable.
• NOAA had the fourth earliest maximum so far (Feb 22), by one day (1987, 1994 and 1996 all peaked on Feb 21). By today only 1999 and 2010 has a later max dates (Mar 30, 31). The extent on the 23rd was 89k below maximum, and on the 24th 227k below. Will probably not exceed maximum, but...

Although the early maxima will probably hold, I find it interesting that there is a undisregardable (is that a word?) probability that the earliest maximum suddenly becomes the latest maximum, at least for extent.

Caveat: this is statistics (numerology?), no regard for current weather or forecasts.

Thanks for the overview, plg.

[viddaloo]

Thanks Vid, 
The estimate  using this data is an IJIS extent minimum in 2015 around 400K km^2 above 2012 and 500 below 2007.
NSIDC by  comparison is  predicting a result much  closer than that to 2012 and we are still adding days to 2015 values in both sets.

You're welcome, DavidR.

While I think it's a great observation, I wouldn't go as far as trying to predict which position we'll have for minimum — other than maybe 'bottom 5'–like precision — and the reasons are as follows: April weather, May weather, June weather, July weather, August weather and September weather 

I do agree we have a better chance than ever of hitting the criteria for a 'Blue Ocean Event'.

[viddaloo]

Here are the 5 longest plateaus of the 9 years since 2007, and 2015 set a new record length today with 43 days. All of the 4 others have exceptionally high melt in one way or another, for 2015 of course it remains to see.



I'm guessing that what we see here is winter refreeze into February and early March, up to a certain point where the ocean heat doesn't allow more extent growth that year. It is then back and forth for the ice extent on this plateau for up to 1.5 month, but a lower plateau (warmer water) means both a longer plateau and an earlier start of that plateau. (Note: The maximum doesn't have to be early for the plateau to kick in early, and because you have a plateau for 1–2 months, the exact timing of the max is more or less random and insignificant.) In early february the refreeze follows a track that is approximately similar for all years, which means the plateau kicks in earlier if that year has a lower extent plateau. At the far end of the plateau the paths are more diverse, with perhaps a trend towards an earlier abandonment of the plateau and start of the rapid melt? 2011, that held the record for low plateau before 2015, got off to an early start of the melt on Apr 3 from a very low level, and took advantage of albedo and other positive feedbacks to go as low as 2nd lowest at the time. I'm guessing that if we leave the plateau within the next 7 days, 2015 will have a great shot at getting an equally good or even better start of the first phase of the melt.

Anyone ready to make a guess when we'll leave the 2015 plateau? In real terms it implies going below 13.74 million km2 for good.

[Lord M Vader]

If JAXAs value of 13,92 million km² holds and we put an average of the latest melt seasons DavidR is presenting it's reasonable to believe that about 9,5-10,2 million km² of the sea ice will be lost during the melt season. That would yield a reasonable SIE minimum value in the range of 3,7-4,4 million km² in September. If such a unlikely big melt like 2012 would be imminent we would be down to a new record minima of roughly 2,8 million km².

As the sea ice is thicker than it was in the beginning of 2012 I highly doubt we'll see such a huge melt this year! Therefore, I think it's fair to believe 2015 will end up being the second or third lowest on record!

Best, LMV

[Nightvid Cole]

If JAXAs value of 13,92 million km² holds and we put an average of the latest melt seasons DavidR is presenting it's reasonable to believe that about 9,5-10,2 million km² of the sea ice will be lost during the melt season. That would yield a reasonable SIE minimum value in the range of 3,7-4,4 million km² in September. If such a unlikely big melt like 2012 would be imminent we would be down to a new record minima of roughly 2,8 million km².

As the sea ice is thicker than it was in the beginning of 2012 I highly doubt we'll see such a huge melt this year! Therefore, I think it's fair to believe 2015 will end up being the second or third lowest on record!

Best, LMV

It's actually not thicker than 2012 in most of the Arctic Ocean. The average thickness is higher only because of a band of really thick ice jammed up against northern Greenland and the CAA. If you look at the PIOMAS thickness plot for February, the area of ice more than 1.75m thick is quite a bit *smaller* in 2015 than 2012. In 2012 most of the Arctic Ocean was over 1.75m thick. This year, not so much.

Most of the ice pack, with the exception of that thick band near the CAA and Greenland, is actually more vulnerable than in 2012.

[viddaloo]

Another way of reading the Plateau plot is like an EKG hospital screen for the 'winter power': The agonising death and last breath of winter ....

[BornFromTheVoid]

This evenings ECM shows just about the worst conditions we could get for sea ice retention with very mild air being pulled in through the Bering strait, very mild air being pumped north over the Kara and Laptev seas (850hPa temp anomalies of over 12C) with a strong pressure gradient north and south of Fram which would flush out an awful lot of ice.... luckily, that pattern doesn't start until 7 days out, so it's likely to change over the next few model runs.

In the meantime, some mixed conditions about. Warm air and southerly winds will arrive in the Barents and Kara region from today, so renewed melt is likely there. Conditions remain similar to recent days over the Bering sea, so probably little coverage change, but the ice should be spreading out and thinning from the warmth associated with the record strong +ve PDO.
Baffin sea remains cool throughout, so probably a slow decline there, which Okhotsk sees mostly mild conditions which should promote a gradual melt.
Overall, I'd say a steady decline, close to average, over the next 5 days.

[iceman]

    .... a strong pressure gradient north and south of Fram which would flush out an awful lot of ice....

Yes, that will be an impressive surge of thicker ice through Fram Strait next week.  Looks like a harbinger of an early volume max.

....
Baffin sea remains cool throughout, so probably a slow decline there, which Okhotsk sees mostly mild conditions which should promote a gradual melt.
....

My take on Baffin/Newfoundland is erratic extent increases through the end of the month, after which the region gets hit by more warm moist winds off the Gulf Stream.  It would be surprising, though, if any gains there exceeded the coming losses in Barents and then Kara.

[cesium62]

In my never-ending quest to find graphs half as cool as Viddaloo's...

I've been looking at the volatility of the NSIDC yearly lines relative to the historical average.  In this graph, we've converted daily extent into daily deviation from the historical average, and then we've plotted the mean for each year as the middle yellow wiggly line.  The other four wiggly lines are 1 and 2 standard deviations above and below the mean.  Light colored linear trend lines are overlaid on top.

So, 2012 was an exceptionally volatile year.  (1996 was nearly as volatile.)  2012's extent stretched above the historical average now and then, and reached down to 3 standard deviations below the historical average.  2013 and 2014 (and, so far, 2015) have been about as non-volatile as years get.  Squishing away the volatility seems to make the extent fall more linearly over time.

The mean yearly extent has fallen from about 1.5 standard deviations above the historical average to 1.5 standard deviations below the historical average over the past 35 years.  But it looks like it will need to fall at least another 3 standard deviations before anyone will claim ice free seas.  (-6 standard deviations in September is something like 1m km^2 of extent.  Give or take. Modulo math errors.)

[plg]

Latest IJIS data (Mar 26) shows that the extent is within 58156 from the Feb maximum, so we may yet see the earliest max transform into the latest...

However, it is reasonably safe to assume that it will be the lowest on record, as it is still 127.4k from taking a second place which seems unlikely given the forecasts discussed above.

[Frivolousz21]

Current weather forecasts are showing some major torching going down over parts of Eurasia heading threw the 1st week of April.

This is also when solar insolation between 50-70N is really sky rocketing day after day.

Their is potential depending on the persistence of this pattern for a major early season snow melt over the Western 1/2 of Eurasia.

Both the Euro and GFS and their ensembles are in almost lock step showing the pattern over Eurasia come to a halt and lock in with big time WSW/SW/SSW/S flow over the Western 1/2 of Russia as well as a large ridge promoting sun.


I am not saying this will happen and the snow over a large area will just melt super fast.  But that WSW eventually veering S flow over the same regions for so long with strong sun picking up as well.  850s(1500m) get around 0c an above at times all the way to the valley South of the Kara with Southerly winds relentlessly.

Of course real melt will be hard to come by in the Valley but West of the ridge on the hills on the West side of the valley snow melt will be vigorous.


The faster it melts the faster the April sun which is modest to strong at the end gets to work on unthawing the ground.  Greening up and warming up. 

The warmer the air from the W and S going towards the arctic circle the faster that snow will melt and the process feedbacks if driven.





The Western side is already well below normal.  The eastern side is above normal in the valley.

The valley would green up nice and quick if an early Spring arrives.

The snow on the West side is toast this upcoming week.

https://www.ccin.ca/home/sites/default/files/snow/snow_tracker/plot_anom_sdep.png

[Peter Ellis]

I think the interpretation of the plateau plot varies depending on what's causing the plateau. Is it an early re-freeze and a later start to the melting season (i.e. bulking up the "shoulders" of the plot), or is it a failure of the freeze right at the peak of the plot (i.e. chopping off the "head" of the plot).

This could easily vary between years.

[DavidR]

I think the interpretation of the plateau plot varies depending on what's causing the plateau. Is it an early re-freeze and a later start to the melting season (i.e. bulking up the "shoulders" of the plot), or is it a failure of the freeze right at the peak of the plot (i.e. chopping off the "head" of the plot).

This could easily vary between years.

It could, however 2015 had a fairly  normal  refreeze up to the max being consistently below the post  2007 average, then it  stopped.  In 2015 there hasn't been a specific cold snap  across the Arctic that drove the extent  out of the current range.

The current plateau will undoubtedly be the longest in recent  memory.

The plateau for average extent is 37 days. The rapid freeze peak is the outlier not the norm.

[crandles]

http://globalcryospherewatch.org/state_of_cryo/snow/fmi_swe_tracker.jpg

Ooooh, I like that. Can that be put on the graphs page. Perhaps also


http://globalcryospherewatch.org/state_of_cryo/snow/fmi_swe_tracker.jpg

latter one can be compared to last years version at
http://www.globsnow.info/swe/GCW/GCW_Snow_Watch_plot_20140616_SWE_NRT_v1.3.png

v2 seems to stop in 2013.

[viddaloo]

The current plateau will undoubtedly be the longest in recent  memory.

The plateau for average extent is 37 days. The rapid freeze peak is the outlier not the norm.

DavidR, not sure I understand what you say about these 37 days? Clarify?

As I understand it, the record–low plateau of 2015 is caused by ocean heat content. Trend seems to be towards earlier abandonment of the plateau for lower (and thus longer) plateau years. Yet I guess the date for leaving the plateau at the 'right shoulder' is less important for melt than the (assumed) fact that ocean heat is running the show? If it is, and we're already low at the plateau, we will go very, very low.

[Neven]

Ooooh, I like that. Can that be put on the graphs page.

Scroll all the way down, been there since a month or so. 

Perhaps also


http://globalcryospherewatch.org/state_of_cryo/snow/fmi_swe_tracker.jpg

latter one can be compared to last years version at
http://www.globsnow.info/swe/GCW/GCW_Snow_Watch_plot_20140616_SWE_NRT_v1.3.png

v2 seems to stop in 2013.

That's interesting, they had a snow water equivalent map there with the red trend line going off the charts, but it's not there anymore. I discussed these high values a couple of weeks ago with some people. Must've been wrong then.

I'll update and include that graph soon. Thanks for the heads-up, crandles.

[viddaloo]

Even with yesterday's fairly big gain, we're still looking at a possible weekly melt of 677 km² if the last 3 days melt like 2011 did:

[Nightvid Cole]

Even with yesterday's fairly big gain, we're still looking at a possible weekly melt of 677 km² if the last 3 days melt like 2011 did:

How can you say "0% of 2015 max extent gone" (in caption) when we have not yet matched the Feb 15th peak?

[viddaloo]

How can you say "0% of 2015 max extent gone" (in caption) when we have not yet matched the Feb 15th peak?

Well, it's not exactly me saying it, I just press 'reload' on my PHP script.

0% gone means less than 1% of the extent is gone. Mystery solved? 

[DavidR]

The current plateau will undoubtedly be the longest in recent  memory.

The plateau for average extent is 37 days. The rapid freeze peak is the outlier not the norm.

DavidR, not sure I understand what you say about these 37 days? Clarify?

As I understand it, the record–low plateau of 2015 is caused by ocean heat content. Trend seems to be towards earlier abandonment of the plateau for lower (and thus longer) plateau years. Yet I guess the date for leaving the plateau at the 'right shoulder' is less important for melt than the (assumed) fact that ocean heat is running the show? If it is, and we're already low at the plateau, we will go very, very low.

Vid,
As well as recording individual years I also calculate the average daily extent for each day of the year since 2007.  Measuring the 200K plateau on that gives a reading of 37 days.

Crandles has demonstrated that the correlation disappears if you take the measurement  back to 1988. I've got some ideas around that, that I am going to explore over the weekend. 

[Neven]

That's interesting, they had a snow water equivalent map there with the red trend line going off the charts, but it's not there anymore. I discussed these high values a couple of weeks ago with some people. Must've been wrong then.

I'll update and include that graph soon. Thanks for the heads-up, crandles.

That graph was back, with the red trend line still off the chart, but now it's gone again...

[viddaloo]

Cool. My pet–project for the weekend will be to work out a 13–step index for a new prediction tool for yearly melt/minimum. Currently testing it for seasons already in the record, so that I have the correct answer and can trim the prediction tools to produce better and more educated guesses.

Guess I could make the index as simple as 1 = lowest daily extent and 13 = 13th lowest ... We'll see in May/June when this tool is applicable.

[Neven]

That's interesting, they had a snow water equivalent map there with the red trend line going off the charts, but it's not there anymore. I discussed these high values a couple of weeks ago with some people. Must've been wrong then.

I'll update and include that graph soon. Thanks for the heads-up, crandles.

That graph was back, with the red trend line still off the chart, but now it's gone again...

There it is again:



I don't know why that red trend line is so much off the charts, when this one is right on the average:

[Yuha]

I don't know why that red trend line is so much off the charts, when this one is right on the average:

Maybe it's because the second chart says "excluding mountains" and the map shows a strong positive anomaly around Himalayas.

[jdallen]

I don't know why that red trend line is so much off the charts, when this one is right on the average:

Maybe it's because the second chart says "excluding mountains" and the map shows a strong positive anomaly around Himalayas.

That would fit, with additional moisture mobilized by increased heat.

More snowfall generally fits with increased total heat in the environment.  Unfortunately, it is unlikely to create sufficient negative feedback to slow the changes we are starting to see.

[viddaloo]

JAXA dropped 35763 km² to 13848141 Friday, increasing 2015's lead over 2011 to 52874 km² less gained:



This is the 3rd biggest extent drop in March, and the biggest since March 4th.

[Neven]

I don't know why that red trend line is so much off the charts, when this one is right on the average:

Maybe it's because the second chart says "excluding mountains" and the map shows a strong positive anomaly around Himalayas.

Good one, Yuha! Should've thought of this myself, as I had noticed that a lot of snow had fallen in the Himalayas.

[plg]

Last update on the maxima (unless something extremely unusual happens):

Tomorrow is the latest day for maxima in all sattelite records for NSIDC (Mar 31), JAXA (Mar 31) and Cryosphere (Mar 29).

Current status:

• NSIDC: First place on Mar 27, 144k below current 2015 maximum, needs 253k for second place.
• JAXA: First place on Mar 28, 166.092k below current 2015 maximum, needs 351.761k for second place.
• CRYO: Second place on Mar 27, 180.9273k below current 2015 maximum, needs 223.5699k for third place.

So, it is very safe to assume that they are first, first and second respectively, and reasonably safe to assume thet the maxima already have been reached.

This is like watching some sporting event such as down-hill skiing, where you are never sure of the outcome until the end. On the other hand I never watch sports, so the analogy falters.

I guess we can be sure the melting season has begun.

[Espen]

At least you know something about down-hill

[plg]

Not really, but you are unavoidably exposed to it if you live in Scandinavia... At least you understand who won, it is not like cricket where you must be a fourth generation english with a twisted sense of scoring to understand the result. 

[viddaloo]

Week 13 Comparison: 2015 so far lost 79793 km² more than 2011.

[Frivolousz21]

The models show the Kara gets absolutely blitz as we go on the next 10-14 days.


With incredibly consistent modeled winds out of the WSWW to SSSW over the entire period.


With huge temp anomalies as well.

Of course there isn't likely to be any sort of surface melt over the sea ice this early.  Snow cover to the S/SW/WSW of the Kara will get completely smoked during this period. 

[crandles]

The models show the Kara gets absolutely blitz as we go on the next 10-14 days.

With incredibly consistent modeled winds out of the WSWW to SSSW over the entire period.

With huge temp anomalies as well.

Of course there isn't likely to be any sort of surface melt over the sea ice this early.  Snow cover to the S/SW/WSW of the Kara will get completely smoked during this period. 

Small area of above 0 forecast for Kara though that is 850 temp not surface. But even if that doesn't happen temps slightly below zero but sun shining can cause some melt or sublimation can't it?

[Shared Humanity]

...absolutely blitz....


...completely smoked.... 

I missed you.   

[Anne]

...absolutely blitz....


...completely smoked.... 

I missed you.   

Yep. The melting season is truly under way!

[epiphyte]

...absolutely blitz....


...completely smoked.... 

I missed you.   

Yep. The melting season is truly under way!

Indeed. One can already feel oneself starting to sweat in the sauna that is Friv's depiction of the arctic summer

But seriously - Any bets on the 5-10 day forecasts getting closer to reality than they managed last year?

[slow wing]

Haha! I enjoy your forecasts as well, Friv

[anthropocene]

But seriously - Any bets on the 5-10 day forecasts getting closer to reality than they managed last year?

It depends which forecasts you mean. The forecasts published by meteorological organisations - since I doubt a major step change in processing power or methodology has been made since last year then I would expect their forecasts to have the same quality as last year (As an aside, a reduction in the quality of the long-range forecasts may be a signal of a change in the climactic system).

If you mean Friv's forecasts on the ice getting "scorched" or "blasted" - we'll never be able to tell until Friv starts using scientific terms which means the statements can be tested. It may also take Friv to realise that for ice to melt any additional (heat) energy has to be transferred to the ice (i.e. since this is mainly by conduction the heat has to get in physical contact with the ice). Showing forecasts of what is happening at 500hPa or 800hPa shows little of what is happening at ice level. Also as Crandles quite rightly points out - large temp. anomalies but with the temp.  still well below zero won't melt much ice either. 

[Frivolousz21]

: Frivolousz21  March 29, 2015, 08:50:11 AM

    The models show the Kara gets absolutely blitz as we go on the next 10-14 days.

    With incredibly consistent modeled winds out of the WSWW to SSSW over the entire period.

    With huge temp anomalies as well.

    Of course there isn't likely to be any sort of surface melt over the sea ice this early. Snow cover to the S/SW/WSW of the Kara will get completely smoked during this period. 

Good to see you all as well.

[icefest]

but then again, warming up the central basin and decreasing refreezing in leads isn't going to make the ice much stronger.

[viddaloo]

2015 won the tight race for Week 13 by melting 122361 km² more than 2011 (but Week 12 was won by twice as much by 2011).

Oh, and the Plateau just ended. The one in 2015 was the longest in recorded history.

[jbatteen]

It depends which forecasts you mean. The forecasts published by meteorological organisations - since I doubt a major step change in processing power or methodology has been made since last year then I would expect their forecasts to have the same quality as last year (As an aside, a reduction in the quality of the long-range forecasts may be a signal of a change in the climactic system).

Well, they did actually give the GFS a major boost in processing power and resolution here in the last couple months.  The quality of the forecasts has definitely improved for my region, Minnesota.

That said, the sensor input network to the models is the weak link in Arctic forecasts.  The North American continent is covered in sensors and balloon launches, and as soon as incoming systems from the Pacific make their way onto the sensor network of the continent, their potential impacts further east in the continent become much more clear.  But the Arctic and most of the surrounding region save for Western Europe and North America is very poorly observed.  If we could carpet the North Pacific, the North Atlantic, Northeast Asia, and the Arctic itself in sensors, forecast reliability would dramatically improve.  As it is, I still don't trust any Arctic forecast more than 24-48 hours out and even that is liable to change in the details.

[Shared Humanity]

Just to be clear, when I said I missed Frivolousz21, I  meant that I really missed his regular posts on 5 and 10 day forecasts. I did because his forecasts always allowed me to focus on particular regions of the Arctic as the melt season progresses. Criticism about "scientific method" is ridiculous. Frivolousz posts weather forecasts, something I don't do, don't even know where to go to find them. It keeps me informed of, you got it, the weather forecasts which I value and is a service to this blog.

We have seen the effect of extreme weather on a particular melt season. Why would we not want to see an ongoing update of the forecasts?

I wish he/she would do the same through the winter.

[viddaloo]

Just to be clear: The 'plateau' around the 2015 maximum (Feb 15) is still ongoing, using DavidR's metrics of 5–day averages:

Code: [Select]

2003: 15066086 36 days (from 54 to 89) (max–min: ?)
2004: 14701388 23 days (from 51 to 73) (max–min: 9017725)
2005: 14396094 17 days (from 60 to 76) (max–min: 9216794)
2006: 14132380 13 days (from 63 to 75) (max–min: 8507334)
2007: 14209677 30 days (from 46 to 75) (max–min: 10143938)
2008: 14774776 27 days (from 56 to 82) (max–min: 10274153)
2009: 14657047 27 days (from 60 to 86) (max–min: 9602992)
2010: 14688540 33 days (from 65 to 97) (max–min: 10066448)
2011: 14127729 42 days (from 52 to 93) (max–min: 9858530)
2012: 14709086 30 days (from 63 to 92) (max–min: 11531631)
2013: 14523635 26 days (from 57 to 82) (max–min: 9714347)
2014: 14448416 10 days (from 74 to 83) (max–min: 9564296)
2015: 13942060 47 days (from 42 to 88) (max–min: ?)
In my graph above I use daily values, not 5–day means, which means we'll earliest be off the plateau in a day or two, given further drops.

I doubt it will last a monumental 50 days. In any case, it seems to be a harbinger of a GREAT melt. Torching or blasting? Perhaps! 

[Frivolousz21]

The models just lay the smack-down on the Western 1/2 of Eurasia the next two weeks. 


Snow cover is going to get totally crushed.

The forecast is calling for above freezing highs in the valley South of the Kara everyday.

There is already a very very large region over the Western side of Eurasia that is about to  vanish off the Rutgers/nic maps where it's barely 50%.

[epiphyte]

Just to clarify - it was the weather-center forecasts I was asking about !

... Especially ECMWF - I know they upgraded their compute last year (they replaced their IBM gear with two Cray XC30s, tripling their performance from 70-200 sustained Tflops running their model).

Aside - twenty-something years ago I was the Cray on-site analyst at ECMWF. My first task there was to install the OS on their new 8-processor Y-MP. At that time it was the most powerful computer in Europe (that anyone was willing to talk about, at least). Their new cluster is >100,000X faster than that.