ASI 2015 update 2: take me to the other side

During the melting season I'm writing (bi-)weekly updates on the current situation with regards to Arctic sea ice (ASI). Central to these updates are the daily Cryosphere Today sea ice area (SIA) and IJIS sea ice extent (SIE) numbers, which I compare to data from the 2005-2014 period (NSIDC has a good explanation of sea ice extent and area in their FAQ). I also look at other things like regional sea ice area, compactness, temperature and weather forecasts, anything of particular interest.

The animation on the right consists of NSIDC
sea ice concentration maps, one for each ASI update.

Check out the Arctic Sea Ice Graphs website (ASIG)
for daily updated graphs, maps, live webcam images and
the Arctic Sea Ice Forum (ASIF) for detailed discussions.

June 4th 2015

Usually, not always, if one thing is going on in the Arctic, the opposite is happening on the other side. In fact, something like it - albeit in the longer term - is mentioned in the latest NSIDC monthly summary:

[A] new study shows that when winter sea ice concentrations are above average in the East Greenland, Barents and Kara seas, ice concentrations tend to be below average in the Bering Sea. This spatial pattern of anomalies linking the North Atlantic and North Pacific is related to the sea level pressure pattern that drives surface winds and their associated movement of atmospheric heat. These conditions are in turn linked to cooler or warmer than average sea surface temperatures that provide memory, influencing regional sea ice concentrations the following autumn. Thus, while the atmosphere is critical in setting the spatial patterns of sea ice variability, the ocean provides the memory for reemergence.

As I wrote two weeks ago in ASI 2015 update 1, the melting season started early in the Beaufort and Chukchi Seas, but while the multi-year ice was getting its first beating of the melting season, the ice in the East Siberian and Laptev Seas was still in deep freeze. This resulted in relatively thick ice (see the latest PIOMAS update released earlier today) and a delay when it comes to melt onset in that quarter of the Arctic.

But as always, things have started to flip, and the action is being transferred to the other side. Below I'll explain the details and the forecast for next week. Also be sure to read RobertScribbler's take on the matter when you're done reading this update, as he also covers snow cover, which, to quote from this article in the Washington Post, "unlike sea ice, gets no attention".

Sea ice area (SIA)

Although Cryosphere Today has been updating on a daily basis again, I've taken the liberty of using the latest two pre-calculated data points by Wipneus so that I can show the June CT SIA map:The 2015 trend line is clearly in the bottom half of the pack, but it refuses to follow the order that has been clear on sea ice extent graphs for several weeks now, as seen below.

Sea ice extent (SIE)

IJIS SIE data is being reported by the Japanese Arctic Data archive System now (check out their website, they have some great visual maps), and here's how things look right now:The 2015 trend line is well clear of all the other trend lines. In fact, I had to adjust my May graph when I came back from my holiday to accommodate it. It will be interesting to see whether the trend line can keep up this tempo throughout this crucial month.

Cryosphere Today area per IJIS extent (CAPIE)

When SIE drops relatively fast, and SIA less so, CAPIE naturally goes down slower. Remember, this percentage is calculated by dividing SIA numbers by SIE numbers. Because open water between floes are counted for SIA, but not for SIE, it will always be lower. Added to that, melt ponds fool satellite sensors into thinking there is open water where there is none. Open water in grids gets counted as such for sea ice area, but for sea ice extent there has to be a lot of open water for all of the grid to be counted as open water. This explains why there are two different ways of measuring the amount of sea ice based on sea ice concentration data (read the CAPIE segment in ASI 2014 update 2 for a more thorough explanation). 

So far this year CAPIE has been relatively high, as can be seen on the graph (and also on the compactness graph that Wipneus updates, using more sophisticated data):

The trend line has started to drop with CT SIA noting a couple of bigger drops, but it's still higher than all other years in the 2006-2015 record. This means that the ice pack is either very compact, without much of a mix of ice floes and open water in the marginal ice zones, or that there is not much melt ponding going on. It's probably a combination of the two, but given the assumed importance of melt ponds for preconditioning the ice pack for the rest of the melting season, this might be a first sign that it won't be easy for the 2015 melting season to force itself into the top 3.

But June is the month when most of this melting momentum is built up, so it's too early to tell.

Regional SIE and SIA

This year I'm using the regional SIE and SIA graphs on the updated Arctic Sea Ice Graphs website, as provided by commenter and amateur Arctic expert Wipneus, who also sends me custom-made maps showing the most interesting region of the fortnight.

The regional map of the week is of Hudson Bay, as that's where a lot of action has been taking place in the past two weeks (red = less ice than two weeks ago, blue = more ice than two weeks ago):

As can be seen on the Uni Bremen sea ice concentration maps comparison page it's unusual for western Hudson Bay to clear out so early and fast. This is one of the main reasons that IJIS SIE is so much lower than other years, which means that the decline rate could stall when the Hudson is clear of ice. But other regions are of interest as well, with Hudson Bay on top:

The effect of the high temps and insolation of the past couple of weeks can clearly be seen on the Beaufort and Kara Sea maps, with the Chukchi going down fast as a prolongation of the rapid clearing of the Bering Sea.

Here are the regions that are acting as a counter-balance, the East Siberian and Laptev Sea as mentioned earlier, but also a Greenland Sea with little sea ice is a sign of little Fram Strait export (counter-intuitive, I know), as we saw during the past two melting seasons:

Again, all these regional graphs and more can be followed on the ASIG.

Weather conditions

As usual we look at sea level pressure (SLP), followed by air and sea surface temperatures. First the two-week animation of Danish Meteorological Institute SLP images:

It's clear how high pressure areas opened the windows on the Beaufort and Chukchi Seas, letting the Sun shine in. But without concomitant low pressure areas on the other side there hasn't been all that much wind and thus movement, just as we saw during all of the 2014 melting season. Nevertheless, some preconditioning has been taking place in a zone that's rife with thicker, multi-year ice.

In the final frames of the animation a big cyclone is moving in via the Barentsz Sea. Let's check out the 6 day forecast according to the ECMWF model to see what's in store (click for a larger version):

The cyclone is going to move over the centre of the Arctic and stay put for a couple of days. Not particularly low in pressure, but pretty big nonetheless. This will make large parts of the Arctic cloudy, but at the same time cause winds that might pull at the sea ice that hugs the Siberian coast, and suck in warm air from the Siberian mainland that is also going to obliterate the snow cover there. Usually this kind of weather causes the rate of decline to slow down, but further losses in Hudson Bay where the Sun keeps shining for a while longer, might keep the rate stable.

Air temperatures play an important role as well. As said, May started out cold, but warmed up towards the latter part of the month, as can be seen on the DMI 80N temperature graph where ultra-low temps in the region above 80° slowly changed into the above average temperatures we see now, last seen in 2011 and 2012:

Here's a short animation of NOAA/ESRL 1-day anomaly maps showing the changes compared to two weeks ago:

The Alaska 'heat wave' has finally dissipated, with a lot of heat popping up in Siberia now. Here's what GFS temperature forecast maps provided by the excellent Climate Reanalyzer website are showing for the coming 7 days:

Above freezing temperatures, almost Arctic-wide, with Eastern Siberia heating up big time (again, read RobertScribbler) as well as the land masses surrounding Hudson Bay. It's going to be interesting to see how this all plays out.

Sea surface temperature anomalies on the Atlantic side of the Arctic are increasing some more, especially in the Barentsz Sea, but there's some deep red showing up in Bering Strait as well, all well ahead of last year:

Conclusion

The Arctic has stopped being cold and is projected to stay warm for the coming week. At the same time it seems there is not much melt ponding happening, which is a crucial factor in building up melting momentum for the rest of the melting season. Extent numbers are very low, and area is catching up, because of early, fast and continued melt in Beaufort, Chukchi, Kara and Hudson, but on the other side the East Siberian and Laptev Seas are still asleep. That's where the action is moving to now, so let's see how many thousands of square kilometres of melt ponds can be squeezed in during Junction June.

Later this week I'm going to try and have a closer look at weather conditions in May 2015 compared to other years. In the meantime, I can highly recommend this blog post by Chris Reynolds on his Dosbat blog that discusses what's been going on this past May across the Arctic: Four Dipoles.

Enjoy the Arctic sea ice, it's the only Arctic sea ice you've got.

Comments

[Neven]

For those who like that kind of thing (I know I do), I have opened two polls on the Arctic Sea Ice Forum, one for the CT SIA daily minimum and one for the NSIDC SIE September minimum (monthly average).

They're at the top of this list of threads here.

[navegante]

CAPIE values should drop during the cyclone

http://goo.gl/muUvfj

Also, any thought on the potential Ekman pumping due to this storm, or not relevant at this time? I am no expert at all

[Neven]

I don't think it's relevant right now, navegante. I believe Ekman pumping requires (lots of) open water.

It's not that strong a cyclone anyhow, but large, yes.

[NeilT]

I'm wondering if this is correct... It doesn't look like it from the webcam.

http://seaice.alaska.edu/gi/observatories/barrow_sealevel

[Jim Hunt]

The Barrow ice mass balance buoy is a few miles north of the webcam location, so it's not out of the question. Here's a picture of Barrow beach from May 26th:

http://forum.arctic-sea-ice.net/index.php/topic,1149.msg53090.html#msg53090

You will notice a fair bit of liquid around even then. They are working out the position of the bottom of the ice using the thermistors, which gets tricky once the floe temperature is much as the same as the water underneath it!

[wayne]

There is nothing in science like repetition,
the horizon shifts in High Arctic Canada were captured many many times near Highest latitude point of USA: Barrow Alaska... You can study it in great details thanks to University of Alaska Archives.

http://eh2r.blogspot.ca/2015/06/from-barrow-strait-to-barrow-alaska-sea.html

[Bill Fothergill]

An interesting announcement has recently come from NOAA pertaining to the so-called hiatus.

http://www.ncdc.noaa.gov/news/recent-global-surface-warming-hiatus

[Jim Hunt]

Speaking of which Bill, regular readers of "Tamino" will already be familiar with this graphic graphic of the statistical argument:

https://tamino.wordpress.com/2015/01/20/its-the-trend-stupid-3/

[Jim Hunt]

And Richard Betts reminds me of this animation from Ed Hawkins:

http://www.climate-lab-book.ac.uk/2013/what-will-the-simulations-do-next/

[wayne]

Indeed Jim, but to be more specific Global Temperatures variations are due to dynamical meteorology. Of which our greatest latest example in Arctic sea ice was 2013. When our not so bright fake skeptics, ie mainly annoying self proclaimed "pending next ice age" experts, decried sea ice recovery and as it was written "It got colder.....", all the while it really was cloudier by multiple Cyclonic incursions over the Arctic Ocean. The warming factor was still there, it morphed into these incursions which as some may know, were higher energy events, warming symptoms may not be measured only with thermometers. And so from fall 2013 there was an apparent "lull" in sea ice melting, only to be obviously a clever trick by nature to make contrarians look silly.

[Rob Dekker]

Thank you Neven, for again a great overview of the early melting season. I share your conservative view that the low extent (especially from IJIS) is probably not a significant indicator.

Low ice "extent" and average "area" suggests that the ice pack is compact and does not yet expose much water (melting ponds) in the ice pack, and thus amplification is limited.

Thus,assuming "average" summer weather, I believe that the early melting start may taper off over the next month and get more in line with extent numbers of 2011 and alike.

But the Arctic is notoriously unpredictable, and certainly has a few tricks up her sleeve.

Also, a big THANK YOU to Chris Reynolds, for the May summary post :
http://dosbat.blogspot.co.uk/2015/06/may-2015-status-four-dipoles.html

What a clear and wonderful read, explaining so comprehensively how the air pressure differences over the Arctic affect regional melting patterns.

Hats off !

[Rob Dekker]

Also, I'd like some clear mind advice on something.

I developed a super-simple model that uses Northern Hemisphere spring snow cover and a metric for melting ponds as a predictor for the September ice minimum :

http://neven1.typepad.com/blog/2013/07/problematic-predictions-2.html

This model works very nicely, as it obtains better correlation numbers than plain extrapolation of the long term trend.

For Northern hemisphere snow cover I use Rutgers Snow Lab monthly numbers, and for "melting ponds" (and other water close to ice) I use NSIDC numbers for (ice extent - minus - ice area) as a metric.

This year, as Neven reports, ice "extent" in May is running very low, while ice "area" is sort of average.

However, and here is the issue :
NSIDC does not seem to share that observation.
Here are the numbers :

..
2012 5 Goddard N 13.11 10.99 --> E-A = 2.12 Capie:A/E= 0.838
2013 5 Goddard N 13.08 11.20 --> E-A = 1.88 Capie A/E = 0.856
2014 5 Goddard N 12.77 10.99 --> E-A = 1.78 Capie A/E = 0.860
2015 5 NRTSI-G N 12.65 10.78 --> E-A = 1.87 Capie A/E = 0.852

Numbers obtained from here :
ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/May/N_05_area.txt

So it seems that NSIDC does not really rate May 2015 as specifically high on Capie index, nor out of the ordinary for the "extent minus area" indicator that I use as a metric for "melting ponds" in May.

So, for starters, I wonder why that difference in data sets (between NSIDC and IJIS extent and CT area) came about.

Second, the problem may be related to the way that NSIDC calculates their sea ice "area" numbers. From the same NSIDC txt file I linked above :

The "extent" column includes the area near the pole not imaged by the sensor. It is assumed to be entirely ice covered with at least 15% concentration. However, the "area" column excludes the area not imaged by the sensor. This area is 1.19 million square kilometers for SMMR (November 1978 through June 1987), 0.31 million square kilometers for SSM/I (July 1987 through December 2013), and 0.029 million square kilometers for SSMIS (January 2008 to present). Therefore, there is a discontinuity in the "area" data values in this file at the June/July 1987 boundary and at the December 2007/January 2008 boundary.

which does not really make sense to me.
Is the irregularity on "area" around the dec 2007/Jan 2008 boundary, or around December 2013 ? Or both ?

I'm confused. Anyone know what's going on here ?

[Neven]

Thanks, Rob.

I need to correct something I've said:

It's not that strong a cyclone anyhow, but large, yes.

I was wrong. It's pretty strong for a cyclone. Not GAC-2012 material, but not exactly weak either. I didn't look at the forecast careful enough. According to Environment Canada the cyclone is at 975 hPa right now, and still dropping:

[Jim Hunt]

The cyclone was down to 972 hPa in the 06:00 UTC analysis.

http://forum.arctic-sea-ice.net/index.php/topic,1149.msg53445.html#msg53445

Rob's point about snow cover makes we wonder about butterflies flapping their wings in Siberia. How much precipitation will this storm produce, and of that how much will be snow and how much will be rain?

And of course how much will it churn up the ice!

[Bill Fothergill]

Rob D wrote...

" ... which does not really make sense to me.
Is the irregularity on "area" around the dec 2007/Jan 2008 boundary, or around December 2013 ? Or both ?"

It looks as though someone at NSIDC has done a bit of cut&paste without checking if the paragraph still makes sense.

I was aware of the "pole hole" in coverage, but it has been a couple of years since I actually bothered to read the caveat. The version I have pasted into my spreadsheet states that...

" ... This area is 1.19 million square kilometers for SMMR (from the beginning of the series through June 1987) and 0.31 million square kilometers for SSM/I (from July 1987 to present). Therefore, there is a discontinuity in the "area" data values in this file at the June/July 1987 boundary."

At that time, the correction was pretty unambiguous: one added 1.19 million sq kms to any Arctic area value up to, and including, June 87; after that, one only added 0.31 million sq kms.

I would hazard a guess that, from Jan 2008 onwards, the offset value reduces to 0.029 - but it could certainly have been phrased better.

Does that help?

I have certainly found their helpdesk extremely helpful whenever I have had a query.

cheers bill f

[D]

Yes, Neven, it is a strong cyclone for June. In 2013 a cold stormy June led to recovery of the ice pack. Stormy Junes over the Arctic ocean are generally associated with recovery years. However, I think we need to be careful to look at the jet stream patterns. In June 2013 the jet stream collapsed around the Arctic ocean keeping cold air in and warm air over the continents. This year the jet stream is very wavy, pulling warm air into the pole.

We need to be careful to look at the details and watch out for generalizations.

Arctic amplification is based on more adsorption of solar radiation as snow and ice melts. Melt ponds are one part of that equation but so are ice extent and area of snow cover over land.

I still expect the large amount of excess heat in the oceans around the Arctic to keep the melt going at a record rate. The rapid loss of snow this spring first on the Alaskan side, now on the Siberian side is contributing to Arctic amplification this summer. However, weather and ice are complex and hard to predict. A stormy June might slow down the progress of melting. It might buy the ice records a little more time. We'll see.

-George aka FishOutofWater

[jdallen_wa]

Indeed, FishOutofWater.

This isn't the GAC of 2012, but it also isn't the ice-saving weather of 2013.

Currently according to Climate Reanalizer there is rain falling across the Laptev. The next three days show much of the same, with warm wet air streaming across the Kara through the Laptev to the ESS bringing rain and accompanied by temperatures of 1-4C. The remaining snowpack is being obliterated.

In short, there is nothing about this storm which favors the ice.

[Wipneus]

Rob Dekker,

That data file does not give you what you think it does. Instead of average {extent|area} computed from daily sea ice concentration, it is {extent|area} calculated from average monthly sea ice concentration.
Think of it: even if area equals extent every day of the month - 100% concentration within the ice extent, never any melt ponds or leads- grid cells that are only covered with ice during part of the month lead to an average concentration less than 100%. Thus monthly area < monthly extent.

I think, the monthly E-A may be at least biased by the change in extent in that month, explaining at least some correlation in an obvious way.

[Martin Gisser]

Jim Hunt,
indeed a graphic graphic! The paradigmatic graph to end all hiatus delusions. Horror of the Lomborgs of our world.

[Neven]

Cyclone is now at 971 hPa:

[Colorado Bob]

Seaweed colonizing ice-free parts of Antarctica

With glaciers melting, the original white, mostly lifeless Antarctica is now becoming darker and lively with seaweed.

http://earthsky.org/earth/seaweed-colonizing-ice-free-parts-of-antarctica

[Blaine]

@Neven, @Navegante:

To the extent that the ice is free-floating, it will not retard Ekman pumping. In fact, by transferring stress downwards, it will increase the depth of the coherently moving surface layer, which will increase the Coriolis deflection angle between the wind and the direction of ice motion, and increase Ekman pumping. Of course, the ice will also retard surface compression (which violates the free-floating assumption), which will decrease net Ekman pumping. The sign of net effect is not obvious, and I don't expect that it would come close to canceling out the effect altogether.

Ekman pumping due to cyclones is not usually significant for a completely different reason: timescale mismatch. Cyclones typically just don't last long enough to see significant thinning of the surface layer. Certainly a repeated series of cyclones in the same location could see a significant effect, but this doesn't usually happen.

Wave-driven vertical mixing on the other hand, is non-linear in the applied force and is severely depressed in ice-covered regions. So Neven's conclusion that vertical mixing caused by a cyclone is very much less where there is significant ice cover is absolutely correct, even though this is not caused by a an Ekman pumping effect.

In any case, there is very little heat available at middle depths at this time of year, so I would not expect to see a significant melting effect now.

What would be possible now is a significant increase of open water caused by ice compression. By causing surface divergence and convergence areas, the the storm should crunch some surface ice together. Ice area should go down with an increase in average thickness and no change in volume. This would of course result in an albedo decline and and increase in the melting rate. This would of course be balanced against the direct effects of the snow and cooler and cloudier weather caused by the storm itself.

[Rob Dekker]

Bill Fothergill said

It looks as though someone at NSIDC has done a bit of cut&paste without checking if the paragraph still makes sense. ... I would hazard a guess that, from Jan 2008 onwards, the offset value reduces to 0.029 - but it could certainly have been phrased better.

Does that help?

Yes, it does, and thank you for your reply.
I also noticed the difference between the notes in the old and the new files, and that the new 'area' numbers are adjusted upward (by indeed about 0.3) for 2008 onward.

So I think you are right that the new discontinuity is in 2008 only, and not in 2013. But I will contact the NSIDC helpdesk just to make sure.

Wipneus said :

Think of it: even if area equals extent every day of the month - 100% concentration within the ice extent, never any melt ponds or leads- grid cells that are only covered with ice during part of the month lead to an average concentration less than 100%.

I had to read that sentence a couple of times, but I understand what you mean. And yes, this may be a problem for even my simple model, and it may explain differences between the NSIDC's area numbers and other data sets.

Let me think about this a bit and then I will try to come up with a better method for determining "extent minus area" as a metric to determine leads and melting ponds.

Thanks !

[Wipneus]

Rob, I was tired glad you could figure it out.

Make sure that you did see my compactness graph at
https://sites.google.com/site/arctischepinguin/home/amsr2/grf/amsr2-compact-compare.png

Three different sea ice concentration datasets, computed by very different methods (not to mention the two different satellite sensors) all agree that the compactness in May (and currently) is unusually high.

Daily data for NSIDC is here:
https://sites.google.com/site/arctischepinguin/home/sea-ice-extent-area/data/nsidc_nt_nrt_main.txt
https://sites.google.com/site/arctischepinguin/home/sea-ice-extent-area/data/nsidc_nt_final_main.txt.gz

Data for the Jaxa L3 SIC is here:
https://sites.google.com/site/arctischepinguin/home/amsr2/data/Jaxa_AMSR2_L3_10km_Area_Extent-v0.0.txt

Data for the Uni Hamburg 3.125km AMSR2 SIC is here:
https://sites.google.com/site/arctischepinguin/home/amsr2/data/UH_AMSR2_3.125km_Area_Extent-v0.0.txt

[Neven]

Our friend has bottomed out at 970 hPa:

[navegante]

D, the persistent cyclone, as far as I remember, was weak and . . . persistent :) Came from Siberia like a thief in the night and stood for weeks ocuppying the Arctic. This one is more like SWAT from Northern Europe breaking into the house, so to speak :) Still we will see if the SWAT calm down and decide to stay for a coffee or two

[navegante]

Although the truth is, the forecast for Beaufort and especially Chukchi for the next five days or so does not look very conductive for melting. Lows, lack of insolation . . . the momentum there may be lost or delayed, the clock is ticking and sun is needed.
A great steady warm flow toward Fram may be developing though.

[RunInCircles]

Help me understand. One of the features of Cyrosphere today is their view of snow cover. I have never seen this much snow cover left in June. Other comments indicate snow cover is low in the northern hemisphere but this is not even close to what I am seeing on their map? Can somebody explain please?

[Chris Reynolds]

Thanks Rob, glad to see it was a useful perspective.

[Kevin O'Neill]

RunOnCircles:Look near the top of the Cryosphere Today Homepage on the lefthand side, above the global maps and below the charts; you'll find this note:

"Note: snow cover data not updating ... we hope to have a new data source by July, 2015."

[RunInCircles]

Thanks Kevin

[Werther]

Whatever ‘meltpond-june’ may or may not bring, the recent large Low has had an effect on ice distribution in the CAB.

This 500 km radius circle around the pole is from today’s MODIS, enhanced picture (bright -10/contrast +22/mid-tones -64).

It shows a lot of torn leads and open water, as the surface winds have dispersed while compacting elsewhere. It looks like the large swath of thinner ice in the ‘Laptev-bite-to-come’ has been messed up most.
I’ll post a larger version on the Forum.

[Neven]

Thanks for that, Werther. Awesome.

Things have slowed down in the Arctic right now, but the second round is about to begin.