Arctic Sea Ice

On persistent cyclones

As the persistent Arctic cyclone - or PAC-2013 - of the past couple of weeks winds down, I want to discuss what I've found on the subject in a couple of research papers. But first want to refer to two excellent blog posts from last week doing just that, on the Robertscribbler blog and FishOutofWater's blog on Daily Kos.

This animation of DMI SLP maps shows the birth and demise of PAC-2013:

Perhaps I should say near-demise, but the weather forecast says it's basically over for this cyclone. Of course, other cyclones come into play in other parts of the Arctic, but it's not the PAC-2013 anymore. After approximately 25 days of wandering around, weakening and re-strengthening to a pretty powerful cyclone at times, churning ice, keeping the Arctic cold, we've witnessed another remarkable cyclone in less than 1 year's time.

There are a couple of questions that have been on my mind ever since the cyclone started to show itself persistent:

1. Is PAC-2013 unprecedented?
2. Is it somehow caused by the progressive loss of Arctic sea ice?
3. If so, will we see these PACs more often?
4. What will the effect on the sea ice be, short-term and long-term?

To answer some of these questions I did a search on Google Scholar with summer arctic cyclone as search words. Because the Arctic is changing so fast, I was mostly interested in recent research, but I remembered there was one paper from 2008 that I read last year during GAC-2012, that also has interesting things to say about Arctic summer cyclones. So, I'll start off with that one and quote the most interesting parts.

New map on the block

Further down is a short update on the effects the SAC-2013 (Small Arctic Cyclone of 2013) has had so far, but first I want to draw some attention to a great new product that has recently come online, a sea ice concentration map with the highest resolution so far. When I started this blog the NSIDC used a 25 km grid, meaning that every pixel on their SIC map represented an area of 25 x 25 km, or 625 square kilometres, on the ground. IJIS had a 12.5 km grid and the University of Bremen even had and still has a 6.25 km grid.

Last year the GCOM-W1 (Shizuku) satellite was launched, with the AMSR2 microwave sensor aboard, and soon after that the Japanese space agency JAXA started to put out data. Alexander Beitsch from the University of Hamburg's KlimaCampus optimized the processing of this data, improving grid resolution to a stunning 3.125 km. This was part of his PhD in the framework of the IRO2 project that is being funded through the German Federal Ministry of Economics and Technology (BMWi). Sehr gut und danke schön!

There was just one problem with the sea ice concentration map that was produced, and that was that it was huge. 17 MB for just one image makes it difficult for us amateurs to play around with it. Luckily, the Arctic Sea Ice blog has some very smart bunnies on its member list, and it was Wipneus (well-known for his work on PIOMAS data) who managed to reduce the file size by ten times. For the time being he is putting those images on a separate Google Site. Thank you, Wipneus!

If this is real...

As usual, it's all about the if. Allow me to explain what this is about:

In the first Arctic Sea Ice update of the 2013 melting season that was posted a couple of days ago, I announced that a cyclone was forecasted to move over the Arctic Basin and stay there for a while. It's been there for a couple of days now as can be seen on this animation of Danish Meteorological Institute SLP images:

It's not particularly strong (especially not compared to GAC-2012), but it stays in the same spot for quite a while and so is bound to have some effect on the ice below. Now, according to the Naval Research Laboratory's ACNFS forecast model, this effect is quite pronounced. And to show you just how pronounced, here's their sea ice concentration animation from May 22nd, with a forecast up to June 6th:

2012 record not due to GAC-2012

A new research paper by scientists of the Polar Science Center of the Applied Physics Laboratory of the University of Washington, has been published this week in the online version of Geophysical Research Letters. It's called The impact of an intense summer cyclone on 2012 Arctic sea ice retreat and discusses the role the Great Arctic Cyclone of 2012 - or GAC-2012 - played in the incredible, unequivocal and unspinnable record streak
we witnessed towards the end of the melting season.

From the UW press release:

Cyclone did not cause 2012 record low for Arctic sea ice

It came out of Siberia, swirling winds over an area that covered almost the entire Arctic basin in the normally calm late summer. It came to be known as “The Great Arctic Cyclone of August 2012,” and for some observers it suggested that the historic sea ice minimum may have been caused by a freak summer storm, rather than warming temperatures.

But new results from the University of Washington show that the August cyclone was not responsible for last year’s record low for Arctic sea ice. The study was published online this week in Geophysical Research Letters.

“The effect is huge in the immediate aftermath of the cyclone, but after about two weeks the effect gets smaller,” said lead author Jinlun Zhang, an oceanographer in the UW’s Applied Physics Laboratory. “By September, most of the ice that melted would have melted with or without the cyclone.”

Recent research [covered on the ASI Blog here, N.] showed that the Arctic cyclone was the most powerful ever seen during the month of August, and the 13^th most powerful of all Arctic storms in more than three decades of satellite records.

“The storm was enormous,” said co-author Axel Schweiger, a polar scientist in the Applied Physics Laboratory. “The impact on the ice was immediately obvious, but the question was whether the ice that went away during the storm would have melted anyway because it was thin to begin with.”

The UW team performed the climate scientist’s equivalent of a forensic exam: They ran a computer simulation of last summer’s weather and compared it against a second scenario that was identical except that there was no cyclone.

Results showed the storm caused the sea ice to pass the previous record 10 days earlier in August than it would have otherwise, but only reduced the final September ice extent by 150,000 square kilometers (almost 60,000 square miles), less than a 5 percent difference. By comparison, the actual minimum ice extent was 18 percent less than the previous record set in 2007.

The storm apparently caused 150K km2 of extra melted ice and so the record(s) would have been broken anyway, as the difference with the 2007 minimum was much larger. The main reason the storm had such an effect, was that the ice was exceptionally thin. Just 10 years ago a storm, even of this size, wouldn't have had such an impact.

Read for instance this Skeptical Science article from October, with the figure below showing September Arctic sea ice extent data from NSIDC (blue) and years with Arctic summer storms similar to the 2012 storm depicted in red.

Remember that we started noticing that the ice seemed to be exceptionally thin weeks before the storm hit. It was the only way to explain the steadily rapid drop in extent and area levels, even though the weather wasn't conducive to that rate of melt.

Here's a quote from yours truly on July 28th, ten days before the Great Arctic Storm started to be great:

All Arctic storms, great and small

If I had to name one website that does the best reporting on Arctic sea ice, its disappearance and the implications thereof, it would have to be Climate Central. Climate Progress is an excellent source as well, but Climate Central doesn't let even the tiniest detail escape. Its senior science writers Andrew Freedman and Michael D. Lemonick cover everything people need to know about one of the major current events on our planet.

Take for instance Lemonick's latest article on this GRL research paper discussing the Great Arctic Cyclone of 2012. I'm using the name, coined first I believe at the time by commenter R. Gates, but I won't be comfortable with it until I haven't seen a repeat in the next couple of melting seasons. Things are usually called Great in hindsight, but what if we see more Great Arctic Cyclones pop up in coming years? Will the next one be called the Similarly Great Arctic Cyclone of XXXX, or worse, the Even Greater Arctic Cyclone of XXXX? But I digress...

Great it definitely was. We assumed as much, but now it has been confirmed by Simmonds et al. that the GAC-2012 is the most extreme Arctic summer storm on record. It even ranks 13th when you add all those fierce winter cyclones. Mind you, since 1979 no fewer than 19,625 Arctic storms have been recorded.

One other interesting piece of info from this paper is that apparently "the authors couldn’t establish that the unusually large areas of open ocean contributed to the storm’s intensity". I would think that all that open water - that was exceptionally warm around that time - acting as fuel for the cyclone is a no-brainer, but it seems it's not as easy to establish a link. Here's an image of DMI SST anomalies at the time of the storm that was used for ASI 2012 update 9. I can't help but think that all of that red must have had something to do with the intensity and longevity of this freak storm.

I guess we all remember how this monster looked (via AtmosNews):

Lemonick, however, wrote an even more interesting article around two weeks ago, called Arctic Storms: A Climate Danger Nobody’s Talking About. It's about Polar lows, winter storms, a phenomenon I learned about while researching back in August what kind of a storm the Great Arctic Cyclone actually was.

Here's the first half of his excellent article:

What role did the Arctic storm play in the record sea ice minimum?

There's another good piece on SkepticalScience that analyzes the past melting season (Tamino is on a roll as well with Arctic air temperatures, here and here). Here's an excerpt:

While the 2012 Arctic storm was a strong one, as Newman noted, it was not unprecedented.  To confirm Newman's claim, we examined daily surface pressure maps from NCEP-NCAR and NCEP-II DOE reanalysis data.  Low pressure systems are typically associated with a tight pressure gradient that results in strong surface winds.

In the 34 years between 1979 and 2012, we identified 19 deep low pressure systems (defined here as having a central pressure less than 980 hectopascals [hPa]) in the Arctic between 15 July and 15 September.  Most of these deep lows occurred over the Barents, Laptev, and Kara seas, so they are not suitable analogs for the 2012 storm.

However, there were nine previous storms during this timeframe in the same general area as the 2012 event.  Of these, we identified five best analogs for the 2012 event. These occurred in 1980, 1990, 1991, 1994, and 1997 (Figure 3 - click for a larger version).

Figure 3: NCEP/NCAR sea level pressure maps in (from right to left and top to bottom) on 07 August 2012, 28 August 1980, 12 September 1990, 06 August 1991, 17 August 1994, and 03 September 1997.

Read the whole article here.

ASI 2012 update 10: (wh)at a loss

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 IJIS sea ice extent (SIE) and Cryosphere Today sea ice area (SIA) numbers, which I compare to data from the 2005-2011 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 that can be of particular interest.

Check out the Arctic Sea Ice Graphs website for
daily updated graphs, maps and live webcam images.

August 25th 2012

I apologize for having provided so little analysis lately, but things are moving so fast that analysis can't keep up. Now I know what an IPCC regional model for the Arctic must feel like. ;-)

Basically, I'm at a loss for words, and not just because my jaw has dropped and won't go back up as long as I'm looking at the graphs. I'm also at a loss - and I have already said it a couple of times this year - because I just don't know what to expect any longer. I had a very steep learning curve in the past two years. We all did. But it feels as if everything I've learned has become obsolete. As if you've learned to play the guitar a bit in two years' time, and then all of a sudden have to play a xylophone. Will trend lines go even lower, or will the remaining ice pack with its edges so close to the North Pole start to freeze up?

Basically I have nothing to offer right now except short posts when yet another of those record dominoes has fallen. Hopefully I can come up with some useful post-melting season analysis when I return from a two-week holiday.

I'm at a loss at this loss. The 2007 record that stunned everyone, gets shattered without 2007 weather conditions. The ice is thin. PIOMAS was/is right.

Sea ice extent (SIE)

WindSat or no WindSat, the 2012 trendline on the IJIS SIE graph has dropped like a rock:

I had to adjust the Y-axis on this graph already once (compared to the one in the last ASI update), and it looks like I'll have to adjust it again. I already announced the new IJIS SIE record minimum yesterday, although it was based on the preliminary number. But after the revision the record is still standing and IJIS still hasn't stopped producing century breaks.

The current difference between 2012 and other years (without the unrealistic last data point that gets revised upwards) is as follows:

ASI 2012 update 9: stormy weather

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 IJIS sea ice extent (SIE) and Cryosphere Today sea ice area (SIA) numbers, which I compare to data from the 2005-2011 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 that can be of particular interest.

Check out the Arctic Sea Ice Graphs website for
daily updated graphs, maps and live webcam images.

ATTENTION: There are new polls in the right hand bar, closing August 20th.
One is for Cryosphere Today minimum daily sea ice area, the other for NSIDC minimum monthly sea ice extent (the one that is used for the SEARCH SIO projections). More info in this blog post.

August 11th 2012

Another two weeks have passed since the last ASI update, and my, what an exciting two weeks they have been. For me personally, the most exciting period since the inception of this blog. And perhaps the most disconcerting so far.
As I wrote two weeks ago:

For a fast decrease you need a strong high in the 1025-1035 hPa range, preferably over the Beaufort Sea. Or a nice cyclone in the 985-995 hPa range to stir things up and cause some of that flash melting that we love so much on the ASI blog because it looks so awesome.

What we got, was a huge cyclone in the 963 hPa range. If you've missed this mega-storm despite all the news stories in the mainstream media (yes, I'm being sarcastic), you can check out the Arctic storms category and see the day-to-day reports here on the ASI blog. Flash melting usually takes place from one sea ice concentration map to the next. This time it's days in a row and it still isn't over yet.

This highly unusual summer storm has massacred all that weak ice on the Pacific side and even caused a large swathe of ice floes to detach itself from the main ice pack. The Stronghold I wrote about a month ago is no more. Trend lines have dropped precipitously on every single sea ice extent and area graph out there.

What is important to remember, is the large discrepancies between this melting season and that of 2007. The latter had almost perfect weather conditions during almost the entire melting season, with lots of sunshine and compacting winds. 2012 had a reasonably good June, but after that the weather switched and hasn't switched back since. Those ideal weather conditions in 2007 pulled in large amounts of warm water through Bering Strait decimating the ice pack on the Pacific side of the Arctic. If 2012 is pulling anything through Bering Strait, it's cold water. The ideal weather conditions in 2007 pushed out very large amounts of multi-year ice through Fram and Nares Strait. 2012 doesn't come close to it. But still, even before the storm, 2012 was following the 2007 trend line as if the Arctic wasn't cloudy and winds weren't blowing the wrong way.

The fact that 2012 followed 2007 so closely, was the strongest evidence possible that a large part of the ice pack is so thin that it will melt out no matter what the weather conditions. The storm only emphasized this, giving all that thin ice a push into the abyss.

Keep that in mind when you see fake skeptics use this storm to mislead their audience into thinking that the storm, a freakish fluke, is the cause of however this melting season is going to end. Although I think they're going to keep quiet about the Arctic, like they are doing most of the time now. As if the Arctic isn't part of our planet. A part of paramount importance.

Sea ice extent

Ever since the switch from the defunct AMSR-E to WindSat the IJIS SIE graph is acting different than all the other graphs (see this blog post), but the storm has finally pushed the 2012 trend line in first position here as well:

The 2012 trend line is now below all the others, even 2007, for the first time since the start of the melting season. It noted 6 century breaks in the last 8 days, and it's not entirely sure whether it will stop there. We're looking at a very exciting end to the melting season with this graph.

The current difference between 2012 and other years (without the unrealistic last data point that gets revised upwards) is as follows:

Arctic summer storm open thread 1

As the storm is still there, big, but no longer strong, and comment threads fill up quickly, I figured it's time for an open thread. And yes, it's a bit of an update as well, with interesting comments and graphs.

But first the updated animation. Now that the swathe of ice floes in the East Siberian Sea has detached itself, it's time to slowly say bye-bye to the ice atoll in the upper left corner:

sea ice concentration maps courtesy of the University of Bremen

Here's a comment by Rob Dekker that I want to put the spotlight on, because it's very interesting and gets to the core of the consequences of this storm:

Further detachment

This isn't an official Arctic summer storm (ASS) update, but it comes damn close to one. The thing is that the last domino hasn't fallen yet, as the storm is still there, though less powerful.

In yesterday's Arctic storm update I mentioned how a large swathe of ice floes had almost detached itself from the main pack. Well, I think it's as good as completely detached now:

We already speculated about such an event back in 2010, the first melting seasson that was covered on the ASI blog, in a blog post called Breaking away from the pack:

Not so long ago, in one of the comment sections, we discussed whether it would be possible for a large piece of the ice pack to separate itself from the main body of ice in the Arctic Basin.

(...)

I'm not sure, but I think the ice pack falling apart with patches of highly concentrated ice breaking away from the pack could be a precedent in the satellite era.

(...)

What if that forecasted high-pressure system over the Beaufort Sea starts to develop for real next week and the Beaufort Gyre kicks into gear again, spurred by westerly winds? One would think the small patch(es) in the Beaufort Sea would get pressed into the central pack again, but perhaps the bigger patch in the East Siberian Sea would get swirled loose by the winds blowing westwards.

Okay, it took eastward blowing winds. Close enough. ;-)

I zoomed in a bit and made an animation of the recent changes on the Uni Bremen sea ice concentration maps: