« Emerging Research Questions in the Arctic | Main | ASI 2013 update 7: cold and cloudy »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Fairfax Climate Watch

here's some perspective on Serreze's outlook, compared to that of Wieslaw Maslowski, from SCIENCE VOL 337 28 SEPTEMBER 2012:

" “There are a lot of deficiencies in the state-of-the-art [climate] models,” says oceanographer Wieslaw Maslowski of the Naval Postgraduate School (NPS) in Monterey,California. If the rapidly declining ice volume is taken to be the better guide, he says, Arctic sea ice could be gone by the end of the decade. "

" Using “only ice extent is not sufficient if you believe volume can change much faster,” Maslowski says. Large uncertainties remain, he notes, but their extrapolation gives a date of 2016 for a nearly ice-free Arctic Ocean, with the end likely to come by the end of the decade. "

" The climate models don’t handle changing rules very well, but Serreze takes a shot at a date for an ice-free Arctic Ocean by adjusting the models’ projections downward by folding in current trends in ice area and volume. “I’m on record [as of 2007] saying 2030 is a reasonable time for” ice-free conditions, he says. "

" Serreze and others think Maslowski’s volume extrapolation exaggerates the problem. “It could happen [by 2016],” Serreze says. “I just don’t think so. I think he’s being too aggressive. ”


Very good article by Serreze, completely agree about his last years August 5 storm assessment (in fact I agreed about this last year after it happened). Also right on about cyclonic scattering and cooling, may I add only good for the summer period, is literally opposite during winter dark season. I however disagree a little about this years nearly stable cyclones, as opposed to moving ones, this season has been pretty strange with stable cyclones, but he is right that in summer there are usually lots of them.

Big point of contentions, the ice models; I wonder if we will ever see one in action? Like those lucky chaps have privy to modelling projections. Other is the effective look of sea ice as is right now,
the badly broken up state of the pack, this is highly unusual, boosted extent a great deal, although I think 2013 melting was great. An area calculation with a 2 km open water grid threshold (instead of 15%) perhaps would reveal this. The latest fad big map in my mind is with ice buoys:


The normal clockwise gyre current seems to have just restarted.
This is a significant change which will affect the final results.

Chris Biscan

Guess he has never heard of up-welling.

John Christensen

Great article, and thanks for the reference Neven!

With the choice of title, it seems Mark Serreze is setting the record straight compared to the 'sea ice mulching cyclones' story that came out late July..

Chris Biscan is raising a valid point though, and we have often seen the reference to the Ekman pumping phenomena.

With a number of well-defined cyclones this summer, there should be a lot to study in the coming months and years to determine net impact from cooler conditions, spreading of ice, overturning, Ekman pumping, precipitation, etc. that is caused by the cyclones.

This far, PIOMAS does not seem to indicate that the cyclones have caused significant Ekman pumping, since volume has held up as well as has been the case for SIA and SIE, but with continued overcast weather we will probably get better data in the coming weeks.


Chris B you serious?

He says that as the ice becomes thinner it doesn’t matter if there’s a storm or not. That's very much the point, and in accord with Neven's opinion (and of other contributors here).


I didn't want to post the whole article, as I want to direct some traffic to the Icelights blog, but in the final part there's some more interesting stuff:

"Statistically speaking," Serreze said, "summers with lots of cyclones have less ice loss than summers with fewer storms. That’s pretty clear." That’s what happened this past June. A stormy pattern slowed the rate of ice loss. "Having said that," Serreze said, "the impacts of an individual storm may not follow that rule, and maybe importantly, the rules are starting to change."

When a storm breaks up the ice causing ice sprawl, it accelerates ice loss because the darker spaces of open ocean water, absorb more solar energy and increase melting. "If you looked at it that way," Serreze said, "okay, I’d buy it. But that’s not the only thing that’s happening." Stormy patterns bring on cool conditions and more precipitation, which tends to increase ice extent. However, individual cyclones may start to change the rules, putting more emphasis on ice break up as a factor in ice loss. Scientists don’t quite know yet if that is the case.

Serreze warned, however, that at some point, the ice becomes so thin it doesn’t matter if there’s a storm or not. "It’s just going to melt anyhow," he said.

I thought we had already reached the point where the ice is so thin it's going to melt anyhow, but as usual, the Arctic decides. I'm not reaching any conclusions yet (as there's still a couple of weeks to go), but the start of the melting season seems very important for the way the rest of the melting season develops. It's difficult to compensate a slow start.

Rob Dekker

Mark Serreze said :

How much did the Great Arctic Cyclone of 2012 contribute to ice loss? Less than 5 percent, according to a study led by Jinlun Zhang and published in Geophysical Research Letters. The scientists point out that 2012's record loss was 18 percent greater than the previous low, set in 2007, meaning the record low was going to get there with or without the big one.

I greatly respect Serreze's opinion, and Zhang's paper results, but, to play devil's advocate, the question that will undoubtely be (and should be) asked is : If the GAC 2012 caused only 5 % ice reduction, then why is 2012 so much lower than 2013 ?

Rob Dekker

Could it be that we already entered the period of "increased variability", the start of which was predicted by GCMs for the 2030's :
but with 2012 materialized early in reality ?

John Christensen


When using the CT compare of August 1st 2012 and 2013 (http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=08&fd=01&fy=2012&sm=08&sd=01&sy=2013), you see that what made 2012 a record year happened before August - most likely the complete lack of cyclonic activity during late spring/early summer (if I recall right).

We have been studying the Arctic very intensely, but much more so after 2007, and it seems (any references anyone?) that the usual summer cyclones have been less frequent in later years.

Summer 2013 has seen the return of usual summer cyclones, but with less ice and more open water, it would make sense that we will see less of these cyclones in the future - unless land temperatures go up sufficiently to substitute for the rise in Arctic temperatures.

Rob Dekker

Thanks John, a reduction in spring / early summer cyclone activity over the past couple of year would make sense.
Do we have some statistical evidence of that happening ?

John Christensen

That's what I have been looking for as well, but even on this blog, it seems we are 'rediscovering' the summer cyclones.

It would be great if Mark Serreze would make an update to his article:

The Summer Cyclone Maximum over the Central Arctic Ocean
Mark C. Serreze and Andrew P. Barrett, 2008, Journal of Climate

They were investigating the period of 1958-2005, but since then it appears focus has been on individual cyclones and not the pattern.

Gerhard Trausner


Serezze may be right. The storms have kept the ice. But still, there is a crucial point. The SST. I dare say that if the SST would have been only 1 ° C warmer on average, everything would have looked quite different.
In the dissertation I read somewhere that the storms are periodic.
There was: every 7 years is a storm season.


First I would like to ask; what is Mark Serreze's definition of a cyclone? Does virtually every depression going below 1000 pHa count as a cyclone, or do you need to have a low pressure that genetares 25m/s winds?

What we have been discussing for the past month, and what many of us have found astonishing, is the strength and sometimes the persistence of the individual cyclone, and the number of which the really strong cyclones have been appearing. Last years GAC was the strongest summer-cyclone every recorded in the Arctic, and this year we have had three cyclones with a minimun pressure just 10 pHa weaker than GAC, one of whom swirled around for three to four weeks. If Serreze says that is not unusual, I simply do not believe him. Though, if he defines even a very weak low pressure system as a cyclone, and counts these strong storms as just another cyclone, then I can very much belive that there is nothing unusual with the sheer number of cyclones this year. But then again that is not the reason why so many of us has declared this season highly unusual.


I've been a professional MET for nearly 40 years now - with the last 30 spent 'watching' the arctic. The number,intensity and most specifically persistence of arctic basin cyclones is NOT that unusual. Just because ere have not seen this before may have more to do with the length of time many on this blog have been observing the arctic.
Eight years ago I hypothesized in a paper published for the energy industry that the arctic meltdown would quite likely lead to stronger storms in the fall initially, and later (as in 10+ yrs from now)in the summer. But 'frequent' & 'persistent' cyclones are a totally different story and have far more to do with global circulation patterns that have temporal life spans at the annual and decadel level.


Like doom, I was underwhelmed by the piece.

Cyclone commentary seems to bring out a lot of she said, he said divorced from statistically rigorous analysis, personal reminiscences about events no one experienced that occurred in a vast under-instrumented area.

That October 2012 article on the Great Cyclone that everyone disses is actually built off an exhaustive Arctic cyclone catalogue. I don't recall them actually providing a link to the database but it cannot be withheld from a request under the journal matériel clause.

It is reanalysis-based but the lag time is short enough that this summer's events would be included.

The database contains a half-dozen fields characterizing the cyclone, such as maximal low, pressure gradient, wind speed, duration, and so forth.

These have a lot of redundancy but upon principal component analysis, I wouldn't be surprised if the first component alone captured 80% of the covariance and could be used as a single number to quantitate and plot cyclone intensity over the years.


A-Team, August 4 last year:


August 12


Is hard with clouds to see a difference, but it doesnt seem too much. I rather agree with the idea that the ice was dispersed, and was already extremely thinned and or broken. It was also positionally apt to break up the ice more, as opposed to this year cyclones putting a stop to the Arctic Basin clockwise gyre. Overall 2012 and 2013 are similar but for stable anti-synergistic cyclones.


Summer brings clouds whose water vapor interferes with analysis, then and now, so take-home lessons of cyclones for the ice are not fully realized.

Despite incredible efforts of Danp on best-swaths and pixel by pixel cloud mask, which have gone retrospective over on the forum. For sure, the pipeline products we use have not found a solution to impenetrable clouds.

For example, what is going on here -- your choice of melt / divergence in an unexpected place vs cloud artefact implausibly affecting very different wavelengths:

 photo newOpen_zps1d71562e.jpg

Chris Reynolds

From the article:

“Statistically speaking,” Serreze said, “summers with lots of cyclones have less ice loss than summers with fewer storms. That’s pretty clear.” That’s what happened this past June. A stormy pattern slowed the rate of ice loss.


NSIDC Extent June and July losses (Jun 31 minus June 1) - 2013 is the greatest loss since 1989.

CT Area Losses, June and July is one of the top five greatest losses, very similar to 2011 and 2012, all of which are only beaten by 2007 and 1985.

So arguably 2013 is a year when summer ice loss beat the weather. Which raises the question; what kind of rout would there have been with weather like 2007 to 2012?

Sorry not got time to post graphs.


According to a 2008 NASA study, there has been a quite significant increase in storm activity in the Arctic:



Count me also as distinctly unimpressed. I basically expect that anything written by professional science journalists will read as if it was composed by a group of monkeys amusing themselves with one hand while typing with the other, but as director, the NSIDC is basically Serreze's own blog and he should at least have a correct and informative version there.

How much did the Great Arctic Cyclone of 2012 contribute to ice loss? Less than 5 percent...

This is a misrepresentation of the results of the linked study. Previous record, 4.13Mkm^2, new record, 3.41Mkm^2, amount contributed by the storm 0.15Mkm^2 = (0.15/(4.13-3.41)) = 26%.

Are the storms getting stronger? “No,” Serreze said. “Most of the evidence is actually showing that the frequency of stormy months is decreasing, favoring fair weather conditions. It’s not clear why, but it’s what’s been observed lately.”

There really is little consensus on this issue. See for example here or here. Even the most recent paper I can find from Serreze himself finds a slight, though not significant, increasing trend through 2005. Any decreasing trend is thus confined to 2006-2012 but not 2013. I really wonder if he's including 2013 in "lately" at all, or still discussing only through 2012.

“People seem to have this thought that all this storminess is unusual,” said Mark Serreze, an Arctic climatologist and center director at NSIDC. “Well it’s not. It simply isn’t. Summer is the time for cyclones.”

Really? From Serreze's paper, the average 500 millibar height pressure at the low near the pole during the "21 stormiest summer months" from 1958-2005 is in the 5400-5430m contour. From here, the average height there during the entire past 90 days is in the 5300-5350m contour, making the entire past 90 days stormier by this measure than his reference 1-month stormy periods.

It has also already been pointed out here that at least in his paper, his statistics use a cyclone count times duration measure which takes no account of storm strength.

Pete Williamson

Blaine the papers hint at were Serreze is coming from.

"The most obvious feature is strong variability."
"Arctic cyclone activity displays significant low-frequency variability"

If you understand what those sentences mean and then try to put the past two summer season into that context then you can see why it might be easy to be misled over the direction of the trend. It seems like a lot of climate science is like this, crudely put we need lots of data in order to pull any signal from the noise.

Just from the papers you link to it would seem safest to say there is no clear trend. We would have to see the detail behind his claim of a drop in cyclone activity to judge it in the context of others work. All that said I would love to see an arctic cyclone activity data set that continued through to present day, the recent years of low ice extent would give more insight. I actually emailed John Walsh (bad me for scientist bothering) to see if he had extend his CAI beyond the 2002 year in the publication, unfortunately not!

I really don't get why people are reacting so badly to what Serreze has said here.


Well there seems to be a dipole forming , on one side the CAA gets the cyclone, on the on the Northern Siberia gets the high pressure. Cold over the CAA, warm over Russian Arctic zone, is not the pattern we are familiar with. I think compaction will be greatest in the Russian Sector, there is a lot of open water to crush all the way to the pole.

The comments to this entry are closed.