Tuesday 23 June 2015

What role did the GAC of 2012 play in the 2012 crash?

A correspondent asks...
I don't know if you have already read and discussed conclusions presented in this article, but it suggests that GAC -2012 was not a major factor in record ice retreat. I thought it may be useful to the discussion on preconditioning of ice melt and summer ice cover.

"The impact of an intense summer cyclone on 2012 Arctic sea ice retreat". Zhang et al 2013.

That is a really good question, and I haven't blogged about why I think that the 2012 Great Arctic Cyclone (GAC 2012) played a major role in the 2012 melt.

What Zhang et al find, using the PIOMAS model is that "The simulated Arctic sea ice extent minimum in 2012 is reduced by the cyclone but only by 0.15 × 10^6 km2 (4.4%). Thus, without the storm, 2012 would still have produced a record minimum." So how can I possibly maintain that the 2012 Great Arctic Cyclone (GAC 2012) played a major role in the 2012 melt? This is especially when PIOMAS does such a good job, and I tend to view PIOMAS as the best available long term proxy for volume loss as a reanalysis product for sea ice. I have read Zhang et al 2013 several times, and I do get it. I have also read Parkinson & Comiso 2013, & Simmonds & Rudeva 2012.

GAC 2012 entered the Arctic Ocean around 3 August and eventually dissipated in the Canadian Arctic Archipelago on 14 August, its progress took it through the Arctic Ocean.

By 6 Aug 2012 it had neared its greatest intensity.

This state persisted until 8 August 2012.

By 10 Aug 2012 the cyclone was declining.

The period around 6 to 8 August is superimposed on the following graph of NSIDC Extent anomalies. Anomalies are the difference from an average seasonal cycle, here the average is for 1981 to 2010.

So that is basically it. Before the cyclone hit extent anomalies were roughly level, indicating losses around the average. This state followed an earlier June loss of extent. In general that pattern is similar to other recent summers, such as 2010, 2011, 2013 and 2014. However once the GAC 2012 hit a new regime of loss at substantially greater rates than the average started, and it started exactly when GAC 2012 hit.

When I look at the above graph, to me it begs the question of Zhang et al 2013: Did GAC 2012 really just cause 0.15M km^2 of loss? It's not a rigorously scientific position, evidence from a reliable model counters it, but when I look at what happened in terms of anomalies of sea ice extent, it is too much for me to swallow that the Great Arctic Cyclone of 2012 had little impact.


Neven said...

Well, one thing is for sure, all the extent and area records would've been broken, even if the GAC hadn't happened. It would've been different if the mega-storm had occurred at the start of September, but it happened one month earlier, meaning there was plenty of time left for the very vulnerable ice pack to sustain further losses.

The fact that the ice pack was so weak became apparent to me several day before the cyclone even showed up on the forecast (see my blog post at the time, ASI 2012 update 8: it shouldn't, but it does). I wasn't sure before that, but by then there was no other way to explain the steady rate of loss, despite weather conditions that weren't conducive to melt. And compactness was extremely low.

I've adapted your graph and drew a second trend line to show what 2012 probably would've been without GAC-2012, to be viewed here.

Whether GAC-2012 caused an additional loss of 150K or 250K or 350K, I don't know. But I'm pretty certain all the records would've been broken, just a bit later.

Nightvid said...

I have to say, I think that GAC2012 sped up a loss that was already "scheduled" to occur. As early as June 30, 2012 it appears that the ice pack was already pre-destined for melt-out in that area. To see this, compare the CT maps for 2012 vs. 2011:


As you can see, the concentration in that area was already bizarrely low at the end of June. As the season progresses but still prior to GAC2012, the concentration loss only becomes more dramatic. For example, on August 2, we have


At the very least, this should be strongly suggestive that the ice there would not survive the season. By early August, the concentration was only about 50% over a wide swath of that, and there were still 5+ weeks left in the melting season at that point.

Chris Reynolds said...


The adapted graph is something I had thought of posting, the reason I decided not to is that I couldn't be sure what role vertical mixing played in the drop after the immediate impact of the storm.


Precisely why June Compactness over the peripheral Arctic Basin is THE key metric. As I just said over at the forum, I agree that without the 2012 storm a record would have been set. I just find it hard to accept that the impact was as little as 0.15M km^2.

For comparisons with this year it might be worth looking at the peripheral seas (Beaufort, Chukchi, ESS, Laptev) compactness, hint hint. I am downloading the data from Wipneus's page every few days, but I am not commenting on the progress until the end of this month. So all I can do is hint... ;)

Nightvid said...


I suppose it depends on where you draw the delineation between "central" and "peripheral" areas. If you are using CT's boundaries, a lot of that is actually CAB, not peripheral seas. Of course if you drew the boundaries to make the CAB smaller (i.e. further into the Arctic Ocean), then that would be much less so.

Nightvid said...


Have you looked at the SIPN report yet?

Chris Reynolds said...

It's taken so long coming out I had given up checking.

Anyway I'm watching Motorhead at Glastonbury on BBC4 now, so it will have to wait.