Regular visitors to Cryosphere Today will have noticed that the headline graph of sea ice area covering the whole Arctic has developed a funny little squiggle at the end, and the sea ice loss appears to have come to an abrupt stop, at least for the time being. So what is going on?
If you look at the various regional graphs, something emerges. Almost every area has flatlined at or near zero. It’s hard to reduce ice in areas where there is next to none. The Arctic Basin has increased its sea ice cover. two possible explanations: as it’s the first area where refreezing begins, and surface temperatures are at their lowest, it is possible that some consolidation of the ice pack has occurred, and there are fewer pools or polynyas in this area at the moment. The whole of the Arctic ice also drifts around a bit. With so much open water this year, it is likely that the currents have pushed what little ice was outside the area into the Basin. Indeed, if you then check out the Kara and Laptev seas, you’ll see that the ice area here has continued to decline; some from drift, and some from melt, perhaps.
Two other areas where there have been changes recently are the Canadian Archipelago/Beaufort Sea and the Greenland Sea. The first of these suggest that the ice melt is continuing, and pack ice is drifting towards the Pole. The second is a lot more worrying in its implication.
If you go back to the NSIDC illustration of the season’s ice movement, there’s a strong pattern of motion across the North of Greenland and around the corner, into the Greenland Sea. Using CT’s large file of satellite images, or the NSIDC image, a lot of this looks like multi-season (old) pack ice being flushed out of the Arctic and into the North West Atlantic.
All of this means that it is possible that we may have already seen the minimum ice area day pass. I’d expect, though, to see a few more days of losses at least from that Kara/Laptev area, so there may yet be a slight dip before the thaw ends and the refreeze proper begins.
This Winter, then, the sea ice area will go back to its larger value. How much the maximum will be is dependent on a lot of factors, but a recovery to near-average maximum seems less than likely as things stand. And much more of the Winter ice will be first-year ice. Vulnerable ice. Even if the Winter was to see a return to nearer-normal average areas, the picture for next Spring’s melt season looks ominous already. Unless another major variable comes into play, and suppresses the sub-Arctic SST anomaly, reduces the Bering inflow, or cools the whole region by a few degrees compared to the past six years, then ice loss will be rapid and comprehensive early in the season.
It is not clear what impact this will have on the Northern Hemisphere’s climate or weather, nor what the feedback effect is likely to be. I can’t see that it’s likely to be good for us.
All of this is pure speculation. Unless it happens.
Stay cool.
5 comments
Comments feed for this article
September 5, 2007 at 11:17 pm
Steve Bloom
NSIDC states that the rate of extent loss was back up again as of Sunday, well after CT started showing a flat trend. Of course area and extent will never track each other perfectly, but I assume the sort of difference we’re seeing now is why NSIDC considers area to be less reliable. It does seem reasonable to expect the timing of melting and reformation to be different with so much missing ice. A particular point I wonder about is whether the winter ice will be less stable with so much of it being thin; i.e., will we see continued flushing of older ice, polynya formation, etc. even in the midst of the winter season?
September 6, 2007 at 12:02 am
fergusbrown
What is likely to happen with first year Winter ice is that it will get mashed around by storms and currents and pounded into weird shapes and lumps. In response to this, the CT graphs should show some wild fluctuations in area after heavy storms in places where there is still some sea left. It is imaginable that a series of Autumn/early Winter storms would hamper reformation to a certain degree.
The flushing of the older ice will have to depend on prevailing wind and currents. The worry is, most of it is sitting right near the Denmark Strait, ready for the Spring. It’s likely to be pretty fast during the cold season.
Then there’s the AO to consider…
There are some pretty pictures here, a link I posted on Eli’s blog:
http://www.seaice.dk/iwicos/
September 6, 2007 at 3:36 pm
llewelly
If you look at CT’s long anomaly graph you can see that substantial fluctuation is more or less normal at this time of year. In fact, short-term variation is normal all year long, but the sinusoidal shape of the seasonal variation hides the fluctuation during the melt season and the freeze season – it is only at the peaks on the troughs that the fluctuation is obvious. But if you look at an anomaly graph, as I link above, the short term fluctuations are visible at all times.
By the way, I’m upset that your blog fails to provide preview. html-like codes like links and line breaks are interpreted differently by different blogs, (there is a standard, but I guess they don’t care) so preview is essential. Or do you prefer sloppy comments?
September 6, 2007 at 3:46 pm
fergusbrown
Thanks for the link to the ‘tale of the tape’; I look at it regularly, as it provides more detail than the main graph, but I should have provided a link myself.
On the preview option: I’ve no idea whether that’s my fault or wordpress’s; I’ll look at my control panel and see what I can do. As I only get a few comments a day, though, it’s no great problem to edit them if needed, though I avoid that if possible. I’m a writer rather than a computerer, but there’s no real excuse for not knowing these things.
I tend to write the blog on Firefox, but read it often on IE: the firefox version is prettier, but not much.
regards,
September 7, 2007 at 2:00 am
llewelly
Thank you for looking into the preview issue. It seems to me that some months ago most wordpress blogs had preview, but now most do not. I miss it. As far as browsers go, my computers run linux and freebsd, so they do not run IE.