Now that I've called the minimum, it's time to have a look at the differences between the two years with the lowest minimums in all data sets. This post will be a potpourri of different factors.
If we look at the six most important data sets - which IMHO are IJIS extent, NSIDC extent, Cryosphere Today area, Uni Bremen extent, DMI extent and PIOMAS volume - we see that this year a new record has been reached in three of them:
If we compare the sea ice concentration maps from the University of Bremen we can see how close 2007 and 2011 match, despite differences in final shape:
Though it doesn't immediately jump out, the biggest difference between the two is the date the respective minimums occured.
According to IJIS the 2007 minimum was reached on September 24th. The 2011 minimum was reached on September 9th (caveat: If I've called the minimum correctly; in theory it could go lower still). The difference is more than two weeks, as the melting season ended relatively late in 2007 and ended early in 2011. On the animation I have both minimums a day later (don't know why I did that), but it doesn't matter for the general impression we get from comparing the shapes.
It's also clear that there is a lot more low concentration ice on the 2011 sea ice concentration map than in the one from 2007 (more on that below).
This year's big arm in the East Siberian full of weak ice stands out in comparison with 2007. It's compensated by 2007's arm past Novaya Zemlya, ice in the Canadian Archipelago, in the Beaufort Sea and all along the edge of the ice pack on the Atlantic side of the Arctic Ocean.
Checking the Regional Graphs page we see indeed that 2011 is below 2007 in most graphs, except this one from the East Siberian Sea where that weak ice has probably been saved by the weather bell:
A difference of 400K is huge at this stage of the melting season and if we look at these JAXA-EORC sea surface temperature maps we can see a bit why the ice melted out there in 2007, but not in 2011. 2007 had more warm water on the Pacific side of the Arctic, where 2011 had more on the Atlantic side:
But of course this big difference in the East Siberian Sea isn't entirely due to a difference in sea surface temperatures, as compaction caused by winds is also a dominant factor in the last month and a half of the melting season for determining the final shape of the ice pack, as well as extent and area totals. The intensity and direction of winds depend on the way sea level pressure is distributed across the Arctic.
High pressure areas cause winds to blow in a clockwise fashion (and also make for clear skies). The opposite happens when low pressure areas, also known as cyclones on the Northern Hemisphere, dominate. When a high pressure system sits stably over the American side of the Arctic, including the Beaufort Sea, it will cause the winds to blow the edge of the ice pack inwards, towards Greenland and the Canadian Archipelago where the ice gets compressed. If this is accompanied by a low pressure system on the Siberian side of the Arctic (preferably between the Laptev and Kara Seas) a lot of ice can be blown out towards Fram Strait and then melt in the Atlantic. This set up called the Arctic Dipole Anomaly can cause a lot of extent and area decrease at the end of the melting season.
To see what the situation with regards to these weather patterns has been, I have retrieved images from the Daily Mean Composite page, compiled by the Physical Science Division of NOAA's Earth System Research Laboratory, and turned them into an animation that shows the final four weeks before the respective minimums in 2007 and 2011 were reached. Week 1 is the first image of that one month period, week 4 is the last week before the minimum.
I suggest looking first at one image, then the second, and then both if you'd like. Yellow, green and orange are for high sea level pressure, blue and purple are for low sea level pressure: It's clear that 2007 had a big high pressure system hovering over the Beaufort Sea and Canadian Archipelago, and a low pressure system on the Siberian side of the Arctic. It almost exactly fits the description above of perfect conditions that cause a lot of extent and area decrease. This is one of the main reasons that made 2007 such a special year with such a spectacular record.
In 2011 we see the exact opposite of 2007. Except for week 3 low pressure areas dominate the American side of the Arctic, so instead of ice compaction towards Greenland and the (wide open) Canadian Archipelago, we've mainly seen major ice divergence and hardly any flushing of ice through Fram Strait. If you consider this, it's actually quite amazing that 2011 has ended up in a virtual tie with 2007.
This difference between extreme compacting and diverging conditions also expresses itself on our custom made CAPIE or compactness graph, where Cryosphere Today area gets divided by IJIS extent and the resulting percentage tells us something about the degree of compaction (more info here):
Looking at the latest date we currently have, September 11th, we see that 2011 is 3.78% below 2007. That's quite a lot at this stage of the melting season. As the CT area numbers are practically the same, we can calculate what 2011's total extent would have been on this date, would the ice pack have been as compacted as that of 2007: 4,286,763 square km. 2007 had an extent of 4,343,438 square km (almost 57K more) on September 11th.
Finally, as low pressure areas bring in freezing temperatures towards the end of the melting season, we can also see the difference between 2007 and 2011 reflected in the NOAA/ESRL Physical Sciences Division temperature maps for September 10th: Conclusion
2011 had a flying start to the season, pretty much like 2010. But where 2010 came to a standstill around July 1st due to a dramatic switch in weather patterns, 2011 continued melting at a brisk pace for two weeks longer. A similar switch in weather patterns then slowed things down considerably (which can clearly be seen on the IJIS extent graph).
2011 recovered somewhat in August, but some very bad weather for extent and area decrease cut the season short abruptly (2010 had better luck with some weather patterns that extended the melting season and took it into third position in all data sets), much shorter than most recent years.
Where 2007 had an absolute perfect melting season, 2011 had some severe hiccups in the final 8 weeks of the melting season. Despite all of this 2011 beat 2007 on some of the graphs, and came within spitting distance on others.
Like I wrote last weekend in my guest post on ClimateProgress:
This is a sure sign that the ice is very weak and thin in large parts of the ice pack, which means that perfect weather conditions conducive to melting and compacting are no longer necessary to break records. The ice will melt out in place, regardless of what the weather does. That doesn’t bode well for years to come.It underscores the new abnormal in the Arctic: despite adverse weather conditions this melting season is on a par with the exceptional 2007 melting season. If these general circumstances persist, the Arctic will be very close to becoming ice free by the end of summer before 2020. Sooner, if we get a melting season with the same weather conditions as 2007.
No one knows what this will do to Northern Hemisphere weather patterns in the short run, and what happens with the permafrost and methane clathrates in the medium to long run. I don’t find that a particularly comforting thought.
Let's see what the coming winter and next year's melting season brings, shall we? Hopefully something unforeseen suddenly shows up to save the Arctic sea ice. And I'm expecting we finally get some more information out of the CryoSat-2 project.
The ice looked very thin this year. There were multiple lines of evidence for that (this whole post is one), but it'd be nice to finally have some hard observational numbers for the entire pack.