Two weeks ago the Integrated Climate Data Center from the KlimaCampus of the University of Hamburg - which enables easy access to climate relevant observational data from in-situ measurements and satellite remote sensing - announced that they had derived 8-day composites of Arctic Ocean melt pond cover fraction on sea ice for May-September 2000-2011 based on MODIS data.
To quote the ICDC: "The melt pond cover fraction on Arctic Sea Ice is of particular importance for the process of every years' summer melt of the Arctic sea ice cover."
I've looked at the images from the database of the CliSAP/KlimaCampus Product (unfortunately not as nice-looking as the image above I copied from the ICDC melt pond page) and compared the melting seasons since 2007 at three different dates, based on the CAPIE graph:
One of the things that made the CAPIE graph useful in the past two melting seasons was that it gave us an idea about the amount of melt ponds on the Arctic sea ice. A lot of these melt ponds get counted as open water in sea ice area calculations, which makes sea ice area (SIA) go down faster than sea ice extent (SIE). This is the reason why SIE is preferred over SIA as a measure in during the period of greatest insolation and thus melt pond formation.
We divided SIA by SIE and the percentage allowed us to compare years on the graph above. Unfortunately, with the demise of the AMSR-E sensor we no longer receive daily extent data from IJIS, and thus cannot calculate CAPIE (which stands for Cryospheretoday Area Per IJIS Extent). Perhaps something else will come up as a replacement during the 2012 melting season.
The three dates for comparing melting seasons are June 6-7th, when melt ponds have started forming for real; August 9-10th, just before melt ponds start freezing over again; and July 8-9th, midway between the two other dates. Here is the result (click the image for a full version):
As expected 2007 really stands out in all three months, with 2011 and 2008 coming in second and third. Coincidentally more or less the same order of SIE and SIA minima.
I think this is going to be a very handy tool.
For more info on the technical details, read Rösel et al. 2011 (PDF).