The blue line shows the Arctic sea ice extent for 2011 (May-August) and the dashed line for 2007. By comparing 2011 to the year with the lowest extent (2007), NSIDC tries to give the impression that this year is very low as well. The black line is the average for
19971979-2000, and the grey area is two standard deviations. They have been included to give the impression that there is some kind of "normal" Arctic sea ice extent, and that 2011 and 2007 is well below that extent. What NSIDC is doing is called cherry picking: they pick the data that support their thesis and discard the rest. Being an experienced data analysician, I can easily see through such ruses.
NSIDC get their money from federal agencies, and of course they want to keep the cash flowing even when the ice is not melting.