German scientists have just said that Arctic sea ice set a new record low this year, but the US National Snow and Ice Data Center has released its own measurements showing a slightly higher sea ice summer so far . The difference between the two observations is just 100,000 square kilometers--4.24 million on September 8 from Germany, 4.34 million on September 9 from the US. The NSIDC says,
The rate of decline has flattened considerably the last few days: Arctic sea ice is likely near its minimum value for the year. However, weather patterns could still push the ice extent lower. NSIDC scientists will make an announcement when ice extent has stopped declining and has expanded for several days in a row, indicating that the Arctic sea ice has reached its lowest extent for the year and has begun freezing over. During the first week of October, after data are processed and analyzed for the month of September, NSIDC scientists will issue a more detailed analysis of this year's melt season and the state of the sea ice.
This could easily be spun into a "did we or did we not set a record?" situation--and the internet loves new records* and its easy to turn things into a confrontation--but ultimately if 2011 set a new record low for Arctic sea ice melting in the summer or if it was just the second-lowest sea ice minimum on record isn't as important as noting that fact that the trend is still solidly towards ice free summers in the mid-term.
*As for new records, NSIDC lists the 2007 summer ice minimum as 4.17 million square kilometers, so by that measurement the 2011 German observation still wouldn't be a new low. Which is to say, there are different measurement methods yielding different results, even if the overall picture remains clearly defined.