Herewith some analysis of Jaxa sea ice extent numbers. The projections are not meant to be forecasts in any sense of the word. Rather they are attempts to test what melting is required from now to minimum to achieve certain results.
As at June 30 Melt As % of Resulting
Required 07/16 av. Minimum
For 2016 Result 5,013,380 101.1% 4,017,264
For 2012 Result 5,853,189 118.0% 3,177,455
For 2007 Result 4,964,905 100.1% 4,065,739
Av 2007-16 melt 4,960,046 100.0% 4,070,598
For 1 million km2 8,030,644 161.9% 1,000,000
It is obvious that a 2nd lowest miminum only requires remaining melt to be a modest fraction greater than the average remaining melt in the last 10 years.
For the magic 1 million km2 minimum, melt from now to minimum has to be 62% above the 2007-2016 average. This seems unlikely, given variations in previous years.
For a record low (below 2012), remaining melt has to be at least 18% above the previous 10 year average. This has only happened once in the satellite record - in 2012. My memory of 2012 was that it was the late season melt that made 2012 such a record record low. So I looked at the Jaxa record. In 2012, melt from Aug 1st to minimum was 35 % greater than the 2007-16 Average. In 2008 it was + 15%. No other year comes close.
Melt Aug1 to min Km2 cf with 07-16
1980's Average 1,588,705 -26%
1990's Average 1,502,583 -30%
2000's Average 1,728,320 -20%
2002 No data
2003 1,658,226 -23%
2004 2,181,360 2%
2005 1,721,140 -20%
2006 1,345,561 -37%
2007 2,116,693 -1%
2008 2,466,075 15%
2009 1,717,881 -20%
2010 2,036,131 -5%
2011 2,039,229 -5%
2012 2,904,945 35%
2013 1,833,934 -15%
2014 1,848,264 -14%
2015 2,211,140 3%
2016 2,304,571 7%
Av. 2007-2016 2,147,886 0%