Thank you Juan C!
Are JAXA numbers only going back to 2000? While 3,89 Mn km2 from 2013-2014 is the most conservative numbers for the last 20 years, how much is that value valid if we use the much more unlikely numbers for the early years1979-2000?
Here are the ADS (JAXA) numbers for all the years (1st. attachment).
It is interesting 2006, that was a very bad year on May, June and the first half of July, but it stopped the strong melting afterwards. 2020 will have 4.24M km2 at the end of the melting season, if it follows the melting of 2006.
Comparing 2005 and 2006, 2005 was the September lowest on record at his time and 2006 was in the path of being the new lowest on record, before it stopped its fast melting (2nd. attachment, NSIDC Graph).
As an anecdote, -in my opinion- 2005 & 2006 were the years that made some scientists think that the IPCC models were very conservative. They published their study on 2007, but it was before the huge melting of the 2007 summer. Of course, after the melt of 2007, this study and specially the graph, became famous.
I miss this graph, that compares the IPCC models with the real values (3rd. & 4th. attachments). I hope they will have it again to evaluate the new models (Sounds hard, but it is a fact to me: I don't trust the old & new IPCC models, if I don't have this graph to show how they really work).
This study was:
GEOPHYSICAL RESEARCH LETTERS, VOL. 34, L09501, doi: 10.1029/2007GL029703, 2007
Arctic sea ice decline: Faster than forecast
Julienne Stroeve,1 Marika M. Holland,2 Walt Meier,1 Ted Scambos,1 and Mark Serreze1
Received 15 February 2007; accepted 26 March 2007; published 1 May 2007.
P.S. The years in yellow have not ADS (JAXA) official data. Average used.
P.S. [2] I started on topic and I ended off topic.
I recommend that we follow this subject in "Freeform season chatter and light commentary":