I have a vague recollection of someone (a year or two ago) comparing early ice melt with season-complete ice melt and found no correlation. Can someone with data run a comparison and enlighten us?
(It's the 'only' hope we have of my 2013 projection of ice freedom in 2019 not coming to pass.)
There were some comments about this on NSIDC Analysis.
By memory, the lowest years on September are 2005-2007, 2010-2012 and 2015-2018. Looking at the ice on March on those years, the great exemption is 2012, that had more ice than 2010's average on March and end up being the lowest on September (and by far). But all the other years that were low on September, they were also low on March.
So, I don't think that low values on March will mean that we will not have a low record on September. If we have melting conditions on June-August, 2019 will be terrible.
2019 is going to make me eat my nails, with the -165,712 km2 drop that we have today and the heat that we still have on the Arctic.
P.S.: I didn't include 2010 on the graph and it is close to 2012 on March. So there is another year that it was above average on March and end up with a low value on September.