I think this winter is going to get studied like crazy,
for quite a while. It’s a very interesting time.
Jennifer Francis, Washington Post
The extraordinary temperature anomalies in the Arctic since the start of the year haven't gone unnoticed in quite a few media outlets, and I apologize for not having joined the fray of actuality. On the other hand, context trumps actuality, as we need to compare to previous years and get a feel for what this prelude to the melting season may mean. In that sense, I'm early with this year's winter analysis (compared to last year).
Let's start studying like crazy, shall we?
It's a lot of text and images, so if you're feeling a bit tl;dr-ish today,
skip to the conclusion at the bottom of the page.
Surface air temperature
Here are the monthly temperature graphs for November-February in the Arctic Circle, from 2005/2006 to the past winter, based on the NCEP reanalysis dataset:
Last November saw the highest average monthly temperature in the 2005-2015 record, followed by a lower December, relatively speaking. Things then get a little bit crazy after the turn of the year, with the January 2014 record getting broken by almost 3 °C! February isn't far behind either, almost 1.5 °C higher than the already 'warm' February of 2014. This is unprecedented.
To see where temps were least low, I've created average temperature maps using the Daily mean composites page from NOAA's Earth Science Research Laboratory website, comparing the 2015/2016 freezing season to those preceding the years with the lowest minimums on record (click for a larger version):