wonder how much elevated SST would affect the forthcoming freezing season
Good question. While we wait for jd to crunch the numbers, it is worth asking whether we should be monitoring -- and comparing year-on-year -- the heat burden in near-surface waters that accrues during melting season. The temperature of open surface water is observable by satellite and the depth of the mixing layer is known though we don't have the temperature profile without operational buoys.
The overall heat budget of the Arctic Ocean is greatly influenced by incoming Atlantic Waters; though these are initially at ~300m depth, they don't exit with all the heat they brought in. Over the short term, this lost heat doesn't much influence temperatures of surface waters.
Other complications come in from sunlight penetrating thinner ice quite effectively, warming the water below. And that sunlight can fuel algal growth on the underside of the ice which then adsorb the energy right there rather than it going deeper in the water column.
In any event, Mercator Ocean only offers 2017 and 2018 for today's date (plus various water depths, not shown). The animation below compares them by arithmetic variations (such as subtraction of pixel RGB values) directly on the graphics. That leads to new colors though and so 2D color legends.
Water temperatures now should have considerable predictive power as to which peripheral ice sectors will melt out over the next month (if only we knew how the floes would move relative to warmer water temperatures). At the end of melt season, these surface temperature pockets suggest lagging areas of freeze-up (if only we knew what air temperatures would be associated with them).
Last winter, elevated Bering Sea temperatures (and lack of ice) were unprecedented. The Chukchi didn't fully freeze until early January. How much of that should be attributed to local SST vs imported SST vs mixing layers vs currents/eddies vs air temperatures?
Compare current to 2007
How? It's in a distorted cylinder projection, doesn't go above 80ºN and is unreadable/unreliable there, plus it would have to be cut down to comparable ice covers which isn't practical without having the data in polar stereographic projection. Are the grid cells really of adequate resolution for the Arctic seas and islands -- I can barely make out Greenland.
Overall, there are a great many climate statistics like this, really focused on the equator and mid-latitudes, with the poles just bones thrown to the dog. We are not so interested in oblique observation of the Arctic from satellites in equatorial orbits, as is done for example with lightning. Data quality deteriorates with latitude -- even the 80º cutoff is likely a stretch.