For newbies like myself this might be interesting -- it involves the shape of things to come. My apologies to all the old hands on deck.
The limit of the ice along the Atlantic coast is pretty much defined by bathymetry, which shows how important melting from beneath is. The limit is approximately the 500m submarine contour (edge of the Kara/Laptev shelves). The reason for this boundary is that warm but salty and dense Atlantic waters take the plunge at this point. The 2012 NSIDC source below states that if this boundary is violated, then we'll know a serious sea-change has occurred.
This melting season, that boundary has not been violated by Atlantification (to my interpretation), but it is being pressed (edging ever northwards). The ice is melting up to that boundary more quickly to date, as far as I can tell from the Bremen AMSR2 records, than in any previous year.
Now to the point of all this. The boundary is obvious in the Barents Sea.
I think the boundary is also now becoming evident on the edges of the northern Kara in the Worldview image below (contrast enhanced b/c of clouds). What I think I see here is the compact continuous pack ice running from the left of FJL, diagonally up to the top left of picture off the coast of the Severnaya Zemlya islands. The rest of the northern Kara Sea (all the ice to the right of that compact CAB ice) seems to consist of a loose rubble of floes. So it looks as if the ice melt is continuing apace up to the region of the 500m submarine contour.
I am not claiming this is unusual, but it does seem to be happening earlier than usual even in what seems to be a fairly standard melt year for the past few years. There may also be continuous open water earlier than usual between the Kara and the Laptev Seas.
NSIDC bathymetry source:
http://nsidc.org/arcticseaicenews/tag/bathymetry/