thinking Beaufort Finger' was this feature?
In that case (which overlooks NBZ's mention of Fram), you are quite right in that that piece of MYI broke off the thick CAA ice and swung around in the half-Gyre that we have had in recent years.
Once upon a time, ice circled (and thickened) in the Beaufort Gyre for years and years, leading to a pronounced dome of lower salinity (see Proshutinsky papers on Ekman pumping that Fish has been citing:
http://www.whoi.edu/page.do?pid=66578 for an update).
Looking once again at that
all-time great animation of Tschudi's sea ice age data by the climate.gov team over the last 26 years, one wonders how the textbook Beaufort Gyre story retains any traction in the new Arctic.
From my perspective, the current drama unfolding on the Chukchi/Beaufort is mostly just the windy specifics of the expected demise of the weak FYI ice there, whereas more noteworthy export of SYI has been taking place across the Svalbard-FJL-SZ maginot line.
It's not easy to translate wind forecasts or reanalysis products (ie pressure gradients) into quantitative ice movement, the problem being it depends on directional surface friction (eg pressure ridges and ice edges). This roughness determines the effectiveness of coupling (momentum transfer) and we know next to nothing about what's currently down there (satellites mostly view radially, flattening) nor new effects from more pervasive younger ice.
So other than waves induced by wind over long fetches open water directed straight on to the ice edge, well-covered by JimH, we might just as well skip the persistent weather pattern -- even though it is causative -- and just directly track the ice going into the Barents.
The idea here is not so much that the bathymetric break to the north of Svalbard-FJL-SZ will end up ice-free again as it is that so much ice has been pushed out to the Barents this year.
The animation below shows the last three weeks. Here, as in a June post, there's been a
cloud removal process applied to affected passive radar sea ice concentration scenes (which active radar like S-1AB does not need).
GreatD has started a whole forum on these enhancements. It's worth noting the commonsensical choices are already provided as menu items at ImageJ and Gimp. The one below (underneath as-is) isn't striving so much to recovered masked concentration values as it is seeking persistent feature visualization and their motion. The upper ten percent of ice concentrations are not resolvable visually so the whites have been replaced by a tan color.
The second product shows where it makes a difference to let a darker pixel from the previous day over-ride the current day's. Posters positing that clouds darken could perhaps make their case with Saldo's S-1AB 3-day mosaics (which are greatly under-utilized on these forums).