"N_Cook" wrote in message news
I've no knowledge on that, this is just my halfpennyworth. If the ice is
over seawater and when it breaks off it forms an ice-berg, then classed
as sea-ice I would have thought.
Penetrating radar has long since defined the boundary of the land.
================

Yes, certainly, the true land boundary of the Antarctica is well known. But
that's not to say that the ice-shelves are not added to the land boundary
for the purpose of the sea ice maps.

Take a look for example at the extent of the Ross ice shelf, as per:

https://en.wikipedia.org/wiki/Ross_Ice_Shelf

Then compare that to the current NSIDC map.

The Ross sea ice looks to have very largely gone, yet there's no suggestion
AFAIK that the Ross ice shelf is actively melting. So it's presumably still
all there, yet not showing up on the NSIDC map.

And the introductory notes to sea ice at eg:

https://nsidc.org/cryosphere/seaice/index.html

make a major distinction between sea ice and land-derived ice, eg:

'Sea ice is simply frozen ocean water. It forms, grows, and melts in the
ocean. In contrast, icebergs, glaciers, ice sheets, and ice shelves all
originate on land. '

'The most basic difference is that sea ice forms from salty ocean water,
whereas icebergs, glaciers, and lake ice form from fresh water or snow. Sea
ice grows, forms, and melts strictly in the ocean. Glaciers are considered
land ice, and icebergs are chunks of ice that break off of glaciers and fall
into the ocean.' (Though this para doesn't specifically mention ice shelves,
AIUI the shelves are made from glacier ice and so aren't sea-ice by
definition.)