Wow! Great progress above.
Do we have Chris’s 2013 regional mask in the form of a piomas grid cell color markup (first image below, from wip’s #1613)?
While regional seas will inevitably be somewhat arbitrary, other groups using other definitions have data resources eg for ‘Chukchi Sea’ that we would like to integrate with sea ice thickness.
I looked at the Beaufort Gyre to see if that could form an objective basis for defining ‘Beaufort Sea’ but it is completely unworkable. The two deep passages across the Lomonosov Ridge that affect ocean flow structure do not lend themselves to defining regional sea boundaries either.
Probably the best physical considerations for defining natural boundaries are bathymetry and slope, notably the 150m depth contour which defines the continental shelf break-point to the abyss. This generally steers incoming currents and governs tidal vertical eddy and surface wind mixing, and so has a lot to do with peripheral ice bottom melting. The shelf shows up in M Jakobsson’s segmentation analysis of 2003 but the traditional seas not so much.
https://www.researchgate.net/profile/Martin_Jakobsson2/publicationsAs Hyp notes above, a lot of other mechanisms come in to play as thinner and weaker first-year ice comes to predominate, notably ice freeboard dipping below sea level, turning snow layers into slush and then into 'snow-ice', which might then be covered by later snow. Rain and melt events in both the Arctic and Antarctic may become more frequent, with new layers having odd effects on inputs to the piomas algorithm which is not currently configured to interpret them.
N-ICE2015 also documented storm-induced break-up of snow-loaded floes and loss of buoyancy after basal ice melt. The ice cores below show, in polarized light, how complicated the vertical ice structure can become even 100 km internal to the ice edge.
Snow contribution to first-year and second-year Arctic sea ice mass balance north of Svalbard
MA Granskog et al
http://onlinelibrary.wiley.com/doi/10.1002/2016JC012398/full free
Observations of flooding and snow-ice formation in a thinner Arctic sea ice regime during the N- ICE2015 campaign: influence of basal ice melt and storms
Christine Provost et al DOI 10.1002/2016JC012011
NSIDC has a new (for me) display tool for Tschudi's sea ice age algorithm (which is largely based on older ice being lighter than newer on radar). Sea ice age has been shown, unsurprisingly, to be highly correlated with ice thickness. However the maps correlate rather poorly with piomas thickness. It would be good to have both as compatible overlays; the one below is in the widely used epsg4326.
http://nsidc.org/soac/sea-ice-age-year.html#seaiceagesequential