Coastline is missing ice shelves, CCAMLR coastline mismatch with ADD-derived coastline

[raymondben]

Does the simplified ADD map in quantarcticR provide a better alternative, perhaps?

[raymondben]

WIP: data-raw/SOmap_data.R now has separate data objects for ice shelves/tongues/etc, and land

[Maschette]

Do I recall you finishing this? I feel like I saw it somewhere.

[raymondben]

No, not finished yet. The data objects are there, but they aren't used in the code yet. It's one of the many things I'll get back to when time allows (but anyone else can take it on in the meantime ...)

[raymondben]

Any chance we can re-define the CCAMLR planning domain data so that they do not include the coastline (and same for any other management layers in SOmap_data that have coastline in them)? If they are just lines over the ocean (e.g. like the IWC domains) things will be easier and faster

[mdsumner]

YESSS - and we could add something to "bake in" a coastline at user request. sf is fast enough for that to be feasible now (though it's still not always reliable)

FWIW, I have quite a few of these kinds of layers defined from scratch, from a few points and lines (MEASO, aceecostats, and some CCAMLR things)

[Maschette]

Simple answer, Possibly? it has always been unclear about where the CCAMLR boundary starts in relation to ice shelves but in theory we could remove their coast line and use our own as long as we are quite explicit about it in the documentation. It would mean you couldn't easily use those layers for analysis like i currently do.

[mdsumner]

Well, it's the convention text that holds the boundary - but they then publish a shapefile with some unknown provenance with a baked in geography. I think it's clear where our responsibility lies, and btw of course there is this long ago thing:

https://github.com/mdsumner/cconvention

    #The waters bounded by a line starting at 62°S 30°E
    # thence due east to 62°S 73°10'E
    # thence due south to 64°S 73°10'E
    # thence due east to 64°S 80°E
    # thence due south to the Antarctic Continent
    # thence westward along the coast of the Antarctic Continent to 30°E longitude
    # thence due north to the starting point.

SW even wrote a recursive descent parser to extract the coordinates from the convention text, but I'll leave that aside for now

I guess I'm volunteering ...

[Maschette]

My main thought would be we should have both in data so people can access if they want it, but our default would be the correct continent file.

[raymondben]

Ice shelves added as of 83df0f4
Have not yet resolved the CCAMLR vs ADD coastline discrepancies

[raymondben]

2f504df uses a simplified (less complexity, smaller object size) of the domains layer, but I think ideally we still want to remove the coastline from any layer except the coastline layer

[mdsumner]

I'll revisit this stuff