[Imports] Antarctica coastline/shelf-ice import

[Jochen Topf]

Hi!

The Antarctica coastline currently existing in OSM was imported years ago from
rather bad and outdated data. And the huge shelf-ice areas are mostly missing.
There is much better data available that can be used to replace the old data.
Christoph Hormann and I are planning an import that will improve this situation
greatly.

Why an import?

The existing data is the result of an import and was not changed much after
that import. It is much easier to replace it wholesale with very good data
derived from public datasources than to try to fix it piece by piece. There are
a few areas where the current data is the result of more detailed work and in
those areas the existing data will be preserved.

What are we importing?

Antarctica is very different from most other places in the world, because most
of it is covered in ice, ice which is in places over 2000m thick. Around the
continent of Antarctica itself, this ice forms huge permanent ice-shelves (not
to be confused with seasonal sea ice). For most purposes these ice shelves
"behave" like solid land. There are permanent settlements, airfields, etc. on
them. Currently those ice-shelves are not mapped at all which means that some
research stations look like they are in the water.

We are importing
a) OSM ways for the "rock coastline" where ocean water meets solid land.
b) OSM ways for the "calving line" where ocean water meets the permanent ice-shelves.
c) OSM ways for the "grounding line" where the floating ice-shelves meet the ice
   sitting on the continental rock.
d) OSM multipolygon relations for the ice-shelves.
All these ways and relations will get tags. See
http://wiki.openstreetmap.org/wiki/Antarctica/Tagging for details.

One important difference from the current tagging is that we propose to tag
the "calving line" instead of the "grounding line" as natural=coastline. This
makes a lot of sense, because for most purposes the "water" side of the new
coastline allows typical water use (such as moving about on a ship) while the
"land" side of the new coastline allows typical land use (such as building
a settlement). This change in tagging will affect all OSM map of Antarctica
right away. After the import we have the data available to add specific
rendering for shelf-ice allowing a much improved OSM map.

We are also proposing to remove the huge natural=glacier multipolygon that
covers large parts of Antarctica. It is quite unusual to tag the "default" land
cover this way (there is no "desert" multipolygon covering half of North Africa
or "forest" multipolygon covering Siberia). The problem is that every time
somebody adds some land cover information (for instance natural=rock for some
places where there is no ice cap) it has to be added as inner ring to this huge
multipolygon. Sooner or later this multipolygon will be too big to handle properly.
And, for some reason, the currently existing multipolygon does not cover the islands.
Also the way this is currently renders is rather confusing, because this "glacier"
is only rendered from zoom level 6 on, the whole of Antarctica suddenly switches
from the usual "land" color to blue when zooming in.

We have added the proposed import to the "Import/Catalogue" and we have
reviewed the "Import/Guidelines" and "Import/Past_Problems" wiki pages.

What about licensing?

The data is derived from NSIDC (The American "National Snow and Ice Data Center")
data, it is in the Public Domain.

How are we planning to do the import?

Christoph has already converted the data to the OSM format adding all necessary
tags etc. The import will be done through JOSM in one or two sessions and
should be over in a few hours. While Antarctica is huge there isn't all that
much data.  And because not much else is mapped in Antarctica it is rather easy
to go through the few places where old and new data conflict and fix them
manually.

Where do I get more information?

We have written down all the details on this wiki page
http://wiki.openstreetmap.org/wiki/Antarctica/Import_2013
There you can also see more detailed explanations about the coastline, a
comparison between old and new data, download GeoJSON and OSM files with the
new data and read the details on how we are planning to tag the data and a
step-by-step procedure for the import.

Jochen
-- 
Jochen Topf  jochen at remote.org  http://www.remote.org/jochen/  +49-721-388298