I notice that while some GML objects are used, rather key objects
like netcdfGridCoverage are not derived from the corresponding GML
objects (e.g. gml:GridCoverage ) and hence would not be recognized
as such by GML processing software. Is this intentional?
NcML-GML is an extension of ncML data model; thus, netcdfGridCoverage
extends the netcdfType element defined in ncML.
More generally, the ncML-GML aims at bringing together netCDF and
geographic coverage data models, making them complementary. An
important feature of ncML-GML document is that it contains a valid
ncML core document, from which you can easily generate a netCDF file.
Another important ncML-GML requirement was to keep the extension as
simple as possible, introducing only the geographic coverage metadata
supported by our community.
Naturally, this is a progressive process. As we discussed in a
previous email, the opportunity to develop a future ncML-GML version
which is a GML application is already on the table.
Indeed, ncML-GML version 0.7 is more complete and GML-oriented than
version 0.3. Anyway, the starting point and the main motivation for
ncML-GML are, respectively: netCDF + conventions dataset and to
improve the description of netCDF dataset semantics.
GALEON is an important opportunity to test the interoperability level
required and supported by our community.
