[Gdal-dev] Georeferencing netCDF files

[Frank Warmerdam]

Joaquim Luis wrote:
> This doesn't seam logic to me. Why is the georeferencing saved as 
> attributes of the dataset and not
> as global attributes (netCDF parlance)? The only good(?) reason that 
> occurs to me is if different
> datasets can have different reference systems. But does that make sense 
> as well?
> 
> Having the referencing as a global attribute would facilitate the 
> parsing and allow gdalinfo to
> inform us on spatial reference system right from the basename of a 
> multiband file instead of having to fish
> it from its SUBDATASETS.

Joaquim,

In the GDAL data model different datasets can have different coordinate
systems, so it seems natural to associate the coordinate system as
closely as possible to the dataset.

Generally speaking when you see the gdalinfo report of a file that contains
subdatasets and no data itself, the coordinate system, metadata, size and
other attributes should be ignored.  The dataset is just a "stub", existing
to list the actual subdatasets.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org