Re: NSDL metadata for THREDDS datasets

Ben Domenico writes:
# We are aware of the detailed standards of the FGDC, GCMD, those
# evolving at ISO as well as the ESML definition at UAH. [...]
# But, as a place to get started and gain some experience with the
# metadata generation tools themselves, we're looking for something
# simpler

The danger in choosing something simpler "to get started" is that once
chosen it is very difficult to get rid of.

# http://www.smete.org/nsdl/workgroups/standards/current_element_set.html

This appears only to address educational documents or modules, not
datasets.  Perhaps that is all you need, but you also wrote:

# In particular, we plan to start with the types of data found on many
# THREDDS servers:
#   -- gridded data output from weather forecast models,
#   -- GOES satellite imagery, NEXRAD level 3 images, and
#   -- point source weather observations and balloon soundings from weather 
#      stations
#   -- climate data

I recommend you consider whether a subset of FGDC CSDGM or ISO 19115
would meet your needs for dataset description.  Only some of the fields
therein are required.  See <http://www.fgdc.gov/metadata/contstan.html>.

Note also that specifications like the OGC Web Map Service have explicit
support for linking to FGDC or ISO metadata, and that the Office of
Management and Budget's e-government "Geospatial One-Stop" initiative
will be using FGDC metadata.

Regards,
Jeff DLB


      Dr Jeff de La Beaujardiere * Computer Scientist
        NASA Goddard Space Flight Center * Code 933
       address@hidden * +1 301 286 1569
                http://iniki.gsfc.nasa.gov/



--MAA93115.1008956153/iniki.gsfc.nasa.gov--