Bob Drach (drach@xxxxxxxx) and I have installed a new version (1.3) of the GDT
netCDF conventions for climate data. The main difference of this version from
the previous one (which was released in January) is in the treatment of the
"long_name" attribute.
One of the aims of the GDT convention is to make it easier to exchange climate
data, especially GCM output, among different institutes. For this reason we
want to provide a way to give each quantity (temperature, pressure, etc.) a
name from a standard list, so that users of data from different sources know
which quantities are comparable. In version 1.2, we proposed that the long_name
should contain a standardised name. However, since the long_name is an existing
attribute in wide use without any standardisation, this proposal could be
inconvenient. Therefore in version 1.3 we have restored our earlier proposal
that the standardised name should be stored in a separate "quantity" attribute.
The quantity attribute is optional in all cases. It is intended for easier data
portability. For convenience and backward compatibility, it is permitted in 1.3
to record the standardised name in the long_name instead of the quantity
attribute, if you don't want to use the long_name for some other purpose.
We would welcome any comments on GDT 1.3. It can be read via
http://www.unidata.ucar.edu/packages/netcdf/conventions.html or directly from
PCMDI as http://www-pcmdi.llnl.gov/drach/GDT_convention.html or the UK Met
Office as http://www.met-office.gov.uk/sec5/CR_div/GDT_convention.html.
Jonathan Gregory (jmgregory@xxxxxxxxxxx)
