[CF-metadata] CF-metadata Digest, Vol 67, Issue 6

From: Heinke Hoeck <heinke.hoeck>
Date: Wed, 22 Oct 2008 09:08:42 +0200

Dear Martin,

I like to give an example with two constituents:

dimensions:
lat=72;
lon=124;
maxlen=30;
variables:
char constituent_ozone(maxlen);
          constituent_ozone:standard_name='IUPAC_system';
char constituent_NO2(maxlen);
          constituent_NO2:standard_name='IUPAC_system';
float quantity_ozone(time,lat,lon,height) ;
          quantity_ozone:standard_name =
"mass_fraction_of_constituent_in_air" ;
          quantity_ozone:units = "1" ;
          quantity_ozone:coordinates = "constituent_ozone" ;
float quantity_NO2(time,lat,lon,height) ;
          quantity_NO2:standard_name =
"mass_fraction_of_constituent_in_air" ;
          quantity_NO2:units = "1" ;
          quantity_NO2:coordinates = "constituent_NO2" ;
data:
   constituent_ozone='ozone';
   constituent_NO2='NO2';

Best regards
Heinke


>
> Dear Heinke,
>
> if constituent='ozone' is in the data, then how do you deal with a
> file that has several constituents in it? This is the norm rather than
> the exception in chemistry modelling. If you suggest
> constitutent = '[ozone, NO2, CO, ...]'; (or something like this
> in correct syntax)
> then I believe it will be impractical to construct this constituent
> vector. If at all, then I think the constituent name must come along
> with the variable attributes.
>
>
> Best regards,
>
> Martin
>
>
>
>
>
Received on Wed Oct 22 2008 - 01:08:42 BST