Hello Kristian, Jonathan and Jim,
I also agree with Jonathan's observations and I think that some
changes would be required to the conventions to make it clear that the
ancillary_variables attribute is meaningful when data variables are
used in other contexts (namely auxiliary coordinate and, why not,
measure variables). Something like:
* Change section 3.4 Ancillary Data to read from the beginning:
"When one data variable provides metadata about the individual
values of another data variable, auxiliary coordinate variable or
measure variable it may be ..."
* Allow, in appendix A, the ancillary_variables attribute to be
meaningful on auxiliary coordinate variables and measure variables
as well as data variables. We'd need a some new letter codes for
that, though ...
What do you think?
All the best,
David
---- Original message from Jonathan Gregory (06PM 06 May 15)
> Date: Wed, 6 May 2015 18:09:31 +0100
> From: Jonathan Gregory <j.m.gregory at reading.ac.uk>
> To: cf-metadata at cgd.ucar.edu
> Subject: Re: [CF-metadata] Ancillary variables in coordinate variables
> (latitude, longitude, ...)
> User-Agent: Mutt/1.5.21 (2010-09-15)
>
> Dear Kristian
>
> > In these case the feature type is a trajectory, so
> > the latitude and longitude are auxiliary coordinates.
> >
> > You said that for auxiliary coordinates it could be ok, Would this mean
> > that it could be added to the CF conventions?
>
> I don't think any change would be needed to the CF convention. CF allows a
> given netCDF variable to serve as both a data variable in its own right and
> an aux coord var of another data variable. An example of where this might be
> useful is for an ocean model on density levels, in which depth as a function
> of density is of interest as a data variable, but could also be an auxiliary
> coordinate variable for other data variables. Therefore I think it would be
> legal for your aux coord vars to have ancillary vars. Since the lat and lon
> along a trajectory are actually dependent data variables, this feels quite
> reasonable to me for your case. I wonder what others think.
>
> Cheers
>
> Jonathan
>
> > On 5 May 2015 at 18:58, Jonathan Gregory <j.m.gregory at reading.ac.uk> wrote:
> >
> > > Dear Kristian
> > >
> > > Ancillary variables are only for data variables in the current convention.
> > > It wouldn't be illegal to use the attribute on coord variables, but CF
> > > doesn't define what that would mean. (CF always permits any
> > > non-standardised
> > > attributes.)
> > >
> > > I wonder whether these are 1D coord variables (in the Unidata sense) or
> > > aux coord vars (named by the coordinates attribute)? In the latter case, it
> > > could be OK, because other data variables could be named as aux coord vars.
> > >
> > > Best wishes
> > >
> > > Jonathan
> > >
> > >
> > > ----- Forwarded message from Kristian Sebasti??n <ksebastian at socib.es>
> > > -----
> > >
> > > > Date: Thu, 30 Apr 2015 08:43:54 +0200
> > > > From: Kristian Sebasti??n <ksebastian at socib.es>
> > > > To: cf-metadata at cgd.ucar.edu
> > > > Subject: [CF-metadata] Ancillary variables in coordinate variables
> > > (latitude,
> > > > longitude, ...)
> > > >
> > > > Dear CF community,
> > > >
> > > > We have some dataset with quality controls applied to the coordinate
> > > > variables, such as latitude and longitude coordinate. The result are
> > > > quality control variables that we associate as ancillary variables of the
> > > > coordinate variables with the ancillary_variables attribute. For example,
> > > > the LAT coordinate variable has the ancillary variable QC_LAT. The
> > > dataset
> > > >
> > > http://thredds.socib.es/thredds/dodsC/drifter/surface_drifter/drifter_svp052-ime_svp017/L1/2014/dep0001_drifter-svp052_ime-svp017_L1_2014-05-25.nc
> > > >
> > > > The cf-conventions clarifies the use of the ancillary_variables attribute
> > > > for data variables but not for coordinate variables. My question is, Is
> > > the
> > > > ancillary_variables attribute in coordinates variables compliant with the
> > > > cf-conventions?
> > > >
> > > > Best regards,
> > > >
> > > > Kristian
> > > >
> > > > --
> > > >
> > > > Kristian Sebastian Blalid
> > > > SOS Division: Data Center Technical
> > > > Tel: 971439860 - Fax: 971439979
> > > > E-mail: kristian.sebastian at socib.es
> > >
> > >
> > >
> > > > _______________________________________________
> > > > CF-metadata mailing list
> > > > CF-metadata at cgd.ucar.edu
> > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > >
> > >
> > > ----- End forwarded message -----
> > > _______________________________________________
> > > CF-metadata mailing list
> > > CF-metadata at cgd.ucar.edu
> > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > >
> >
> >
> >
> > --
> >
> > Kristian Sebastian Blalid
> > SOS Division: Data Center Technical
> > Tel: 971439860 - Fax: 971439979
> > E-mail: kristian.sebastian at socib.es
>
>
>
> ----- End forwarded message -----
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
David Hassell
National Centre for Atmospheric Science (NCAS)
Department of Meteorology, University of Reading,
Earley Gate, PO Box 243,
Reading RG6 6BB, U.K.
Tel : +44 118 3785613
E-mail: d.c.hassell at reading.ac.uk
Received on Thu May 07 2015 - 06:31:45 BST
