Hi Alejandro
> This is a satellite track. Model data will be sampled along this
> satellite track, and the data will be stored as a 'curtain' of vertical
> profiles, with each point defined by (lon,lat,time).
I suspected as much :-)
> This coordinate is defined in the MIP tables for CMIP5
> but the standard name is missing (the MIP name is 'location'). My
> understanding is that all CMIP5 variable and coordinate names have to be
> mapped to CF standard names, but I'm not 100% sure about this. I will
> find out whether a standard name is always required.
Well, currently, I think "the official position" might be that standard names
are only for physical variables, but this is a rather interesting and important
special case ...
... and in the absence of a convention extension which covers this
properly, the following is my gut feeling:
In terms of information content it's just a special case of a timeseries of
station data, but where the word "station" is inappropriate, what we've
actually got is the subsatellite_location
Example 5.4
http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.4/ch05s04.html
suggests that one doens't need standard names for the variables
for the station list, and my gut feeling is that we probably don't
from a CF point of view for this example either.
It become a CMIP5 issue. I suspect for this instance we just need
to agree the CMIP5 long name which is more useful than
"location", and I think it should probably be something like
subsatellite_location (and we probably need which satellite
in the metadata too).
But none of the above is authoratative :-)
(Which MIP table are you looking at? Can you give me the
context, perhaps in offline email, this is of some interest to me for
a number of reasons.)
Bryan
>
> Cheers,
>
> Alejandro
>
> On Wed, 2009-09-16 at 09:49 +0100, Bryan Lawrence wrote:
> > Hi Alejandro
> >
> > > This is going to be used as the horizontal
> > > dimension for a stream of atmospheric vertical profiles (e.g.
> > > temperature(position,height)). It will be an index running from 1 to N
> > > (number of profiles). Then, the variables [latitude(position),longitude
> > > (position),time(position)] will give the geographical position and time
> > > of the profiles. Is there any CF standard name suitable for this?
> >
> > this is really a special case of the observational data issue.
> >
> > we have this with profiles from flights, ships etc, intersections between observations on the ground and swaths overhead etc ...
> >
> > So, in this case, is there an intrinsic meaning (as Jonathan is implicitly asking)? What
> > does this position actually mean?
> >
> > Bryan
> >
> >
> > On Wednesday 16 September 2009 08:58:54 Jonathan Gregory wrote:
> > > Dear Alejandro
> > >
> > > Ah, I see. No, I don't think there is a standard name for that. If the index
> > > had some meaning e.g. if it was a standardised station number, a standard name
> > > could indicate that. If it is purely an index with no intrinsic meaning, I am
> > > not sure that it would add any information to give it a standard name. What do
> > > you think? In a sense, I would say that the variable exists only because it is
> > > not possible to construct the netCDF file without it. All you're really
> > > interested in is its dimension.
> > >
> > > Cheers
> > >
> > > Jonathan
> > > _______________________________________________
> > > CF-metadata mailing list
> > > CF-metadata at cgd.ucar.edu
> > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > >
> >
> >
> >
--
Bryan Lawrence
Director of Environmental Archival and Associated Research
(NCAS/British Atmospheric Data Centre and NCEO/NERC NEODC)
STFC, Rutherford Appleton Laboratory
Phone +44 1235 445012; Fax ... 5848;
Web: home.badc.rl.ac.uk/lawrence
Received on Wed Sep 16 2009 - 03:50:31 BST
