Dear John
> Heres another person trying to do staggered grids in CF.
> I forgot what we decided was the right way to do it.
The coordinates have been defined in the right way. CF does not provide a way
to record a relation between different grids. This was discussed in January in
the thread "WRF staggered grids and vertical coordinates". From a couple of my
emails in that thread:
> The point is that this relation (the staggering) can be deduced by
> inspecting the coordinates. Given that this is the case, I don't see the
> need to introduce any new convention to indicate it explicitly, because this
> would be redundant. ... Maybe it would be helpful to an application, but I'm
> not sure it's really necessary, is it? As you said earlier, the methods used
> are quite straightforward: you interpolate one field onto the grid of
> another one, or maybe you interpolate two fields onto a common grid which
> uses x from one and y from the other. To do this, no special machinery is
> required. All the program needs to know is which grids are to be used as
> target for the interpolation. I would argue that's not metadata, but a
> convention for how the data should be used - it's not a property of the
> dataset but of the program.
What do you think?
Looking at the file, I noticed a few other things which might be helpful to
point out:
* units = "nondimensional".
This is not a legal udunit. It should either be omitted or coded as units="1".
* calendar = "365.25 days in every year"
This isn't a standard CF calendar. What is the intention here? I think a year
has got to have a whole number of days in it (possibly a varying number if
there are leap years). Perhaps it means that every fourth year is a leap year
i.e. the julian calendar.
positive = "downward flux, heating"
This isn't a legal value for positive, which should be up or down. It is being
used to indicate the sense of the sensible heat flux. In CF, this is done by
choosing an appropriate standard name. For fluxes, the standard names specify
the sign convention, and we can provide alternative names for the different
choices.
The CF checker
http://titania.badc.rl.ac.uk/cgi-bin/cf-checker.pl (linked from
the CF home page) is useful to check compliance.
Standard names exist for quite a number of quantities in this file but have not
always been used. For those that don't exist, we could define standard names.
I think that would help with the making the file more self-explanatory as it
may not be clear what all the quantities are from the long names.
Best wishes
Jonathan
Received on Tue Oct 12 2004 - 12:11:39 BST
