Thanks, Alison. Wave_frequency would be better, although it
implies, to me, the frequency of wave peaks where we really
could use a term for a coordinate parameter of anything that
can be reported in the frequency domain. So, I'm not sure why
'frequency' itself isn't sufficient in this case, other than a
possible conflict with sound frequency ... although its a stretch
to see how these could be mistaken for one another. A specific,
non-coordinate parameter like brunt_vaisala_frequency_in_air
should of course keep its more definitive standard name.
> Don't forget we can have internal waves (waves propagated on subsurface
> density boundaries like thermoclines) as well as surface waves in water
> bodies.
Yes, I guess that's my point: why would we need a separate
frequency coordinate parameter for frequency of internal
waves? The meaning of the coordinate parameter is the same
in both cases, and the standard name of the measured parameter
explains itself.
We don't specify the variable being measured when we use depth
as a coordinate parameter in CF, so I'm not sure why this case
should be different.
> My working hypothesis is that optimum combination of detail and
> generality is medium + property
Mine is that it's a bad idea to provide information that might be
misleading or conflicting in a NetCDF field when that information
is contained in another field, the standard_name. In this case, the
parameter's standard name, sea_surface_wave_variance_spectral_density,
gives information about the quantity being reported, and the name of
the coordinate parameter should contain exactly enough information
to make the measured parameter definitively understood. If there is
another standard name where this would not be the case, I haven't
come across it. That's the strength of NetCDF, and we may as well
take advantage of its strong points, since we put up with its weak
points as well.
Wave_frequency seems like a better alternative and a good compromise,
to me.
Cheers - Nan
> >>> "Pamment, JA (Alison)" <J.A.Pamment at rl.ac.uk> 03/31/08 11:09 AM >>>
> >
> > One of our parameters is sea_surface_wave_variance_spectral_density,
> > with coordinate variables time and frequency.
> >
> > Sea_surface_wave_frequency is an accepted standard name, but I
> > would like to know why 'frequency' is not preferable, being more
> > generic. As an analogy, we can use 'depth' as the coordinate variable
> > for temperatures, and are not so specific as to require
> > sea_water_temperature_depth. Is specifying sea_surface_wave
> > really necessary, and might it not be misleading if the parameter
> > is not actually "plain" sea_surface_waves?
> I'm not an expert in this area myself, so I'm not quite sure what you
> mean by "plain" sea_surface_waves. However, to respond to your general
> point, I agree that sea_surface_wave_frequency is very specific.
> The original proposal (see
> http://www.cgd.ucar.edu/pipermail/cf-metadata/2006/000958.html) was for
> wave_frequency and it was submitted along with quite a list of other sea
> surface wave names. I suspect that it was changed for consistency with
> the other names. I think that we need to at least use the term
> wave_frequency rather than simply 'frequency' because it is not the
> only kind of frequency in standard names. We have the names
> sound_frequency, which could in fact be described by wave_frequency, and
> brunt_vaisala_frequency_in_air, which is a parameter that relates to the
> static stability of air and is useful for determining whether gravity
> waves will be generated, for example, by flow over orography. I think
> that we could use wave_frequency quite adequately for fluid dynamical
> waves, sound waves and electromagnetic waves independent of the medium
> in which they occur. Would you be happy with wave_frequency?
>
> Best wishes,
> Alison
Received on Mon Mar 31 2008 - 10:08:51 BST
