[CF-metadata] Proposed Standard Names

From: Jonathan Gregory <j.m.gregory>
Date: Mon, 19 Sep 2005 23:36:14 +0100

Dear Mike

Thanks for your reply. I think that clears up pretty much everything. All
these are agreed:

 * speed_of_sound_in_sea_water:m s-1
 * height_above_sea_floor:m
 * volume_absorption_coefficient_of_radiation_in_sea_water:m-1
 * volume_attenuation_coefficient_of_radiation_in_sea_water:m-1
 * volume_scattering_function_of_radiation_in_sea_water:m-1 sr-1
 * omnidirectional_photosynthetic_spherical_irradiance_in_sea_water:W m-2
 * omnidirectional_spectral_spherical_irradiance_in_sea_water:W m-3 I
 * moles_of_oxygen_per_unit_mass_in_sea_water:mol kg-1
 * moles_of_nitrate_per_unit_mass_in_sea_water:mol kg-1
 * moles_of_phosphate_per_unit_mass_in_sea_water:mol kg-1
 * moles_of_silicate_per_unit_mass_in_sea_water:mol kg-1
 * moles_of_nitrite_per_unit_mass_in_sea_water:mol kg-1
 * moles_of_nitrate_and_nitrite_per_unit_mass_in_sea_water:mol kg-1
 * bioluminescent_photon_rate_in_sea_water:s-1 m-3
 * fractional_saturation_of_oxygen_in_sea_water:1
 * volume_mixing_ratio_of_oxygen_at_stp_in_sea_water:1

> I would like to propose that absorption, attenuation, AND scattering should
> be integrals over all wavelengths, unless there is a coordinate variable
> specifying range(s) of wavelengths.

Yes, I agree.

> On the topic of scattering, I had argued that in addition to a general
> "scattering" variable, there ought to be a variable called "backscattering".
> Your reply was that backscattering is just a special case of scattering, and
> that a coordinate variable should be used to differentiate the two. Well, I
> agree that the coordinate variable is important, because there certainly are
> many angles around which scattering is measured. However, I would like to
> propose once again that "back" be retained as a special case of "scattering"
> in much the same way that "upwelling" and "downwelling" are used modify
> "radiation", conveying concepts that can be expressed concisely with a
> coordinate variable, but which are conveyed more clearly with a choice of
> name.

Where there are names containing "upwelling" and "downwelling", we do not also
have a "generic" name that could be either, so it's not an exact analogy. Are
you proposing we should have *both* of

 * volume_backwards_scattering_coefficient_of_radiation_in_sea_water:m-1
 * volume_scattering_coefficient_of_radiation_in_sea_water:m-1

I think this would be OK if the first is defined to apply to the sum of all
backward scattering angles, and is not allowed to specify scattering angle as a
coordinate. If that's a common quantity, I agree it would be useful to have a
name for it, and probably we should have a correspondingly defined "forward"
one as well. We have also agreed we need to define

  * scattering_angle:rad

> I had used the term "surface" in some of my original proposed variable
> names... It is probably redundant in the cases I had suggested, since there
> will often be a coordinate variable that indicates depth or altitude = 0.

Regarding

 * partial_pressure_of_carbon_dioxide_in_air:Pa
 * partial_pressure_of_carbon_dioxide_in_sea_water:Pa
 * carbon_dioxide_partial_pressure_difference_between_air_and_sea_water:Pa

If these are surface quantities, I think we should put "surface_" on each of
them. This is the general rule. The point is that there are many coordinates
you could equivalently use to indicate "surface", so it's inelegant to
introduce a single-valued coordinate variable solely for that purpose, and
makes things complicated for the data-reading program, which would have to
check for many possibilities.

> Finally, for sea water flow ... it applies to instruments that have internal
> measurement elements and require a pump to ensure flow through the
> measurement elements. Commonly, the flow rate through these measurement
> elements is also measured, usually as a quality control check.

Is it appropriate to have a standard name for this? It sounds rather specific
to the measurement technique and instrument. Would it often be useful to label
this quantity in such a way that it could be compared in datasets from
different sources (the main use of standard names)? If not, it might be better
to use a long_name for it.

Best wishes

Jonathan
Received on Mon Sep 19 2005 - 16:36:14 BST