[CF-metadata] CF standard names for chemical constituents and aerosols (resulting from a GRIB2 proposal)

[Philip J. Cameronsmith1]

On Wed, 17 Sep 2008, Martina Stockhause wrote:

> Dear all,
>
> a group of chemical modellers discussed the standardization of the chemical
> constituents and aerosols in the GRIB2 standard. I assist them. As sources we
> used the parameters used in different intercomparison projects: GEMS
> (http:// gems.ecmwf.int/), CCMVal (http:// www. pa.op.dlr.de/CCMVal/), TFHTAP
> (http:// www. htap.org).
>
> Martin Suttie from ECMWF / GEMS created a two tables approach for GRIB2: one
> for the physical quantity and one for the chemical constituent or aerosol.
> Thus, a physical-chemical  quantity (= CF standard name) is a combination of
> these two parameters. Since, we want to connect GRIB2 standard closely to the
> CF standard names, we used CF standard names in the proposal for WMO and
> propose here the missing CF standard names used in our WMO proposal.
>
> 1. NEW physical parameters for chemical constituent X:
> I read the discussion about 'burden' in the mailing archive. It seems that the
> scientists stick to 'burden' as a common term in the community and have
> started to use it for the creation of CF-like standard names (e.g. in
> CCMVal), because CF couldn't offer an acceptable alternative.
> Neither 'global_integral' nor 'burden' could be agreed on in the CF mailing
> list. Sorry, for starting that discussion again without offering a new
> alternative. Since the CF standard names should be self-explanatory for those
> who create and analyse data, we propose the standard
> name 'total_yearly_average_burden_of_X_in_atmosphere' among others:
>
> column_integrated_mass_density_of_X_in_air  : kg.m-2
> tendency_of_atmosphere_mass_content_of_X_due_to_respiration : kg.m-2.s-1
> tendency_of_mole_concentration_of_X_due_to_chemical_gross_production :
> mol.m-3.s-1
> tendency_of_mole_concentration_of_X_due_to_chemical_gross_destruction :
> mol.m-3.s-1
> total_yearly_average_burden_of_X_in_atmosphere : mol
> total_yearly_averaged_atmospheric_loss_of_X : mol.s-1
> surface_area_density_of_X : m-1
>
>
> 2. NEW chemical constituents:
> We suggest to use the keyword 'lumped' to indicate a certain group of species
> showing similar chemical behaviour (PRISM definition). And to
> introduce 'anthropogenic' and 'biogenic' for the origin of the organic
> compound NMVOC.
>
> atomic_oxygen
> nitrate_radical
> nitrous_acid
> radon
> methyl_peroxy_radical
> methanol
> formic_acid
> aceto_nitrile
> ethanol
> acetic_acid
> butane
> alpha_pinene
> beta_pinene
> limonene
> hcfc141b
> hcfc142b
> nitric_acid_trihydrate_aerosol
> hox_expressed_as_hydrogen : HOx radical (OH+HO2)
> total_peroxy_radical : Total inorganic and organic peroxy radicals (HO2 + RO2)
> passive_ozone
> noy_expressed_as_nitrogen : All nitrogen oxides (NOy) expressed as nitrogen
> clox_expressed_as_chlorine : Total inorganic chlorine except HCl, ClONO2: ClOx
> brox_expressed_as_bromine : Total inorganic bromine except HBr, BrONO2: BrOx
> lumped_alkanes
> lumped_alkenes
> lumped_aromatic_compounds
> lumped_terpenes
> nmvoc_expressed_as_carbon : Non-methane volatile organic compounds expressed
> as carbon
> anthropogenic_nmvoc_expressed_as_carbon : anthropogenic
> nmvoc_expressed_as_carbon
> biogenic_nmvoc_expressed_as_carbon : biogenic nmvoc_expressed_as_carbon
> lumped_oxygenated_hydrocarbons

Hi Martina,

Thank you for working on this list. I know it is a lot of work.

Here are my thoughts:

1)  The question of 'which species are included in the lumped/total 
hydrocarbons' has been discussed at some length for CF in the past, but 
did not reach a conclusion (as I recall).   However, I believe some 
progress was made.  In particular, it has been agreed that a group of 
species can be included in a CF standard name if the group is generally 
comparable between models and observations (eg total_reactive_nitrogen).

The challenge with hydrocarbons is that different models can have 
different cut-offs for the minimum number of carbons in the group (e.g. 
all alkanes with N or more carbons), which make the results 
non-comparable.

In the absence of a comprehensive ontology (which was proposed, but is not 
yet available as far as I know) I think the best solution was suggested
by Martin Schultz in his email to the CF mailing list on 6 Aug 2007, ie 
something along the lines of:

     mixing_ratio_of_alkanes_with_at_least_three_c_atoms_in_air
     mixing_ratio_of_alkanes_with_at_least_four_c_atoms_in_air
     mixing_ratio_of_alkanes_with_at_least_five_c_atoms_in_air
     mixing_ratio_of_alkanes_with_at_least_six_c_atoms_in_air

This is not a perfect solution, as Martin discussed in his email, but I 
think it is a fair compromise between accuracy and usability.

2) What is meant by 'expressed_as_carbon'?  Does it mean 'mass of carbon 
contained in species, ignoring all other chemical elements'?  If so, would 
it be appropriate to put this into the units instead of the standard name?

3) As I alluded to in a separate email, I think there is a challenge to 
defining 'passive ozone', given that there are several varieties of 
passive/artificial tracers I have seen in use.  However, I do agree that 
passive tracers can be valuable output from models, and that it is a good 
idea to clearly separate such artificial tracers from the real specie they 
are related to (in this case ozone).

I do not recall anything in the previous discussion which offers a good 
solution to handling the different passive tracers.  It may be that each 
time one of these passive/artificial species comes up we will need to give 
it a new name and define it clearly.

Best wishes,

      Philip

------------------------------------------------------------------------
Dr Philip Cameron-Smith        Atmospheric, Earth, and Energy Division
pjc at llnl.gov                   Lawrence Livermore National Laboratory
+1 925 4236634                 7000 East Avenue, Livermore, CA94550, USA
------------------------------------------------------------------------