[CF-metadata] CF standard names for chemical constituents and aerosols

Pamment, JA (Alison) alison.pamment at stfc.ac.uk
Fri Apr 24 05:47:50 MDT 2009

Dear Martina and all those who have contributed to the lively discussion
in this thread,

Thanks all for your discussion of these names and thanks to Martina for
her summary of the current state of agreements reached.  Thanks also for
your patience while I have been working my way through this quite
complicated discussion!

1. Regarding the physical quantities in Martina's original proposal, I
think many of the outstanding issues have been resolved.  The following
is a summary of the situation:

> column_integrated_mass_density_of_X_in_air  : kg.m-2
It has been agreed that the existing syntax of
atmosphere_mass_content_of_X (kg m-2) will be used for these names.
This is also consistent with some newly accepted names in the
ECHAM5-Hammoz discussion.

> tendency_of_atmosphere_mass_content_of_X_due_to_respiration :
Early in the discussion Jonathan asked whether 'respiration' should
specify heterotrophic, plant or soil as in existing respiration names,
or whether your names represent the total.  I'm not sure that this point
was answered - Martina, please can you confirm which is the correct

> tendency_of_mole_concentration_of_X_due_to_chemical_gross_production :
This has been agreed with the addition of 'in_air' as follows:

> tendency_of_mole_concentration_of_X_due_to_chemical_gross_destruction
: mol.m-3.s-1
This has been agreed with the addition of 'in_air' as follows:

> total_yearly_average_burden_of_X_in_atmosphere : mol
It was agreed that the existing moles_of_X_in_atmosphere (mol) syntax,
used in conjunction with a cell_methods attribute on the time axis,
would describe this quantity.  In the "medium 'atmosphere'" thread it
has been agreed to rearrange the order of the terms in the existing
names, so the correct syntax will in fact be atmosphere_moles_of_X.

> total_yearly_averaged_atmospheric_loss_of_X : mol.s-1
It was agreed that the existing syntax of
tendency_of_moles_of_X_in_atmosphere (mol s-1), used in conjunction with
a cell_methods attribute on the time axis, would describe this quantity.
In the "medium 'atmosphere'" thread it has been agreed to rearrange the
order of the terms in the existing names, so the correct syntax will in
fact be tendency_of_atmosphere_moles_of_X.

> surface_area_density_of_X : m-1
It was agreed to modify these names to read density_of_surface_area_of_X
(m-1).  The definition of the physical quantity is 'the surface area of
aerosol per unit volume of atmosphere'.

In addition to the physical quantities listed above, I note that the
original list of proposed names (attached for ease of reference) also
contains many names with patterns already in use in the standard name
table, such as mass_fraction_of_X_in_air,
tendency_of_atmosphere_mass_content_of_X_due_to_emission, and so on.
The majority of these are well established and require no further
discussion but there are two sets of names in the list which are
affected by discussions in other threads.  The quantities under
discussion are X_mole_flux (in the 'Modification of some existing CF
standard names for chemical constituents'/'CF: surface flux versus
emissions' threads) and atmosphere_optical_thickness (in the
ECHAM5-hammoz thread).  I think we need to resolve the separate threads
before reaching a final decision on those particular names.

Martina's list includes names of tendencies due to the processes
net_production and net_production_and_emission.  The latter is used in
two existing standard names while the former is not currently used in
the table, but I see no problem in including it as a separate process.
We have now agreed in this thread to add the processes
chemical_gross_production and chemical_gross_destruction and this
prompted me to look back at the (lengthy) discussion of chemical names
we had in 2007 when Christiane proposed net_production for use with the
HTAP names.  Originally, Christiane proposed to use
chemical_net_production but the word 'chemical' has been dropped from
the two names already in the table.  I couldn't find any discussion of
dropping 'chemical' and I now think this may have happened accidentally
when I added the names to the table.  I suggest, therefore, that aliases
should be created for the two existing names to change
net_production_and_emission to chemical_net_production_and_emission and
that the word 'chemical' should be added to Martina's 'net_production'
proposals to make them consistent with the gross production/destruction
names.  I'd welcome any comments on this, particularly if anyone can
recall any agreement to drop 'chemical' from the names.

> 2. NEW chemical constituents:

> atomic_oxygen
> nitrate_radical
> nitrous_acid
> radon
> methyl_peroxy_radical
> methanol
> ethanol
> butane

These species names have not attracted any comment. They are all
individual species, clearly defined, and IUPAC names have been used
where appropriate so I think these are all agreed.

We already have 5 names in the standard name table that refer simply to
'oxygen', presumably in its molecular form.  As we will now be
introducing 'atomic_oxygen' should we create aliases for these older
names by replacing 'oxygen' with 'molecular_oxygen'?  This would then
also be consistent with existing and proposed 'molecular_hydrogen'

> nitric_acid_trihydrate_aerosol
In the discussions under the ECHAM5-Hammoz and 'Modification of some
existing CF standard names for chemical constituents' threads it has
been agreed that all aerosol names should specify either 'dry' or
'ambient' aerosol.  Martina, please could you specify which is correct
for this species? (The definitions of 'dry' and 'ambient' aerosol are
given below for reference).

Dry aerosol: '"Aerosol" means the suspended liquid or solid particles in
air (except cloud droplets). Aerosol takes up ambient water (a process
known as hygroscopic growth) depending on the relative humidity and the
composition of the aerosol. "Dry aerosol" means aerosol without water.'

Ambient aerosol: '"Aerosol" means the suspended liquid or solid
particles in air (except cloud droplets). "Ambient aerosol" is aerosol
that has taken up ambient water through hygroscopic growth. The extent
of hygroscopic growth depends on the relative humidity and the
composition of the aerosol.'

This point about the aerosol names also affects some of the existing
chemical species that appear in Martina's proposals.  In the
'Modification of some existing CF standard names for chemical
constituents' thread it was agreed that all older 'aerosol' standard
names (which existed before the dry/ambient aerosol distinction was
introduced) should be made aliases of 'ambient_aerosol' names.  Thus
mass_fraction names for pm1_aerosol, pm10_aerosol and  pm2p5_aerosol
will be changed to refer to pm1_ambient_aerosol, etc. The names in
Martina's proposals should also specify ambient or dry. Is it fair to
say that these size distribution names will always be 'ambient'?

Another small point regarding an existing species name:  please note
that secundary_particulate_organic_matter should read
secondary_particulate_organic_matter, i.e, there is a typo in

> formic_acid
> aceto_nitrile
> acetic_acid
> alpha_pinene
> beta_pinene
> limonene
> hcfc141b
> hcfc142b

None of these are IUPAC names.  Philip has kindly looked up their IUPAC
equivalents, which are long and complex, and has proposed that CF should
take a pragmatic approach to using chemical names.  Others have
expressed strong agreement with this position.  Specifically, it has
been agreed that IUPAC names should be used where practicable, but it is
recognised that non-systematic IUPAC approved names, or on occasion
other common names, will be preferable where these are widely used,
shorter and unambiguous.  Where non-IUPAC chemical names are used in
standard names the definition of the name should include the IUPAC
equivalent.  I think that we have already tried to take this approach
with chemical names in CF and so it is good if we now have formal
agreement that this is the way to go.  Following the earlier discussion
in this thread, these eight chemical names are agreed.

> noy_expressed_as_nitrogen : All nitrogen oxides (NOy) expressed as
The group of species here referred to as noy is already in the standard
name table using the term 'reactive_nitrogen'.  For example, we have the
name mole_fraction_of_total_reactive_nitrogen_in_air defined as 'Mole
fraction is used in the construction mole_fraction_of_X_in_Y, where X is
a material constituent of Y. "Reactive nitrogen", sometimes referred to
as Noy, describes a family of chemical species. The family usually
includes atomic nitrogen (N), nitrogen monoxide (NO), nitrogen dioxide
(NO2), dinitrogen pentoxide (N2O5), nitric acid (HNO3), peroxynitric
acid (HNO4), bromine nitrate (BrONO2) and chlorine nitrate (ClONO2).'
There is only one reactive_nitrogen name in the table.  We could choose
either to create an alias for that one name and change it to use 'noy',
or we could change Martina's proposals to use the term
'reactive_nitrogen'.  I have no strong preference for either term and
would be happy to go with the majority view.

> hox_expressed_as_hydrogen : HOx radical (OH+HO2)
> total_peroxy_radical : Total inorganic and organic peroxy radicals
(HO2 + RO2)
> clox_expressed_as_chlorine : Total inorganic chlorine except HCl,
ClONO2: ClOx
> brox_expressed_as_bromine : Total inorganic bromine except HBr,
BrONO2: BrOx
> nmvoc_expressed_as_carbon : Non-methane volatile organic compounds
as carbon
> anthropogenic_nmvoc_expressed_as_carbon : anthropogenic
> biogenic_nmvoc_expressed_as_carbon : biogenic

'Expressed_as' has been discussed in this thread and also in the
ECHAM5-Hammoz and "position of 'expressed as' in chemical constituents'
threads. There is general agreement that 'expressed_as' should appear in
the standard name directly following the species to which it applies.
Philip sought clarification of the definition of 'expressed_as'.  It was
agreed that 'expressed_as_X' means the mass of X (where X is a chemical
element) contained in a species or group of species, ignoring all other
chemical elements.  This is the sense in which it has been used in
existing names.

Martina provided the following definitions during the discussion:
'anthropogenic' refers to chemical constituents influenced, caused, or
created by human activity;
'biogenic' refers to chemical constituents generated by natural
'oxygenated' means 'containing oxygen'.

> lumped_alkanes
> lumped_alkenes
> lumped_aromatic_compounds
> lumped_terpenes
> lumped_oxygenated_hydrocarbons

All the above 'expressed as' and 'lumped' names refer to groups of
species that are in some sense treated collectively in atmospheric
chemistry models.  There has been quite a lengthy discussion of these
names from which the following main points arise:
a) Martin explained that the extent to which the groups of species are
defined is variable. For example:
'HOx radical' is well defined as a combination of just two species
'RO2' is a combination of a larger number of species (all organic peroxy
'nmvoc' (non-methane volatile organic compounds) is not very clearly
defined (e.g., sometimes it will include alkanes, sometimes not, and
similarly for alkenes, etc.);
'lumped' alkanes may mean alkanes with more than 3 carbon atoms, or more
than 4 carbons, or more than 5, etc.
b) For the less well defined groups the 'lumping' will vary from model
to model.
c) There was discussion of the use of the terms 'lumped' and 'total' in
standard names.  There are only two names currently in the table that
use the term 'total' - mole_fraction_of_total_inorganic_bromine_in_air
and mole_fraction_of_total_reactive_nitrogen_in_air.  There are
currently no names using 'lumped'.  Jonathan commented that the use of
'total' is to emphasise the total of mole fractions of all species that
"qualify" as inorganic bromine, reactive nitrogen, etc.  Martin said
that the difference between 'lumped' and 'total' is the extent to which
the group of species is well defined. 'Lumped' groups are deliberately
vague - they may be defined within a community but they are not well
enough defined to be used across different communities.  Martin
recommended that the definitions of 'lumped species' names should state
that the group is vague.  Philip commented that, notwithstanding the
difference in concept between 'total' and 'lumped', standard names
should employ only one term - 'total' - because that would be less
ambiguous for non-experts, or alternatively we should avoid entirely the
use of any 'grouping' term.  Jonathan agreed with the latter suggestion.
If we were to adopt this suggestion, aliases would need to be created
for the two existing 'total' names.  I.e. they would become
mole_fraction_of_inorganic_bromine_in_air and
mole_fraction_of_reactive_nitrogen|noy_in_air.  Additionally, Martina's
'lumped' species above would become:
and total_peroxy_radical would become simply peroxy_radical.
I would welcome further comments regarding this suggestion.  In
particular, would the names without 'lumped' be acceptable in the
context of GRIB2? One last point: I think that, no matter which
terminology we choose to adopt, the definitions of all group species
standard names should contain a statement of how precisely or vaguely
the group is known.


At present, we have not reached a conclusion on this.  As Philip
explained, a particular species may be passive in some respects, but
reactive in others.  He gave the example of stratospheric ozone in the
troposphere.  It is generally agreed that it is useful to have standard
names for tracers to separate them from fully interactive species.
However, it is unlikely that all the information regarding the
circumstances in which a particular species is passive and reactive
within a particular model could be given in the standard name.  We have
discussed passive_ozone before and reached a similar point before the
proposal was withdrawn.  It seems that we will need to make use of other
attributes besides standard_name if we are to fully describe passive
tracers.  It would be helpful if we could gather a number of use cases
in order to assess the information that needs to be described and
address how to do that within CF metadata.

I have a couple of questions regarding the relationships between some
new species in Martina's list and some that already exist in the table:


We already have names in the table that refer to
alpha_hexachlorocyclohexane and Martina's proposals include more names
for that species, plus names that refer simply to
'hexachlorocyclohexane' (i.e., no isomer specified). Please can you
confirm whether that was intentional (i.e. I want to check that both
sets of names are needed and it isn't just a by-product of generating
the new names with a script).


Again I just wanted to check that these are different species, i.e, are
there other forms of ammonium in the atmosphere besides dry_aerosol?
Does ammonium include ammonium_dry_aerosol?

> 3. NEW CF standard names:
> The individual CF standard names, resulting from the combination of
> quantity and chemical constituent or aerosol, are quite many (see
> file). 
> I apologize for this long list. My idea to move the name of the
> constituent or aerosol from the standard name into an attribute was
> by the scientists. The disadvantage of a CF standard name without the 
> constituent was larger than the advantage of a smaller number of CF
> names to be proposed.
> Since I had to use a script for the generation of these individual CF
> there might be some names in the list, which doesn't make sense.
Therefore a 
> counter-check is necessary and highly appreciated.

There has been considerable discussion of how chemistry standard names
could be generated in future, particularly where several thousand
chemical species may be involved.  There are also other ongoing
discussions regarding the design and 'grammar' of standard names.
Within this thread, it was agreed that we should not introduce
'nonsense' standard names, i.e., we should not automatically include
names resulting from the combination of every chemical species with
every possible physical process.  Personally, I support this position
very strongly.  We expend considerable effort in CF agreeing not only
the names themselves, but also their definitions.  I think that this
work on agreeing meaningful concepts would be greatly undermined if we
were to begin including names for quantities that cannot exist.  Another
reason for not automatically including all combinations is that it may
on occasion lead to the duplication of quantities that already exist in
the table by another name.  For example, the first name in Martina's
list is mass_concentration_of_air; kg m-3, but this is already in the
table as air_density.

With the aforementioned points in mind, and with regard to the preceding
discussions of the physical quantities and chemical species proposed
under this thread, there are many names on Martina's original list that
can now be accepted for inclusion in the standard name table, but
perhaps some that should not be because they may represent impossible
combinations.  Not being a chemist myself, I am not the best qualified
person to decide which combinations of species and process make sense.
To assist in identifying the valid combinations within this set of
proposals I have prepared an Excel spreadsheet containing physical
processes as column headings and species as row headings.  I have begun
to populate the cells of this spreadsheet, for example, I have marked
where the names proposed from the combination of process and species are
already in the table.  I have marked as 'OK' those cells where the
combination can be accepted immediately for inclusion in the standard
name table - these consist mostly of the various measures of abundance
such as mass_fraction, atmosphere_mass_content, etc.  I would greatly
appreciate some help in populating the remainder of the spreadsheet with
'YES' or 'NO' entries indicating whether particular combinations are
possible.  In one or two instances I have written comments regarding
specific names.  Please note that at present I have not included in the
spreadsheet any species or process names that are still under discussion
in this thread (i.e., the 'lumped' and grouped species, 'pmx_aerosol'
and 'passive_ozone' and the net_production, respiration,
optical_thickness and mole_flux processes).

I will email Martina, Philip, Martin and Christiane directly for help
with the spreadsheet - you are the chemists whose names I know :) - but
contributions from others would also be most welcome.  The individual
standard name proposals will be accepted or rejected according to the
responses in the spreadsheet.  Any areas of disagreement or ambiguity
that come to light as a result of this exercise will be brought back to
the mailing list for further discussion.

The standard names Guidelines document will need to be updated in
response to Martina's proposals.  We currently have a section in that
document called 'Transformations' which gives a number of 'template'
phrases that can be used within standard names to describe mathematical
quantities, for example, change_over_time_in_X, derivative_of_X_wrt_Y.
Jonathan suggested some time ago (not on the list) that we could
similarly include templates for the processes associated with the
chemical names.  I will add a section called 'Chemical phrases' and list
in there all the currently agreed phrases (such as
tendency_of_atmosphere_mass_content_of_X_due_to_emission).  There is
already a 'Chemical Species' section, which is well out of date, and to
which I will now add all the currently agreed species.  Both sections of
the Guidelines will then be kept up to date as new chemical species and
phrases are agreed.

Within this thread there has been some discussion of taking a 'two table
approach' to chemistry standard names, i.e., maintaining separate tables
of agreed chemical processes and species.  A number of advantages and
disadvantages of this approach have been discussed.  Philip pointed out
that one of the big advantages of a two table approach is that it makes
searching for existing names easier, e.g., it is easier to find out
whether a particular species has already been agreed for use in standard
names.  I hope that expanding the standard name Guidelines as described
above will go some way towards addressing the issue of searchability.
In fact, I think it would be quite easy to maintain the lists of
chemical phrases and chemical species as separate documents, linked to
from the Guidelines, but which could then be viewed separately if
needed.  Would that be a useful thing to do?  When we are discussing
proposals for new standard names it is often useful (as Martina has done
in this thread) to separate the discussion of phrases and species and I
hope we will continue to do that. Regarding the standard names
themselves, at present the general weight of opinion seems to be that it
is better to keep the process and species together in a single string
within the CF metadata, i.e., no change to the current design of
standard names.  If in future it is felt that there is a strong need to
allow the two table approach as an alternative, a change to the CF
conventions would probably need to be proposed.

Martin and Martina have put forward the suggestion of a panel drawn from
the atmospheric chemistry community to coordinate future proposals to CF
for chemistry standard names.  I certainly welcome the willingness of
the atmospheric chemists to engage so fully with the CF process and
having domain experts to call on when dealing with particular sets of
proposals is very valuable.  It seems likely that there would be some
overlap between the function of a chemistry panel and the existing
standard names committee so I will need to discuss further with Bryan
Lawrence, chair of the CF standard names committee, on the best way to
coordinate activities.

Best wishes,

> -----Original Message-----
> From: Martina Stockhause [mailto:martina.stockhause at zmaw.de]
> Sent: 20 February 2009 08:18
> To: Pamment, JA (Alison)
> Cc: cf-metadata at cgd.ucar.edu
> Subject: Re: [CF-metadata] Standard Names - Status Summary
> Dear Alison,
> regarding the thread "CF standard names for chemical constituents and
> aerosols":.
> I think we agreed on the new physical quantities, for the chemical
> names I am not sure.
> Then the discussion shifted towards the more general topic of how to
> integrate
> chemical CF standard names in future and afterwards to general
> construction rules for all kinds of CF standard names.
> My original proposal got a bit lost in these general discussions.
> Therefore I would like to focus on the chemical names again.
> 1. Physical Quantities
> In the mails from the 24.10.2008 we agreed on:
> tendency_of_atmosphere_mass_content_of_X_due_to_respiration :
> 1
> tion :/
> /mol.m-3.s-1
> ction
> :/
> /mol.m-3.s-1
> density_of_surface_area_of_X, m-1, This is the surface area of aerosol
> per volume of atmosphere (i.e. /m2.m-3=m-1)/
> 2. Chemical Quantities
> For the new chemical species I am not that sure. I recall a discussion
> about 'lumped' but cannot remember an agreement. The full list of
> for chemical species can be found in the wiki:
> http://wiki.esipfed.org/index.php/CF_Standard_Names_-
> _Construction_of_Atmospheric_Chemistry_and_Aerosol_Terms
> Regarding my proposal, I tried to keep track and the new names I found
> in my files are:
> 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
> total_peroxy_radical
> passive_ozone
> noy_expressed_as_nitrogen
> clox_expressed_as_chlorine
> brox_expressed_as_bromine
> lumped_alkanes
> lumped_alkenes
> lumped_aromatic_compounds
> lumped_terpenes
> nmvoc_expressed_as_carbon
> anthropogenic_nmvoc_expressed_as_carbon
> biogenic_nmvoc_expressed_as_carbon
> lumped_oxygenated_hydrocarbons
> 3. Construction of "chemical CF names"
> The discussion about a naming convention and construction guideline
> chemicals and aerosols is summed up by Christiane in the wiki:
> ames.
> The status there is:
> <physical quantity>_of_<chemical quantity>_due_to_<process>
> e.g.
> _emission
> The position of <medium> is not absolutely clear, I would think:
> <physical quantity>_of_<chemical
> If it is not a medium like "air" but a part of the earth system like
> "atmosphere" it is not agreed (as far as I know), whether it should be
> placed like the medium or in front.
> Actually, the chemical names in the CF construction guideline
> http://cf-pcmdi.llnl.gov/documents/cf-standard-names/guidelines are a
> bit out-dated and needs to be revised or replaced by a link to the
> above
> mentioned wiki page of the chemists.
> What is still open to me (or I missed the agreement), is the question
> of
> whether the CF standards names with each individual chemical species
> are
> part of the CF convention or only the physical quantities and
> with a place-holder for the chemical quantity.
> Heinke proposed a possibility how we could deal with the place-holder
> (she called "constituent") in netCDF-headers, about which I could not
> find a real discussion,
> http://mailman.cgd.ucar.edu/pipermail/cf-metadata/2008/002421.html:
> dimensions:
> lat=72;
> lon=124;
> maxlen=30;
> float quantity(lat,lon,height,time) ;
> quantity:standard_name = "mass_fraction_of_constituent_in_air" ;
> quantity:units = "1" ;
> quantity:coordinates = "constituent" ;
> char constituent(maxlen);
> constituent:standard_name='IUPAC_system';
> data:
> constituent='ozone';
> I hope to revive the discussion to reach an agreement of this rather
> old
> thread as well, at least on the first and second parts (1. and 2.).
> Best wishes,
> Martina
> --- CliSAP - Integrated Climate Data Center ---
> Martina Stockhause
> University of Hamburg
> CliSAP
> Grindelberg 5
> D-20144 Hamburg
> Germany
> phone: +49-40-42838-7582
> FAX:   +49-40-42838-7711
> e-mail: martina.stockhause at zmaw.de
> -----------------------------------------------

Alison Pamment                          Tel: +44 1235 778065
NCAS/British Atmospheric Data Centre    Fax: +44 1235 446314
Rutherford Appleton Laboratory          Email: alison.pamment at stfc.ac.uk
Chilton, Didcot, OX11 0QX, U.K.


Scanned by iCritical.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: proposal20080917_new.cf
Type: application/octet-stream
Size: 148410 bytes
Desc: proposal20080917_new.cf
URL: <http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20090424/f8fb4be3/attachment-0006.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: grib2_chem_names.xls
Type: application/vnd.ms-excel
Size: 54272 bytes
Desc: grib2_chem_names.xls
URL: <http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20090424/f8fb4be3/attachment-0006.xls>

More information about the CF-metadata mailing list