Hi John,
Your suggestion that we disallow storage of NBS scale pH in CF by not creating a Standard Name for it worries me. We have significant quantities of NBS-scale pH in BODC (from estuarine systems and the days before ocean chemists discovered the kilogram) that we may wish to put into CF in the future. Whilst conversion based on approximation to free_scale is possible in some (but not all) cases (need additional data), it's something I'd rather not do. I would much rather label precisely and leave any manipulations up to the user who is aware of the proposed usage and what is fit for that purpose.
My attraction to Jonathan's simple 'pH_in sea_water' is that it gives us a way of handling data in CF where 0.1 pH units don't matter. We have one dataset monitoring waters off a chrome plating plant where pH varies between 2 and 6. At the same time I am acutely aware of the effects errors of that magnitude have on deep ocean carbon budget calculations.
Cheers, Roy.
-----Original Message-----
From: cf-metadata-bounces at cgd.ucar.edu [mailto:cf-metadata-bounces at cgd.ucar.edu] On Behalf Of John Graybeal
Sent: 02 May 2009 01:28
To: CF Metadata List; Lowry, Roy K
Subject: Re: [CF-metadata] new standard name request for pH
Hi everyone,
Thanks very much for considering and responding to the pH terms
proposal. To respond, I spent considerable time being educated by our
in-house pH expert, even reading pages myself from the referenced book
[1], and reviewed this email with him. So I claim this email
represents a high level of expertise in this scientific field. I
apologize for the length, but it seemed necessary to present
justifying details.
To recap, the original proposal as revised was for 4 names related to
pH:
A) sea_water_pH_NBS_scale (moles/liter)
B) sea_water_pH_free_scale (moles/kg)
C) sea_water_pH_total_scale (moles/kg) <-- the one we care about
D) sea_water_pH_seawater_scale (moles/kg)
Through this thread we have had several suggestions for replacements
[5], most recently arriving at
a) pH_of_sea_water_defined_by_moles_of_hydrogen_ion_per_unit_mass
b) pH_of_sea_water_defined_by_mole_concentration_of_hydrogen_ion
c)
mole_concentration_of_H_and_HSO4_per_unit_mass_in_sea_water_expressed_as_pH
d) mole_concentration_of_H_and_HSO4_and_HF_per_unit_mass
in_sea_water_expressed_as_pH
e) (added) pH_of_sea_water
I believe the focus of these responses was on the concentration of the
solution, but this mis-states the actual chemical definition of this
particular concept. (Unfortunately we incorrectly presented this in
the original definitions for B, C, and D.) In the reference[1] it is
made clear that:
(a) pH is strongly affected by abundance of hydrogen (actually
hydrate) ions, but is not defined by their quantity; it is determined
according to the *activity* of those ions, and this is not a linear
relationship (though it is close to 1 in some regimes)[2]. We will
change the proposed definitions to replace 'concentration' with
'activity'; my apologies for this error.
(b) The use of the NBS scale is not recommended for sea water[3], as
errors will be induced (up to at least 0.1 units in situ); in this
light we propose to simply withdraw [A] from consideration, especially
as we don't need it.
(c) There are significant differences in the values obtained for these
different quantitiies, up to .12 units; thus to say just
'pH_of_sea_water' is to have a very loose concern about the accurate
meaning of the so-labelled pH values[4]. For comparison, measurements
are often made, and needed, at the level of 0.0005 units.
(d) All 3 of {B, C, D} above are in regular use, but (C) is the most
promising for the common reference[4].
Additionally, the units are expressed as noted above by common
convention (see for example UCUM units section on pH [5]), in order to
make certain conversions handy for the enlightened. Yet the second
sentence of the following statement does not hold true for pH, given
the depths at which these measurements are being made:
On Apr 30, 2009, at 9:22 AM, Jonathan Gregory wrote:
> Boussinesq models (most ocean climate models are Boussinesq) treat
> density
> as constant 1000 kg m-3 except in the computation of pressure
> gradients, where
> it matters to the dynamics. Therefore in dealing with concentrations
> of
> tracers, per kg and per litre are identical and to choose one or the
> other
> would be arbitrary and hence unhelpful for data exchange.
We therefore do not believe a substitution of moles/liter or moles/
m**3 is appropriate for the canonical moles/kg, as this would force
most practitioners to convert their data before naming it with this
name. [7] Obviously if there is a set of practitioners that are
working with seawater pH using a different canonical unit, that would
be another matter to consider; we think it is unlikely that any
observationalists are doing so.
So, in summary, the longer, explicit terms being proposed are so
approximate (for the reasons descrX-MailScanner-NERC-SpamCheck: not spam, SpamAssassin (not cached, score=-3.5,
required 5, autolibed above) as to be inaccurate to
the scientists that use this data. Indeed, they would not use these
terms. Therefore we re-propose the original terms (B), (C), and (D),
or alternatively simply (C), as these represent the common and most
descriptive usage.
John
[1] Zeebe & Wolf-Gladrow (2001). CO2 in Seawater: Equilibrium, Kinetics,
Isotopes. Published by Elsevier as part of their oceanographic series.
[2] Ibid, p 54 ff. The exact quote was helpful to me, so I provide it
here: "In high school we all learned that pH is the negative common
logarithm of the concentration of hydrogen ions. ... Things are a bit
more complicated for several reasons... The symbol H+ represents
hydrate complexes rather than the concentration of free hydrogen
ions.... In a refined theory one should use activity (an 'effective'
concentration...) instead of concentration...." It goes on to explain
the derivation of the various scales, in case that is of interest.
[3] Ibid, p 55
[4] Ibid, p 59
[5] Section 45 in
http://aurora.regenstrief.org/~ucum/ucum.html#section-Other-Legacy-Units
(which unfortunately does not agree with the cited reference [1]
about the meaning of this term.
[6] Proposal chain for reference:
a1) pH_per_unit_volume_in_sea_water =>
a2) pH_of_sea_water_defined_by_mole_concentration_of_hydrogen_ion =>
a3)
pH_of_sea_water_defined_by_mole_concentration_of_hydrogen_ion_per_unit_volume
=>
a4) pH_of_sea_water_defined_by_moles_of_hydrogen_ion_per_unit_mass
b1) pH_per_unit_mass_in_sea_water =>
b2) pH_of_sea_water_defined_by_mass_concentration_of_hydrogen_ion =>
b3)
pH_of_sea_water_defined_by_mole_concentration_of_hydrogen_ion_per_unit_mass
=>
b4) pH_of_sea_water_defined_by_mole_concentration_of_hydrogen_ion
c1)
mole_concentration_of_H_and_HSO4_per_unit_mass_in_sea_water_expressed_as_pH
d1) mole_concentration_of_H_and_HSO4_and_HF_per_unit_mass
in_sea_water_expressed_as_pH
e1) pH_of_sea_water
[7] If I understand CF canonical units correctly, it is OK to have
one's standard-named data in units that are physically equivalent to
the canonical unit -- for example, moles per gram is OK if the
canonical unit is moles/kg -- but it is not OK if the conversion is to
a fundamentally different unit -- like moles / liter in the example.
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.
Received on Thu May 07 2009 - 06:14:26 BST
