Sorry for the delayed reply,
On 28-11-14 11:29, Jonathan Gregory wrote:
> Dear all
>
> I agree with Steve and Roy about this:
>
>>> Now UDUnits has "molecule" and "byte". They (Steve)
>>> are receptive to well-justified proposals, so do not feel daunted
>>> in transmitting to them the suggestion for "photon".
>>
>> Adding such units is the wrong way to go about solving the problem of
>> adding quantity-semantics to data values. It would be much better to have
>> the names of the variables be things like
>>
>> "number_of_tents"
>> "number_of_clouds"
> etc.
>>
>> and whose units were "1" than to try to incorporate such semantic
>> information into a unit
>
> That is exactly what the CF standard says. From Sect 3.1 on "Units":
>
> "Descriptive information about dimensionless quantities, such as sea-ice
> concentration, cloud fraction, probability, etc., should be given in the
> long_name or standard_name attributes rather than the units."
I don't agree completely. A concentration or column amount in mol/m3 or
mol/m2 respectively can be converted to/from number concentrations or
column densities expressed in molecules/cm3 or molecules/cm2. However,
the explicit "molecules" is essential here, otherwise UDUnits will not
be of help at all. 1/m3 is not equivalent to mol/m3.
And yes, I agree that you should not try to do this for 'tents', 'moles'
(the furry kind, [1]) or even aerosol particles. Molecules and photons
are an exception in my view because there are large datasets out there
that use explicit molecules and photons for these 'things'.
Best,
Maarten Sneep
[1]
https://what-if.xkcd.com/4/
--
KNMI
T: 030 2206747
E: maarten.sneep at knmi.nl
R: A2.14
Received on Tue Jan 06 2015 - 10:10:22 GMT
