[CF-metadata] Duplicate vocabulary attenuation/extinction and backscatterin/backwards_scattering.

From: Markus Fiebig <Markus.Fiebig>
Date: Mon, 12 Mar 2012 13:27:40 +0000

Dear Philip, Jonathan, Uma, and Martin,

being busy with writing a longer reply on Philips thorough review of my proposed standard names, I'd like to make a comment concerning the use of "backscattering" and "backwards_scattering". I proposed hemispheric in "hemispheric_backscattering_coefficient", which is integrated over the rearward hemisphere as measured by some integrating nephelometers, from the "backscattering_coefficient" that isn't integrated over any solid angle (to be proposed soon), e.g. as measured by a lidar. In the examples you mention however, the terms "backscattering" and "backwards_scattering" are both already used for a property integrated over a solid angle. I've pondered over this one quite a while, but I keep coming back to the point that a standardised word for distinguishing angular integrated from non-angular integrated scattering properties is needed - like "hemispheric". That wouldn't be a nice solution, since it implied aliases for 5 existing standard names, but it would be structured and consistent. Any better ideas?

Best regards,
Markus


-----Original Message-----
From: Jonathan Gregory [mailto:j.m.gregory at reading.ac.uk]
Sent: Montag, 12. M?rz 2012 14:14
To: Cameron-smith, Philip
Cc: Markus Fiebig; Shankar, Uma; X:m.schultz at fz-juelich.de; cf-metadata at cgd.ucar.edu
Subject: Re: [CF-metadata] Duplicate vocabulary attenuation/extinction and backscatterin/backwards_scattering.

Dear Philip

> 1) _attentuation_ and _extinction_ appear to have the same physical meaning, although the comments note that attenuation is more commonly used for radar and extinction is more commonly used for visible light. The number of std_names using each of them is 2 and 1, respectively, so it should be easy to fix with aliases.
> 2) _backscattering_ and _backwards_scattering_ also appear to have the same definition. The number of std_names using each of them is 1 and 4, respectively, so it should be easy to fix with aliases. Although it is less common, I prefer _backscattering_ because it is a single word, and will generalize better to _forwardscattering_.

These are good points and I tend to agree with you. I am not an expert and if there is a distinction someone will point it out, I hope. If there is no distinction, the reason for the different choices would be because of what is generally said, or from personal preference, I presume. It's a difficult balance between making standard names use familiar terms, or making them use consistent terms, but I would prefer consistency if it is still obviously comprehensible to an expert.

Regarding the Mie scattering proposal, again I am not an expert on the science, but I would comment that X_assuming_Y is generally used to indicate what X would be in some hypothetical situation of assuming_Y e.g.
  surface_albedo_assuming_no_snow
which is not, in general, the same as surface_albedo as it truly is (although it is often the same). It does not indicate a method of calculating X and, as you say, we keep that out of standard names, because in principle a method of measurement should not affect the definition of the quantity being measured.

Best wishes

Jonathan
Received on Mon Mar 12 2012 - 07:27:40 GMT