[CF-metadata] medium 'atmosphere'

From: Jonathan Gregory <j.m.gregory>
Date: Wed, 12 Nov 2008 15:48:48 +0000

Dear Christiane

You have pointed out there are actually two separate issues.

> -> _atmosphere or _in_atmosphere?
> Why are moles_of_G treated differently?

My fault, I'm afraid, as I think I proposed moles_of_G_in_atmosphere. I meant
the same as atmosphere_moles_of_G or atmosphere_number_of_moles_of_G but these
alternatives didn't sound natural, and I thought the proposal was better human
language although it was inconsistent. Perhaps we should change to
atmosphere_moles_of_G. Does that make sense as a phrase?

> For chemical names
> there is another difference, and I was referring to this: atmosphere
> means not-ocean, not-land, not-vergation, etc., in the Earth system, but
> in the atmosphere. This includes in_air, in_clouds, in_precipitation,
> etc. Hence, in_air is a sub-element of (_in)_atmosphere, at least this
> is what I thought we had defined.

I agree that atmosphere means not-ocean etc. It is a "component" in the
guidelines. The standard names for "atmosphere" are properties of the atmos
as a whole. I assume that the "medium" in_air means a local measurement
in air, regardless of whether it is in a cloudy or a cloud-free region.
I suggest that the right way to distinguish these in CF
would be with cell_methods, as they are different parts of the gridbox. We
already have cloud and clear_sky as area_types, so you can specify
e.g. "area: mean where cloud" in cell_methods for an in_air quantity to
restrict it to the cloudy air.

Perhaps clear-sky isn't the best choice, as that refers to a 2D view of the
atmosphere. As these area-types are new, it maybe wouldn't be too late to
change it, for instance to cloud_free.

Quantities measured in precipitation, not in air, are a new medium, I suppose.
Do you need to name such quantities?

Best wishes

Jonathan
Received on Wed Nov 12 2008 - 08:48:48 GMT