[CF-metadata] request for new netCDF standart names

From: Jonathan Gregory <j.m.gregory>
Date: Thu, 9 Feb 2006 21:45:41 +0000

Dear Elke

> 1. forest_area_fraction
Obviously we could define a lot of names like this, for different types of
vegetation, etc. I think therefore we should consider something more generic.
I propose that we introduce a standard name of area_fraction, and require that
a variable with this standard name should have a coordinate which specifies
somehow what quantity is being given an area fraction. For your case you would
use a string-valued coordinate variable of land_cover (recently defined). Your
data variable of forest area fraction would therefore be described thus:

variables:
  float area_fraction(land_cover,lat,lon);
    area_fraction:standard_name="area_fraction";
    area_fraction:units="1"; // optional
  char land_cover(surface_cover,stringmaxlength);
data:
  land_cover="forest";

A scalar coordinate variable could alternatively be used:

variables:
  float area_fraction(lat,lon);
    area_fraction:standard_name="area_fraction";
    area_fraction:units="1"; // optional
    area_fraction:coordinates="land_cover";
  char land_cover(stringmaxlength);

If you have fractional land coverage of grid boxes and wish to specify whether
the forest fraction is expressed per unit area of land or per unit area of the
whole grid box, this can be done with cell_methods="lat: lon: mean over land"
or "lat: lon: mean" (this is a convention we agreed earlier in this list but
have not added to the standard yet).

What do you think?

I agree with Bryan that strings such as the possible values of area_fraction
should be tabulated and standardised eventually, as we have done for other
lists such as region names.

> 2. skin_reservoir_content
>
> A part of this quantity is the canopy water amount but additionally this
> quantity contains the wetting of the bare soil. MKS unit=m

By "wetting of the bare soil" do you mean liquid water standing on the surface
(not soil moisture content)?

> 3. frozen_soil_fraction
>
> This quantity describes the part of the soil which is frozen, MKS unit=1

We already have mass_fraction_of_frozen_water_in_soil_moisture. Is that what
you mean? If not, could you describe more precisely what fraction you mean?

> 4. surface_snow_amount_change
>
> This quantity describes the change of the snow amount at the surface,
> MKS unit=kg/m**2

I don't think we should define such a name. We'd have to do that for every
variable, potentially! But I can understand that you want to distinguish a
change from an unchanged quantity. Perhaps it would be more informative to
express the change as a rate of change in kg m-2 s-1. In that case you could
also use the time bounds to show the time interval over which the change
occurred; that information could be useful.

If we need to provide special names for changes, I think we'll have to
introduce a new mechanism for it. Comments welcome from everyone.

Best wishes

Jonathan
Received on Thu Feb 09 2006 - 14:45:41 GMT