[CF-metadata] area_type: convention and usage: bringing the checker and convention into line

From: Jonathan Gregory <j.m.gregory>
Date: Mon, 23 Mar 2015 15:46:01 +0100

Dear Martin and Jim

Thanks for this point.

> The alternative is to continue to use area_type, and to explain to
> people that they need to request new area types when the existing
> ones don't match their model outputs and explain to them how to make
> the requests. Doable, but it might generate a large and confusing
> number of area type table entries.

In fact area_type was introduced as a standard_name so that we could follow
a generalised approach to area fractions, with a string-valued coordinate,
because in a given grid cell there might be various types of vegetation and
other non-vegetated types like sea and ice. I think this approach is a good
one, and that more area types should be proposed to meet the needs of CMIP6.
I guess that there must be experts on this in some relevant MIPs. This would
enable comparison of datasets from different models, which is the main purpose
of CF. If they all use different descriptions of vegetation types that purpose
isn't served well.

Best wishes

Jonathan
Received on Mon Mar 23 2015 - 08:46:01 GMT