Dear Dan
> I found the list on the CF web site here:
> http://cfconventions.org/Data/cf-standard-names/docs/area-type-table.html
You probably saw John Graybeal's posting that he has put the up-to-date
list at
http://mmisw.org/ont/cf/areatype.
> The only references to area_type in the CF conventions doc (v1.6) appear to be in section 7.3.3 Statistics applying to portions of cells. However our data are point values (obtained by interpolating point observations) so we are not defining bounds for the spatial coordinates and therefore do not have a cell method for 'area'. Is it still possible to use area_type as a coordinate variable in this situation?
Yes. It has a special role in that section, but it is always permissible to
attach coordinate variables to quantities to describe the data variable, and
the area_type is often needed for such a purpose.
> x=180
> y=290
> time=UNLIMITED
> ntypes=1
> maxlen=20
>
> lat(y,x)
> lon(y,x)
>
> surface_type(ntypes,maxlen)
> surface_type:standard_name="area_type"
> surface_type="grass"
>
> surface_temperature(time,ntypes,y,x)
> surface_temperature:coordinates = "lat lon surface_type"
> surface_temperature:cell_methods = "time: minimum within days time: mean over days"
> Personally I'm not that keen on having to add a dimension
You don't have to add a dimension. You can use a scalar coord var, thus:
> x=180
> y=290
> time=UNLIMITED
> maxlen=20
>
> lat(y,x)
> lon(y,x)
>
> surface_type(maxlen)
> surface_type:standard_name="area_type"
> surface_type="grass"
>
> surface_temperature(time,y,x)
> surface_temperature:coordinates = "lat lon surface_type"
> surface_temperature:cell_methods = "time: minimum within days time: mean over days"
This might be a good subject for the FAQ because it has come up before.
Best wishes
Jonathan
Received on Fri Sep 12 2014 - 06:49:23 BST
