[CF-metadata] CMIP6 Sea Ice MIP: Integrated quantities

From: martin.juckes at stfc.ac.uk <martin.juckes>
Date: Mon, 25 Jul 2016 09:15:54 +0000

Dear Jonathon, Dirk,

I agree with Jonathon that it is a good idea to follow the approach used in other names (e.g. number_of_days_with_air_temperature_above_threshold) and use a coordinate for the threshold value .. though I would prefer to make it obligatory to state the threshold in order to avoid ambiguity.

Another issue is the domain .. the variable is being defined as the total area of cells above the threshold within a specific domain (the N. hemisphere of S. hemisphere in the CMIP6 data request). The threshold examples that I could find in the existing standard names are all temporal statistics (e.g. number of days with quantity X above a threshold within a year) and so use a climatological time variable. Here we want to specify the area over which cells are being summed. Would it make sense to use a 2nd coordinate variable, with standard name "region" specifying the region (this would also require adding "northern_hemisphere" and "southern_hemisphere" to the region list)? Alternatively, we could use latitude and longitude coordinate variables with bounds attributes to declare the region.

The region will be in the long name, of course, but it would be good to have an agreed approach to encoding it in CF,

regards,
Martin


Dear Dirk

> For sea_ice_extent, there is currently no definition in the CF
> convention. This variable is somewhat odd in a CF context, as it only
> makes sense when applied over several grid cells.
>
> Hence, we suggest to adopt a definition which implies usage over several
> grid cells:
>
> sea_ice_extent: Total area of all grid cells that are covered by a
> minimum of a given area fraction [usually 15 %) of sea ice in a
> specified region".

I agree that it needs a definition. It is like sea ice area (m2) in being an
extensive quantity which is integrated over the area considered, but I agree
that extent is not meaningful unless there is some discretisation of area.
I would suggest that we could allow for the possibility of the threshold
being other than 15%, following practice for other standard names (as in
recent emails of Martin's), like this

sea_ice_extent: Total area of all grid cells in which the sea ice area
fraction equals or exceeds a threshold. By default the threshold is 15%.
The threshold can be specified by supplying a coordinate variable or scalar
coordinate variable with standard_name of sea_ice_area_fraction.

Would something like that be OK?

Best wishes

Jonathan
Received on Mon Jul 25 2016 - 03:15:54 BST