.. as defined + published by the MetOffice-internal "NetCDF to GRIB2 Metadata Mapping Group".
Work out how we could align with those definitions.
NB these include GRIB2 local codes, which we currently don't provide at all.
They cover all phenomena needed for current model outputs.
This appears to be "definitive" information on mappings, especially for MetOffice local table codes.
However it's not yet clear whether all this information is fully public.
The official WMO data may be held in a similar format, but again this is not totally clear.