[CF-metadata] ECMWF GRIB code - CF Standard Name Mapping incorrect?

From: Malm Jakob <jakob.malm>
Date: Wed, 2 Feb 2011 13:08:16 +0100

-28163-01--10 20:59, alison.pamment at stfc.ac.uk skrev:
> Dear Jakob,
>
> To try to answer your specific query first: the correct spelling of the standard name is lwe_thickness_of_atmosphere_water_vapor_content. You are correct that standard names should always use the American spelling. The units associated with this standard name are m and the explanation is as follows: " "lwe" means liquid water equivalent. The construction lwe_thickness_of_X_amount or _content means the vertical extent of a layer of liquid water having the same mass per unit area. "Content" indicates a quantity per unit area. The "atmosphere content" of a quantity refers to the vertical integral from the surface to the top of the atmosphere. For the content between specified levels in the atmosphere, standard names including content_of_atmosphere_layer are used. Atmosphere water vapor content is sometimes referred to as "precipitable water", although this term does not imply the water could all be precipitated."
Yes, I read this, and didn't quite understand why the tcwv GRIB message
has cfName=lwe_thickness_of_atmosphere_water_vapor_content (using GRIB
API 1.9.0)...
> If you are looking for a quantity with units of kg m-2 then the standard name to use is "atmosphere_water_vapor_content" which has the following explanation: " "Content" indicates a quantity per unit area. The "atmosphere content" of a quantity refers to the vertical integral from the surface to the top of the atmosphere. For the content between specified levels in the atmosphere, standard names including content_of_atmosphere_layer are used. Atmosphere water vapor content is sometimes referred to as "precipitable water", although this term does not imply the water could all be precipitated."
Excellent, I will use this name (atmosphere_water_vapor_content) instead.
> Hence, the correct standard name equivalent to GRIB code 137 depends on the units that are normally associated with that parameter, which is something I need to check.
>
> Regarding your broader point, you are correct that there are differences between the two documents - the one on the CF website was originally created by Jonathan Gregory to demonstrate that equivalences can be made between CF standard names and other ways of identifying parameters. I don't think it was ever really intended as a reference document and in fact some of the mappings are incorrect. Also, it pre-dates the use of GRIB2 as a data format. We should probably remove it from the CF website and instead include a link to the ECMWF table which is more up to date and which, I believe, is being actively maintained. However, I think that some of the equivalences shown in the ECMWF table are also not entirely correct, as demonstrated by your query. I will pass this on to ECMWF and try to get a definitive answer to your question. If you notice anything else in the ECMWF table that looks strange, please let me know and I'll feed it back to them.
Looking forward to hearing more...

Thanks,
Jakob
Received on Wed Feb 02 2011 - 05:08:16 GMT