Separate cf axes for experiment / run?

[ahuang11]

I have two separate model experiments / runs as dimensions and I want to take the difference across them, e.g. ds.cf.diff('run') or something similar (similar request as #140)

I don't exactly see a standard name for separate experiments / runs in the CF Catalog, but maybe I missed it?

[dcherian]

realization?

You can also set a standard name of your choice... We don't validate against the standard name table...

[ahuang11]

What if both runs have their own ensemble members

…

On Wed, Feb 3, 2021, 3:36 PM Deepak Cherian ***@***.***> wrote: realization? You can also set a standard name of your choice... We don't validate against the standard name table... — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#160 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADU7FFXFUIDXILFSAPZWLXTS5G6VHANCNFSM4XBQAJQQ> .

[dcherian]

I don't know of a standard name for what you want. I suggest making one up and setting it as standard_name on the dimensions. Things should just work then.

[dcherian]

maybe @ocefpaf has ideas

[dcherian]

@jbusecke might have ideas too

[jbusecke]

In CMIP6 speak I think run=experiment, but I have no clue about how that relates to cf conventions, def need to read up a bit on that!

[dcherian]

Looks like experiment_id is used (https://esgdata.gfdl.noaa.gov/search/cmip6-gfdl/) and member_id is ensemble member that would be realization in CF standard name language

[ocefpaf]

maybe @ocefpaf has ideas

I wish. I steer away from standard_names!