[CF-metadata] New standard names for OMIP: physics

Jonathan Gregory j.m.gregory at reading.ac.uk
Mon Oct 23 14:15:53 MDT 2017


Dear Alison   cc Steve, Paul

Sorry for slow response. I've read Steve's edited version and I agree with
your and Steve's proposals. Thanks. Changing all the integral_of_X_wrt_Y to
integration_wrt_Y_of_X will make them more readable, I think, and it's a
neater construction. In fact people sometimes write \int dy X intead of
\int Y dx in mathematical formulae, because it's natural to think of \int dy
as an operator.

Regarding the residual_mean_advection names, I agree that this is what we
want for OMIP and FAFMIP. However if you search for due_to_advection you will
find quite a few existing ocean ones e.g.
tendency_of_sea_water_temperature_due_to_advection
(K s-1). Is it clear in the definitions what sort of advection this is (I am
offline and cannot check)? I believe that in these existing names we mean
resolved advection i.e. not including parametrised eddy advection. We do not
mean residual mean advection. Maybe it would be a good idea to change
due_to_advection to due_to_resolved_advection and say in the definition that
this means advection by the model's sea water velocity field, not including any
parametrised subgridscale advection. In the atmos we also have some due_to_
advection names, but models don't have parametrised advection in the atmos,
so there's not the same issue.

If you agree with that, I would also advocate including new standard names
like 3n,o,p but with resolved_advection. They are not needed for the CMIP6
request, but they probably are useful. I'd use them for NEMO and UM ocean
model diagnostics, for example!

Best wishes

Jonathan



More information about the CF-metadata mailing list