Global-mean ocean variables at monthly frequency: rho_dzt only available yearly

[rbeucher]

Several global-mean ocean variables are required at monthly frequency:

• bigthetaoga
• soga
• sosga
• thetaoga

Proposed method: Global weighted mean using:

total_mass = rho_dzt * area_t

Problem: rho_dzt is only available as a yearly mean.

Questions:

• Is using yearly-mean density acceptable for monthly global means?
• Do we need monthly rho_dzt for CMIP compliance?

[dougiesquire]

The ESM1.6 configurations use conservative temperature as the prognostic temperature. The temp_global[surface]_ave variables are the averages of the prognostic variables (e.g. here). So:

bigthetaoga

I think we already have this at monthly frequency, output as temp_global_ave. Note I think the standard_name for this output is incorrect. It should be sea_water_conservative_temperature, not sea_water_potential_temperature.

soga

I think we already have this at monthly frequency, output as salt_global_ave.

sosga

We already save both of the following at monthly frequency:

• salt_surface_ave which is a mass-weighted mean
• surface_salt which is the top layer

I think the former is what we want, though we possibly want to take an area-weighted mean of the latter. I couldn't find any clear CMIP guidance for this.

thetaoga

This is a little less straightforward. I can't see any option in MOM5 to output global average potential temperature when the prognostic temperature is conservative. So we will probably have to do the averaging ourselves, and yes, that would require rho_dzt at monthly frequency (we already save 3D pot_temp at monthly frequency). How was this handled for previous CMIP submissions?

[rbeucher]

Thanks @dougiesquire, that's helpful

[rbeucher]

temp_global_ave is not available in /g/data/p73/archive/CMIP7/ACCESS-ESM1-6/spinup//Dec25-PI-control/output*/ocean/
salt_global_ave is not available in /g/data/p73/archive/CMIP7/ACCESS-ESM1-6/spinup/Dec25-PI-control/output*/ocean/
salt_surface_ave is also not available.

I can see: surface_salt (ocean-2d-surface_salt-1monthly-mean-ym*)

[dougiesquire]

They're in ocean-scalar-1monthly-ym_*.nc along with a bunch of other scalar diagnostics

[rbeucher]

I'm on this

[rbeucher]

OK @dougiesquire , looks like bigthetaoga, soga, thetaoga are required for all ocean levels...so essentially olevel, time variables.

[dougiesquire]

OK @dougiesquire , looks like bigthetaoga, soga, thetaoga are required for all ocean levels...so essentially olevel, time variables.

I'm confused. I thought the *ga variable were global averages over the whole ocean volume?

[rbeucher]

The data request is confusing here I agree... Here is what the UK MO says (NEMO related but helpful) UKNCSP/CDDS-CMIP7-mappings#1381

[dougiesquire]

I've added monthly rho_dzt in ACCESS-NRI/access-esm1.6-configs#641

[dougiesquire]

The data request is confusing here I agree... Here is what the UK MO says (NEMO related but helpful) UKNCSP/CDDS-CMIP7-mappings#1381

Hmmm, yeah. From that issue:

In CMIP6, this mapping would simply have pointed to the NEMO diagnostic temptot, which is the volume-averaged temperature. In CMIP7, bigthetaoga is no longer averaged over depth and is now a global-mean temperature profile, rather than a volume-averaged scalar.

Now that monthly rho_dzt is added, I think we'll have everything we need to compute the new definitions of bigthetaoga and soga. But changing the definition of the variable is just asking for confusion!