[Carbon cycle variables]: Ocean & exchange fields with the atmosphere

[seferian]

Dear all

Following IPO suggestion, I open this issue for two variables concerning the modelled global carbon cycle.

1. Minor change: the description of fgco2 (ocean carbon flux).
   The definition of fgco2 (as in CMIP6 DR) is confusing.
   As it stands, fgco2 is the ocean gas CO2 flux toward the ocean. But it doesn't say if it is net, natural, anthropogenic or both.
   As flagged to the panel, it will cause massive confusion when analysing concentration and emission-driven models.
   Depending on model complexity of the ocean carbon cycle, fgco2 can represent either the net anthropogenic sink only with fgco2 because there was no riverine inputs of carbon OR the net total (anthro + natural) sink only with fgco2 because there were no riverine inputs of carbon.
   => It won't harm to expand the definition of this variable in the metadata in a way that encourages users to refer to the EMD or the reference paper of the models.

2. Variable addition: Missing field for the emission-driven configuration: fco2_all
   So far the current version of the DR refers solely to fco2_antt, fco2_nat and fco2_fos (as asked in CMIP6).
   From them it is really hard if not impossible to recompute the total net flux toward the atmosphere which is highly relevant for emission-driven model
   => Therefore, I would suggest to add for fco2_all to ease the use of emission-driven model in CMIP7/C4MIP, etc.

Many thanks in advance,
Roland

[ckoven]

Thanks @seferian for raising these-

For your point (1): Is it sufficient if fgco2 is defined as the total flux including natural and anthropogenic? Or if that is the definition used, does an anthropogenic-only variable also need to be defined?

For point (2): how would the definition of fco2_all differ from (fco2antt + fco2nat)? Could you write a longer description of the variable that explains how it should be computed?

[seferian]

Hi Charlie
many thanks for your response

for 1), I don't think we need an additional variable for fgco2. a clarification of the definition is sufficient to warn users of model specs.

for 2), fco2_antt+fco2_nat, this wouldn't work for all model if the definition of the variable is kept as is.
Because fco2_antt excludes fires and forest regrowth
fco2_nat excludes unrepresented processes that is required to balance carbon budget.
This point is related to the definition of fgco2.
if fgco2 is anthropogenic C sink only, then fco2_antt+fco2_nat would work.
If fgco2 is anth+nat(including river-induced fluxes), there will be missing terms.

Here as well, we could solve this by expanding the definition for fco2_nat, leaving users to understand what is provide depending on models specs.

with that being said, we produced many other variables that can be post calculated (e.g., nep = gpp -ra -rh). Therefore having a variable that facilitate the use of emission-driven model such as fco2_all would be useful.

Does it answer your questions ?.
cheers