add wave variable list for Med-CORDEX

[sam-somot]

This is the start of the discussion on the wave variable list for Med-CORDEX domain as some of our coupled models do have wave coupled.
For the community, @fulden @gkara00 are in charge on defining this list for the community.
@jesusff advised us to start the list directly in csv format in the github.
@larsbuntemeyer could you advise us how to do it ? or create the starting csv file empty for example and help @gkara00 and @fulden to go on. thx

[jesusff]

That was the wrong Fulden 😉 Let's try this one @batibenizf (not sure either...)

If we go for the merged version (#68 let me know what do you think), the CVS would be just the one for Med-CORDEX, simply setting the "component" to wave or ocean_waves or the name you'd prefer, and adding lines similar to those already there. We need to follow CF standard names*. We can have a meeting to clarify the procedure, anyway.

(*) Search for "wave" here https://cfconventions.org/Data/cf-standard-names/current/build/cf-standard-name-table.html Unfortunately it mixes a lot of radiation standard names (longwave, ...)

[jesusff]

I copy here an e-mail exchange which is relevant to keep track of the progress of this data request. This needs to be taken by the new task team on infrastructure:

24 July 2025 – Ioannis Karagiorgos

Dear All,

As a follow-up to the discussions during the recent Med-CORDEX workshop, I have initiated together with Fulden Batibeniz (in CC) a list of output variables to be provided by the wave model components of the Med-CORDEX Phase 3 simulations.

If you have any suggestions, or if there are specific wave variables and/or output frequencies that would support your research and applications, we would greatly appreciate your input.

Best regards,

John

1 September 2025 – Jesus Fernandez

Dear John, Fulden,

I write regarding some new developments regarding wave variables. During our meeting, we discovered some inconsistencies in the CMIP7 definition of instantaneous vs mean output. I opened an issue on this regard:

WCRP-CMIP/Variable-Registry#35

and Karl pointed me to a much deeper discussion they are having in CMIP regarding wave variables:

CMIP-Data-Request/Harmonised-Public-Consultation#78

Since these variables are new in CMIP, it is crucial to have them defined in a way which is also useful for us in CORDEX. In short, all WCRP variables will be defined in a single variable registry that will be used by both CMIP and CORDEX. Even though aliases can be defined to account for different namings, it is much better to agree and participate in the definition from the beginning.

Could you have a look at the discussion above and see how it fits with your current idea for the wave output in Med-CORDEX? I remember from our meeting the idea of having several variables defined for the same quantity computed in different ways (first vs second moment?), but this is no longer in your Google sheet? did you discarded this?

Cheers,
Chus

4 September 2025 – Ioannis Karagiorgos

Hi Chus,

I found the discussion in the github issue very useful, particularly for the choice of instantaneous wave outputs at the higher 3-h frequency instead of time-mean fields.
Just an update for our case; I have modified the Google sheet after discussions, where I’ve currently grouped variable pairs (swh, mean period, peak period, mean direction, peak direction) for total wave under the CORE group, and for windsea/swell components in TIER1, with outputs at 3-h (instantaneous), daily (mean), and monthly (mean) frequency. For TIER2, I’ve included the request for higher-frequency data of 1-h.
Also, there has been interest in variables not currently included in the CFTables or CMIP request, such as "phioc; energy by wave-breaking inducing mixing" and "tauoc; Ocean stress modified by wave induced stress". These are available, for instance, in the ERA5 dataset. I’m unsure whether it would be appropriate to propose them for TIER1 or TIER2, even without an established CF name, but I think it’s worth considering. Do you have any suggestion about this?
Regarding the different wave mean period variables, I've kept for now only the CF variable "sea_surface_wave_mean_period" for consistency with the CMIP data request. I saw that this is an alias to zero upcrossing period, so the second moment wave period gives the closest result. However, it might be useful to clarify this in the https://github.com/CMIP-Data-Request/Harmonised-Public-Consultation/issues . Is this the right place or is there another point of interaction with the CMIP community?
Thanks a lot for your help!
Best regards,
John

5 September 2025 – Jesus Fernandez

Hi John,

the decision between Tiers 1 and 2 should be based on whether the variables can be consistently computed across models (Tier1) or not (Tier2). If there's a risk that the quantity is not fully comparable across models since it can be derived in different ways, it should go to Tier2.

Regarding new variables, it is fine to propose new ones, even if there's no CF standard name for them. In such a case, we should request the inclusion of the new standard names to the CF community. E.g. here an example: cf-convention/vocabularies#226 We can check with Karl if it is OK to include the extra CORDEX variables in the WCRP variable registry or they prefer to keep it for CMIP for now (they just started this effort). In any case, we can consider them in CORDEX.

Regarding the different ways to compute the variables, yes, there are a couple of issues which are still open and where you could rise this concern:

CMIP-Data-Request/Harmonised-Public-Consultation#78
CMIP-Data-Request/Harmonised-Public-Consultation#61

Thank you!
Chus

23 September 2025 – Fulden Batibeniz

Hi Chus,

Thanks for pointing us to the relevant discussions — and apologies for my delayed reply, I was travelling. I’ve now caught up, and John has already updated the list, which looks good to me.

Best,
Fulden

[gkara00]

I’ve added the first version of the wave variable list to the Med-CORDEX data request CSV, here: Med-CORDEX/internal-data-request#6
Please feel free to leave comments or suggestions!

[pierrenabat]

Hello, I would like to make an update on this wave variable list in the framework of the CORDEX-TTPI. Do you know if it can be integrated in the official CORDEX data request ?

[jesusff]

yes, we should merge this into the official data request. The https://github.com/Med-CORDEX/internal-data-request repo is intended for early exchange of data for particular studies within Med-CORDEX. But in your PR @gkara00 you have the full proposal of variables to be saved from the waves model component, right?

We will discuss tomorrow in the TTPI how to finally handle these component requests in this repo and we will merge your proposal here for further refinement if needed

[gkara00]

yes, @jesusff . It's an updated version of the csv for all the proposed wave variables. Of course, it's under discussion and we can modify it accordingly.