CR-CMIP-1-0-0: greenhouse gas concentrations - feedback, observations and issues

[znichollscr]

Known issues:

• no EESC (equivalent effective stratospheric chlorine)
  • this could be done, but would require some effort

[mo-garethsjones]

Documentation, naming convention and metatdata

Good to see such thorough documentation for the CMIP7 forcing datasets, including for GHG concentrations. There is, however, the need for some description and more detail of what the source IDs given for the scenarios mean.

• For the production data available the source ids given are "CR-vl-1-0-0" and "CR-h-1-0-0". It would be really helpful to users to have some context about what the different parts of the ids represent?

• From comparing with the CMIP7 Scenariomip preprint I assume that "h" stands for the "High (H)" scenario. I also assume "vl" stands for "very low", but the preprint does not have that scenario. There are "Very Low with Limited Overshoot (VLLO)" and "Very Low after High Overshoot (VLHO)" scenarios, but no "vl" scenario. What does that represent? Not only should this be explained in the documentation, but it should be clearly linked to what is in the description paper.

• For the downloaded data from input4mips the file metatdata do not make it clear what the scenario's are for. This is in contrast to what was done for CMIP6 ... but then the "ssp...." naming convention made it very clear! The use of single or two letters naming convention for CMIP7 seems a strange choice.

[znichollscr]

Hi Gareth, thanks for the comment

For the production data available the source ids given are "CR-vl-1-0-0" and "CR-h-1-0-0". It would be really helpful to users to have some context about what the different parts of the ids represent?

There is this page, does that help? https://input4mips-cvs.readthedocs.io/en/latest/extracting-scenario-from-source-id/

From comparing...

Yes there is an issue here. ScenarioMIP have altered their names (vllo is now just vl and vlho is now ln) but the pre-print has not been updated and there is no other really official public source to point at (this comment is the best we have WCRP-CMIP/CMIP7-CVs#1 (comment)). Once the next ScenarioMIP paper iteration is released, this will hopefully become clearer and the link with the paper will be clear so it will be obvious where to find all the details.

In terms of how we write our docs, a question: as an ESM team, do the actual details of the scenarios/names matter for your workflows that actually runs things? I ask because, at the moment we're writing the docs in a way where all the scenarios are basically the same. Whether it's name is "h" or "vl" or "m", it's just a scenario and the details of what it is don't matter for the technical question of "How do I make this thing run?" If that view is wrong, please let me know and we can adjust how we're writing the docs. The only thing I would note is that our pages are really meant to just focus on the technical "How do I make this work" question. The scientific questions of what is this scenario and why are we running it are best answered by looking at papers and out of scope here.

For the downloaded data from input4mips the file metatdata do not make it clear what the scenario's are for.

Yes this is a fair point, although it is somewhat by design. These forcing files get used for a large number of experiments (yes, the pure scenario experiments, but also AerChemMIP variations, DAMIP experiments etc. etc.). Many of these actually aren't in our control of so we don't actually know what experiments the files are used for. There is discussion about how to document this, but we don't have a production solution yet WCRP-CMIP/cmip7-guidance#4

This is in contrast to what was done for CMIP6 ... but then the "ssp...." naming convention made it very clear!

I'm glad you felt this was clear, that isn't the feedback that has come from all areas. All the files do have the target_mip="ScenarioMIP" attribute which is intended to point users in the direction of the 'key' MIP for which the files have been produced (noting that they are also used in other MIPs, see above).

The use of single or two letters naming convention for CMIP7 seems a strange choice

Probably the best place for discussing this is in the ScenarioMIP paper revisions (although I suspect they are already closed). Rightly or wrongly, we just have to take the scenario names and acronyms we're given here.