Is this going to be a run directory option in GCHP?

[lizziel]

Most of the contents of this repo are config files to run. Could this be one of the options when creating a run directory from model source code? If the files are separately maintained then they will quickly become outdated with new model versions. The data needed could be stored separately.

[lizziel]

Revisiting this, I noticed this repo is a submodule in .ci-pipelines in GCHP. The updates for 13.1.0-alpha.2 included a change in run directory (new file) and the CI test run failed since the file wasn't in this repo. It wasn't clear that run directory updates needed to be put in two different repos, geos-chem and here.

I think we should only include data files in this repo (and anything else static across versions), and move the config files to geos-chem with the other run config. We used to keep run config in a separate repo from GEOS-Chem and it was a big headache to keep up-to-date.

[LiamBindle]

I like the idea of transitioning this repo to sort of "standalone testing data" repository. I agree that it's messy to duplicate run directory updates here. Hopefully in the near future we can produce standalone test data for a full chemsitry simulation too.

I think once geoschem/geos-chem#459 is merged we can generate this run directory from the transport tracer simulation with the init_rd.sh script. It's capable of intializing settings at template-filling time, so I think that should make it even easier to maintain.

[lizziel]

Great, I like this plan. Thanks @LiamBindle .