Update M4AGO scheme to version v1.1.1 (equivalent to dev-1.1.1)

[jmaerz]

Hi @TomasTorsvik and @JorgSchwinger , with this PR, the M4AGO scheme tag is updated to an improved, updated M4AGO scheme version. For the updated version, see release notes: https://github.com/jmaerz/M4AGO-sinking-scheme/releases/tag/v1.1.0 and https://github.com/jmaerz/M4AGO-sinking-scheme/releases/tag/v1.1.1

Two notes:

• I have tested this only in the NorESM2.1 version thus far - I am not sure, if anything needs to be done wrt git fleximod
• I have kept the meson.build file backwards compatible to earlier M4AGO versions - not sure, if this should remain like this (it doesn't harm, but could be simplified with the updated version).

Now steps taken to test with the NorESM development tag noresm2_5_alpha08d (for documenting the required steps to test branches)

git clone [email protected]:NorESMhub/NorESM.git
cd NorESM
git checkout noresm2_5_alpha08d
gvim .gitmodules

In .gitmodules, replaced

[submodule "blom"]
path = components/blom
url = https://github.com/NorESMhub/BLOM.git
fxtag = dev1.6.1.8
fxrequired = ToplevelRequired
fxDONOTUSEurl = https://github.com/NorESMhub/BLOM.git

by (Note: new-m4ago is just a branch, not a tag)

[submodule "blom"]
path = components/blom
url = https://github.com/jmaerz/BLOM.git
fxtag = new-m4ago
fxrequired = ToplevelRequired
fxDONOTUSEurl = https://github.com/NorESMhub/BLOM.git

Then, to update the submodules:

./bin/git-fleximod update --optional

After this call, everything should be/was in place to build and run the model as usual. For help on git fleximod, see also NorESMhub/NorESM#599 and https://github.com/ESMCI/git-fleximod.

However, currently the model version noresm2_5_alpha08d fails, when running due to errors that seem not to be related to the updated scheme, but rather due to the ice mesh (see: NorESMhub/NorESM#626).

[TomasTorsvik]

@jmaerz - thanks, looks fine to me.
Betzy is not available, so I can't set up a NorESM2.5 build now (NorESM2.3 should be the same as NorESM2.1). Think it should be fine to merge this, and then deal with a potential problem if it should emerge with NorESM2.5. Or we could just wait a few days until betzy is availabe again.

[TomasTorsvik]

I think it's fine to have the meson.buid file backwards compatible at this stage.

[jmaerz]

Hi Tomas, as summarized in the PR description, I tried a run with the most recent development tag of NorESM. To me, it seemed as if nothing else needed to be changed with respect to git fleximod than just the .gitmodules file as described (which will be required to be changed anyway, when a new tag is introduced for BLOM). Hence, I would suggest to merge the stuff into master by tmw noon so that you can prepare the BLOM tag 1.7, if you agree.

[TomasTorsvik]

@jmaerz - Hi, I think it's fine to merge this. The bug in the noresm2_5_alpha08d run seems to be unrelated to this PR.