claireyung:dev-MC_4km_jra_ryf+regionalpanan

[claireyung]

Adds regional panan on top of dev 25km branch post beta release

(see original PR here #689, which was squashed onto a test branch and those changes were cherrypicked onto a recent 25km dev branch. In fixing conflicts, I tried to make sure MOM_input didn't change from the 25km version (though it looks like I missed a few places), and that MOM_override does all the necessary overrides. But there may have been some answer changing from my test run in the process. https://github.com/claireyung/access-om3-configs/tree/8km_jra_ryf_obc2-sapphirerapid-Charrassin-newparams-rerun-Wright-spinup-accessom2IC-yr9 )

ping @chrisb13 @helenmacdonald

1. Summary:

What has changed? 25km dev branch was modified to make a 1/12th degree (4km average grid size in this domain) regional OM3 pan-Antarctic model configuration with no ice shelves, and open boudnary conditions forced by ACCESS-OM2-01 output. This involved parameter changes, generation of new input files documented in https://github.com/claireyung/mom6-panAn-iceshelf-tools and using a new prerelease version of ACCESS-OM3. It is the version described in https://access-om3-configs.access-hive.org.au/pr-preview-573/pages/IceShelves/.

Why was this done?

2. Issues Addressed:

3. Dependencies (e.g. on payu, model or om3-scripts)

This change requires changes to (note required version where true):

• payu:
• access-om3:
• om3-scripts:

4. Ad-hoc Testing

What ad-hoc testing was done? How are you convinced this change is correct (plots are good)?

5. CI Testing

• !test repro has been run

6. Reproducibility

Is this reproducible with the previous commit? (If not, why not?)

• Yes
• No - !test repro commit has been run.

7. Documentation

The docs folder has been updated with output from running the model?

• Yes
• N/A

A PR has been created for updating the documentation?

• Yes:
• N/A

8. Formatting

Changes to MOM_input have been copied from model output in docs/MOM_parameter_docs.short?

• Yes
• N/A

9. Merge Strategy

• Merge commit
• Rebase and merge
• Squash

[chrisb13]

Thanks @claireyung, great to see this coming together (again!)

@helenmacdonald is going to advise of the branch naming convention...

[claireyung]

Also for future reference, note that there are plenty of others who should be credited in the proposed alpha release e.g. who contributed to the making of the input files and parameter choices for this config, and discussions and help with the input file pipeline/config/bug fixes/optimisation. I hope I didn't forget anyone (sorry if I did) but please edit if I did. Tagging GH handles in a code block so not everyone gets spammed by future messages in this PR.

@angus-g @willaguiar @ezhilsabareesh8 @adele-morrison @helenmacdonald @edoyango @AndyHoggANU 
@dgwyther @fabiobdias @chrisb13 @anton-seaice @dougiesquire @minghangli-uni @aekiss @manodeep 

[anton-seaice]

For this whole section - by convention we leave the filenames the same across all configs, and show which config they are for by using a folder name (Not something @claireyung needs to solve, one of us can do it)

[anton-seaice]

Obviously in the final version, remove the changes to project and shortpath, and format in the same order as the other config.yaml files

[anton-seaice]

We should update script in om3-scripts and regenerate (Not something @claireyung needs to solve, one of us can do it)

[anton-seaice]

Suggested change

	ncat = 7
	ncat = 5

Whats the reason for this? Unless there a niche sea ice things of interest, or maybe you are trying to compare with SIS2 and it uses 7 categories?

[claireyung]

We were recommended this set of params for an Antarctic sea ice specific config by Stewart Allen, I have not done a comparison of whether it makes a difference though.

[chrisb13]

@anton-seaice, thanks for having a look.

@claireyung, just so you're updated.. The plan for this was discussed this morning and we should briefly check in about it. As such, if it's easier, we can wait till we've all met/had a look before you need to action anything.

[chrisb13]

(@anton-seaice, fyi, there is a draft version of some config docs that @claireyung has written here, rendered version.)

[anton-seaice]

@angus-g @willaguiar @ezhilsabareesh8 @adele-morrison @helenmacdonald @edoyango @AndyHoggANU
@dgwyther @fabiobdias @chrisb13 @anton-seaice @dougiesquire @minghangli-uni @aekiss @manodeep

You can add these people into CITATIONS.cff and then we won't forget

[aekiss]

Suggested change

	FATAL_UNUSED_PARAMS = False ! [Boolean] default = False
	FATAL_UNUSED_PARAMS = True ! [Boolean] default = False

best to have this true for safety?

[claireyung]

agree safety is a concern, but it might break the 'keep MOM_input up to date with standard OM3 config for ease of upstream improvements, and just use MOM_override for regional modifications' plan that I'm currently using. E.g. if MOM_override turns a module off e.g. MEKE, there might be original flags relating to that module (e.g. various MEKE choices) in MOM_input that become unused.
Not sure about the best approach.

[helenmacdonald]

Some notes and action items from the meeting this afternoon:

Action items

• Revert any mom_input changes to latest 25 km (but double check that this doesn’t introduce errors)

• Max_psurf will be better set to -1

• Double check that we have the info needed to recreate the input netcdf files. Initial conditions files doesn’t have any metadata (Claire will add this). A quick check suggests that the other files have good metadata.

• Config.yaml: This needs order + naming convention to be consistent with other om3 configs (not urgent). Likely also applies to ISF config. @chrisb13. Done by @dougiesquire in claireyung:dev-MC_4km_jra_ryf+regionalpanan. PR #3. #1077.

• Currently diag_table simlink is linking to a diag_table in a subdirectory (sourcefiles). We want to point to diag table that links in directory above (diagnostic profiler).

• Switch to the global released executable (when available in symmetric)

• A few other people (@minghangli-uni @dougiesquire ) should review

• Add list of people to CITATIONS.cff (see post here )

Issues for later

• On saphire rapid sometimes the model gets non reproducible seg faults during initialisation ( the issue goes away if you try again a few time). It was also an issue for previous COSIMA panantactic. Update: we think this has now been fixed -- so we should apply it if it hasn't already, same for isf config. @chrisb13 / @helenmacdonald. UPDATE FROM @dougiesquire: the ESMF update does not fix the issue, but I can avoid the errors with a few MPI flags that I will add to the config.

• Should we be using MOM_override or include all parameters in MOM_input? Probably a discussion to delay until later (but needs addressing for alpha release).

• After we move parameters to MOM_input then we need to re-add fatal unused parameters @helenmacdonald

• Consider removing expresssr queue for collation?
  , see issue Do we need expresssr queue for collation? #782. Update: check in with Claire. See related discussion. @chrisb13 Done by @dougiesquire in claireyung:dev-MC_4km_jra_ryf+regionalpanan. PR #3. #1077.

Notes
The simulation is a modification of old 25 km global, modified to be pan Antarctic and then updated recently to 25km. MOM_override contains all of the panantarctic specific parameter choices

The scripts for creating initial conditions are currently in notebooks in mom6-panan-inceshelf-tools repo. The metadata doesn’t currently give the githash.

Docs – these contain detailed instructions on how it was setup but currently just a dump so needs formatting/revision. Can deal with this later

This config is slightly different to the 10year run but has run for 1 month.

(edited to add further action items)

[claireyung]

The recent commit adds a couple of extra parameters to MOM_override that make this config more like the one I ran for 10 years. In particular, changes to the 25km model had an unwanted addition, turning neutral diffusion on in this config. I have now turned it off, so the MOM_parameter_doc.all is now pretty much identical to what I ran the 10 year spin up with (exceptions are updating the REGRIDDING_ANSWER_DATE where I had been using a 2018 value and updating ENTHALPY_FROM_COUPLER = False where before I had True (found a discussion here)

Also updated the metadata in the /g/data/tm70/cy8964/mom6/input/input-8km/ACCESS-OM2_IC_bfilled_smoothedland.nc file.

[chrisb13]

Thanks for the update @claireyung !

@angus-g @willaguiar @ezhilsabareesh8 @adele-morrison @helenmacdonald @edoyango @AndyHoggANU
@dgwyther @fabiobdias @chrisb13 @anton-seaice @dougiesquire @minghangli-uni @aekiss @manodeep

You can add these people into CITATIONS.cff and then we won't forget

On this point, can we please add this as an action item to your post @helenmacdonald?

@claireyung, here's an example if that's useful.

[helenmacdonald]

On this point, can we please add this as an action item to your post @helenmacdonald?

done

[dougiesquire]

Just noting that I'm holding off reviewing this until @anton-seaice's comments and the action items above have been addressed. Hopefully that's in line with what others were thinking/expecting.

[anton-seaice]

The Configuration stub 'dev-MC_25km_jra_ryf.md' in the target branch is causing the same page to be included twice in https://access-om3-configs.access-hive.org.au/ - can it be deleted ?

[chrisb13]

Closing this PR as it's being picked up again here: #1077

[dougiesquire]

(Reopened temporarily to give me permission to push the updated repro checksums to Claire's branch)