Coordinating development threads for next phase of control experiment

CMIP7 CMIP7 development
1

As discussed here at the ESM1.6 standup 20/05/2025, this thread is to discuss and plan the upcoming combining of ESM 1.6 dev work to continue the existing Jhan simulation.

Please edit the below table with the name of the proposed update (this post is a wiki so anyone can edit it), it’s priority, target date and related github issue.

When making a contribution to the spin-up, refer to this checklist

Proposed update Priority Coordinator Order GitHub Issue/PR Status Target Date (ready)
Ozone 0 Paul CMIP7 official forcing. Data ready, waiting on software update to allow download
N dep 0 Paul CMIP7 official forcing. Data ready, waiting on software update to allow download
Iron dep 0 Dougie Outside CMIP7 official forcings. For the ocean only. PR here
Cooling in spin-up - possibly tweak ocean albedo in UM Revisit after other priority 0 changes are done
More STASH changes for data request 0* Additional stash outputs for data request · Issue #164 · ACCESS-NRI/access-esm1.6-configs Nice to have for spinup but essential for publication
Any diag table changes needed for data request? 0* Nice to have for spinup but essential for publication
code change: radiation / bgc only experiments 0*
CABLE as a library Still worth investigating
Alternate vcmax feedback 1
Reset sea-level? 2
Canopy height 2 Test spatially and temporally varying canopy height · Issue #108 · ACCESS-NRI/access-esm1.6-configs
Root distribution 2 May be included with parameter changes for Australian pfts
EGNL productivity / seasonality 2
Mediterranean ocean bathymetry change? / additional mixing 0 Chris Change bathymetry to address Mediterraean salinity increase · Issue #127 · ACCESS-NRI/access-esm1.6-configs · GitHub Dropped
COMPLETED
Water balance fixes / adjust lprec 0 Spencer 1 A bit more work for a few days. But we’ll need to adjust lprec anyway. Done. In access-esm1p6/2025.09.002
Switch to oneAPI compiler + use recent dependencies (MPI, NetCDF) 0 Manodeep 1 Added the known working version for oneapi by manodeep · Pull Request #126 · ACCESS-NRI/ACCESS-ESM1.6 · GitHub PR here. Done. In access-esm1p6/2025.09.002
cleaning up the vegetation distribution (so land use change is working) 0 Alex 1b TBD: Produce Land Use ancillaries for the ACCESS-ESM1.6 piControl experiments · Issue #54 · ACCESS-NRI/CMIP7-Input Done. In access-esm1p6/2025.09.002
Australian PFTs 0 Alex 1b Add Australian PFTs · Issue #106 · ACCESS-NRI/access-esm1.6-configs.; https://github.com/ACCESS-NRI/UM7/issues/129. Summary of changes required here: Updates required for land, including Australian pfts · Issue #191 · ACCESS-NRI/access-esm1.6-configs · GitHub Done. In access-esm1p6/2025.09.002
Tweaks to WOMBATlite 1 Pearse Slight tweak to WOMBATlite default parameters for ESM1.6 · Issue #44 · ACCESS-NRI/GFDL-generic-tracers · GitHub No longer required
Atmosphere STASH changes 0 Pull requests · ACCESS-NRI/access-esm1.6-configs Added to branch 16/5/25
Wombat parameter change (alkalinity gradient) 0 Dev preindustrial+concentrations pjb update1 by pearseb · Pull Request #126 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 4/6/25
Wombat parameter change (disscal) 0 Update field_table by pearseb · Pull Request #132 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 11/6/25
GHG 0 Use CMIP7 GHG concentrations for PI control · Issue #110 · ACCESS-NRI/access-esm1.6-configs Added to branch 7/5/25
Volcanic 0 Use CMIP7 volcanic forcing for PI control · Issue #114 · ACCESS-NRI/access-esm1.6-configs Added to branch 14/5/25
Aerosols 0 Use CMIP7 aerosol emissions for PI control · Issue #123 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 6/6/25
Wombat diagnostics to be output for assessing model 1 Update diag_table_wombatlite by dougiesquire · Pull Request #107 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 5/5/25
Read number of phenology types from file Update phenology file by blimlim · Pull Request #133 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 11/6/25
Wombat parameter change (alkalinity gradient) 0 Dev preindustrial+concentrations pjb update1 by pearseb · Pull Request #126 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 4/6/25
Wombat parameter change (disscal) 0 Update field_table by pearseb · Pull Request #132 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 11/6/25
GHG 0 Use CMIP7 GHG concentrations for PI control · Issue #110 · ACCESS-NRI/access-esm1.6-configs Added to branch 7/5/25
Volcanic 0 Use CMIP7 volcanic forcing for PI control · Issue #114 · ACCESS-NRI/access-esm1.6-configs Added to branch 14/5/25
Aerosols 0 Use CMIP7 aerosol emissions for PI control · Issue #123 · ACCESS-NRI/access-esm1.6-configs · GitHub Added to branch 6/6/25
deep_cmt_opt Deep CMT option · Issue #121 · ACCESS-NRI/access-esm1.6-configs Done / no change
Build system changes (see below) 0 Dougie 1 Update mom5 SPR for new CMake build system by dougiesquire · Pull Request #238 · ACCESS-NRI/spack-packages · GitHub Done. August spinup
MOM grid format update (area changes) 0 Dougie 1 Update ocean grid inputs to mosaic format · Issue #31 · ACCESS-NRI/access-esm1.6-configs · GitHub Done. August spinup
CICE5 0 Anton 1 Create test config for cice5 integration · Issue #80 · ACCESS-NRI/access-esm1.6-configs · GitHub Done August spinup
Priority legend
0 Absolutely essential for spinup
0* Absolutely essential for production run
1 Would be great to have
2 In a perfect world

If the feature you wish to add doesn’t have a GitHub issue please create one and add it to the ESM1.6 project board. Instructions for doing this are here.

Got questions? Contentious things to discuss? Pop thoughts in this thread, below.

0* Ice_history.F90 Ice diagnostic edits for ice outputs in production runs parallel to STASHC and diag table changes.

2

Shall we include the switch to OneAPI in that table?

3

Yes please. And any other known/planned optimisation changes.

1 Like
4

I’ve tried to group the table by changes that are more on the technical or I/O side, those that have a larger science component and those that are just on forcing. My thinking was that there is more potential to group some of the technical changes together when updating the configuration but the science changes may need to happen more incrementally.

5

I suggest we use the first hour of our regular NESP project meeting next Wednesday (4/6, 10am) to discuss the ‘science changes’ in the table. The aim would be to prioritise which changes go in when, dependencies between changes etc. If you aren’t already on the meeting invitation but would be interested in attending, please let me know.

6

Revision: 6

Here are checklists to guide contributors on updating the spin-up for ESM1.6:

Worked example: contributor guidelines for ACCESS ESM1.6

If there is no issue for the proposed change, create an issue in access-esm1.6-configs to describe the change and any discussions/figures/justification/testing

Create config change as a new branch in access-esm1.6-configs where the branch is based on tip of dev-preindustrial+concentrations and named using an issue number , e.g. 99-dev-preindustrial+concentrations for issue 99.

Makes changes to your branch and then open Pull Request (see screenshots)
Note that changes to restart files may need to be specified for some config changes and this specifies the restart file used;

Screenshot 2025-08-05 at 10.35.07 am
Screenshot 2025-08-05 at 10.35.07 am3060×572 85.8 KB

Screenshot 2025-08-05 at 10.38.00 am
Screenshot 2025-08-05 at 10.38.00 am3054×582 86.1 KB

Request for review;

Merge your changes into dev-preindustrial+concentrations.

Worked example: How to update the spin up (inputs from reviewer, contributor and runner)

Reviewer/contributor decide if a tag is going to be needed on the merged version on the dev-preindustrial+concentrations branch (a tag is used when the spin up will be updated).

Ask runner to continue run based on tagged commit.

Update experiment tables on ACCESS-Hive forum.

Worked example: runner (@Jhan) guidelines for ACCESS ESM1.6

Pull or clone fresh from the tagged commit of GitHub - ACCESS-NRI/access-esm1.6-configs: Standard ACCESS-ESM1.6 configurations released and supported by ACCESS-NRI (should be on dev-preindustrial+concentrations branch);

Continue run deciding to either use an older restart (to provide some overlapping years), parallel runs or an abrupt change;

Update experiment tables on ACCESS-Hive forum;

Update coordination table on ACCESS-Hive forum.

Please reach out to @Jhan as the runner to coordinate getting changes added. And @anton and @spencerwong for any technical guidance on contributor steps. And @clairecarouge and @cbull are available to help too.

1 Like
7

Spin-up options with (late-September) consolidated code

This table describes alternate spin-up options for code base with ocean albedo scaling, CICE5, oneAPI etc, new vegetation distribution including Australian pfts, water balance fixes. Some forcings still to be updated. Some diagnostics still to be added.

Name Atmos restart Ocean restart Ocean BGC restart CICE restart Notes Branch name
1 land-update 40 April 500 April 500 2025-08-01 0.92 albedo - Jhan. Use year number from April run test-PIspinup-Sept-A
2 land-update 40 April where June started April where June started 2025-08-01 0.92 albedo - Jhan. Use year number from April test-PIspinup-Sept-B
3 0.90 albedo - Martin to test
4 Try a case with Mediterranean mixing?
8

Thanks for starting a table @RachelLaw, would it be better placed here do you think? (One can move posts by clicking on the wrench icon)

9

@cbull - I was thinking discussion of options works here while the other post maintains the summary of the ones that form our spin-up chain - a ‘master list’. This will inevitably get more complicated now we are proposing to pull restarts for the different components from different runs, including some that may not be on the master list yet.

1 Like
10

@anton - for filling out this spin-up table - what options do we have for CICE5 restarts and do we have to be careful about compatibility with other restarts?

11

The only cice5 restarts are

/g/data/vk83/prerelease/configurations/inputs/access-esm1p6/modern/pre-industrial/restart/2025.08.01

which is year 1843

and those from the August spin-up run (years 1843 to 20xx)

The retained cliamte memory in the sea ice is small - it will equalibriate with the ocean quickly. But the year in the cice restart needs to match the other components - it can be changed using nco

12

@RachelLaw Have we dropped the idea of starting from the beginning of the April run when the solar constant change but bring some of the BGC fields back to the start, are you baking in that first 500 years of excess cooling in the Southern Ocean in particular.