Skip to content

Conversation

@dougiesquire
Copy link
Collaborator

Testing of mom-ocean/MOM6#1680 on dev-MC_100km_jra_ryf+wombatlite.

Companion to ACCESS-NRI/ACCESS-OM3#164

@dougiesquire dougiesquire added the mom6-pr-test Documenting tests of a PR to MOM6 upstream label Nov 28, 2025
@dougiesquire
Copy link
Collaborator Author

!test repro

@github-actions
Copy link

✅ The Bitwise Reproducibility Check Succeeded ✅

When comparing:

  • MOM6-PR-test/gfdl-to-main-2025-09-25/dev-MC_100km_jra_ryf+wombatlite (checksums created using commit b110b64), against
  • dev-MC_100km_jra_ryf+wombatlite (checksums in commit 6059ff1)
Further information

The experiment can be found on Gadi at /scratch/tm70/repro-ci/experiments/access-om3-configs/b110b6499fb94adc6d068c5ac0286758ad097eb9, and the test results at https://github.com/ACCESS-NRI/access-om3-configs/runs/56601824542.

The checksums generated by this !test command are found in the testing/checksum directory of https://github.com/ACCESS-NRI/access-om3-configs/actions/runs/19753700037/artifacts/4703324011.

The checksums compared against are found here https://github.com/ACCESS-NRI/access-om3-configs/tree/6059ff10d6aa0cfa30bdf73dabc2ff61e5d5c664/testing/checksum

Test summary:
test_repro_historical
test_repro_determinism
test_repro_restart

@dougiesquire
Copy link
Collaborator Author

Note, I set the following parameters to preserve answers, because some defaults have changed relative to the version of MOM6 currently used in ACCESS-OM3:

- MASS_WEIGHT_IN_PRESSURE_GRADIENT_TOP = False !   [Boolean] default = False
+ MASS_WEIGHT_IN_PRESSURE_GRADIENT_TOP = False !   [Boolean] default = True
                                ! If true and MASS_WEIGHT_IN_PRESSURE_GRADIENT is true, use mass weighting when
                                ! interpolating T/S for integrals near the top of the water column in FV
                                ! pressure gradient calculations.
- NDIFF_ANSWER_DATE = 20240101    ! default = 20240101
+ NDIFF_ANSWER_DATE = 20240101    ! default = 99991231
                                ! The vintage of the order of arithmetic to use for the neutral diffusion.
                                ! Values of 20240330 or below recover the answers from the original form of the
                                ! neutral diffusion code, while higher values use mathematically equivalent
                                ! expressions that recover rotational symmetry.
- WAVE_INTERFACE_ANSWER_DATE = 20221231 ! default = 20221231
+ WAVE_INTERFACE_ANSWER_DATE = 20221231 ! default = 99991231
                                ! The vintage of the order of arithmetic and expressions in the surface wave
                                ! calculations.  Values below 20230101 recover the answers from the end of 2022,
                                ! while higher values use updated and more robust forms of the same expressions:
                                !    <  20230101 - Original answers for wave interface routines
                                !    >= 20230101 - More robust expressions for Update_Stokes_Drift
                                !    >= 20230102 - More robust expressions for get_StokesSL_LiFoxKemper
                                !    >= 20230103 - More robust expressions for ust_2_u10_coare3p5

However, when we actually bring these changes into ACCESS-OM3, we'll almost certainly want to accept the default changes (and corresponding answer changes).

@dougiesquire dougiesquire deleted the MOM6-PR-test/gfdl-to-main-2025-09-25/dev-MC_100km_jra_ryf+wombatlite branch November 28, 2025 05:31
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

mom6-pr-test Documenting tests of a PR to MOM6 upstream

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants