Parameter updates for improved stability in 25 km configuration #556

minghangli-uni · 2025-05-12T06:13:08Z

1. Summary:
Update MOM_input parameters to closely follow the latest settings from GFDL OM5, which have been shown to support stable long-term runs.

What has changed?

Most parameters in MOM_input
coupling timestep in nuopc.runseq
ocn_cpl_dt in nuopc.runconfig
All doc files in docs

Why was this done?
stable long-term control runs without truncation errors.

2. Issues Addressed:

Parameter updates for improved stability in 25 km configuration #555

3. Depedencies (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)?

Answers are not changed compared to the stable long run.

ocean.stats in this PR

  1      0,  700435.000,     0, En 1.1534026194849000E-03, CFL  0.14190, SL  2.8286E-09, M 1.38176E+21, S 34.7210, T  3.5677, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  2     96,  700436.000,     0, En 1.2393540284500390E-03, CFL  0.11140, SL  2.8307E-09, M 1.38176E+21, S 34.7210, T  3.5678, Me  1.89E-19, Se  1.80E-17, Te  1.42E-18
  3    192,  700437.000,     0, En 1.2722991874921168E-03, CFL  0.10365, SL  2.8329E-09, M 1.38176E+21, S 34.7210, T  3.5678, Me  8.66E-20, Se -2.36E-17, Te -1.09E-18

ocean.stats in the stable long run

  1      0,  700435.000,     0, En 1.1534026194849000E-03, CFL  0.14190, SL  2.8286E-09, M 1.38176E+21, S 34.7210, T  3.5677, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  2     96,  700436.000,     0, En 1.2393540284500390E-03, CFL  0.11140, SL  2.8307E-09, M 1.38176E+21, S 34.7210, T  3.5678, Me  1.89E-19, Se  1.80E-17, Te  1.42E-18
  3    192,  700437.000,     0, En 1.2722991874921168E-03, CFL  0.10365, SL  2.8329E-09, M 1.38176E+21, S 34.7210, T  3.5678, Me  8.66E-20, Se -2.36E-17, Te -1.09E-18

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

minghangli-uni · 2025-05-15T01:51:16Z

It is confirmed that a test run using the MOM parameters in this PR reproduces the results of the long-running control simulation (upto 49 years)

minghangli-uni · 2025-05-15T01:55:26Z

The parameter differences compared to the latest GFDL OM5 settings are documented here, https://github.com/minghangli-uni/access-om3-doc/blob/main/tables/MOM_comparison/GFDL-OM5_ACCESS-OM3_25km_nml_diff.md

minghangli-uni · 2025-05-15T02:08:44Z

For the vertex shear model, the following three parameters were not included in this PR, as they are not yet available in the current source. They will be incorporated once the corresponding PR from GFDL is merged.

VERTEX_SHEAR_VISCOSITY_BUG = False !   [Boolean] default = True
                                ! If true, use a bug in vertex shear that zeros out viscosities at vertices on
                                ! coastlines.
VERTEX_SHEAR_GEOMETRIC_MEAN = True !   [Boolean] default = False
                                ! If true, use a geometric mean for moving diffusivity from vertices to tracer
                                ! points.  False uses algebraic mean.
VERTEX_SHEAR_GEOMETRIC_MEAN_KDMIN = 1.0E-09 !   [m2 s-1] default = 0.0
                                ! If using the geometric mean in vertex shear, use this minimum value for Kd.
                                ! This is an ad-hoc parameter, the diffusivities on the edge of shear regions
                                ! are sensitive to the choice.

minghangli-uni · 2025-05-15T02:11:52Z

We may consider turning off SIMPLE_TKE_TO_KD for testing purposes, but not as part of this PR, as the following warning is reported in access-om3.err,

WARNING from PE     0: set_diffusivity_init: SIMPLE_TKE_TO_KD can not be used reliably with USE_REGRIDDING.

aekiss · 2025-05-21T04:53:49Z

For reference: changes relative to current dev-MC_25km_jra_ryf:
https://github.com/COSIMA/access-om3-doc/blob/4c0b90f/tables/MOM_input/dev-MC_25km_jra_ryf-555-params-updates_nml_diff.md

aekiss · 2025-05-22T06:59:47Z

MOM_input

+MAX_P_SURF = 0.0                !   [Pa] default = -1.0
+                                ! The maximum surface pressure that can be exerted by the atmosphere and
+                                ! floating sea-ice or ice shelves. This is needed because the FMS coupling
+                                ! structure does not limit the water that can be frozen out of the ocean and the
+                                ! ice-ocean heat fluxes are treated explicitly.  No limit is applied if a
+                                ! negative value is used.
+USE_RIGID_SEA_ICE = True        !   [Boolean] default = False
+                                ! If true, sea-ice is rigid enough to exert a nonhydrostatic pressure that
+                                ! resist vertical motion.
+SEA_ICE_RIGID_MASS = 100.0      !   [kg m-2] default = 1000.0
+                                ! The mass of sea-ice per unit area at which the sea-ice starts to exhibit
+                                ! rigidity
+


Are these needed for coupled ice-ocean stability (damping gravity waves in ice-covered regions), or can they be deleted (together with DYNAMIC_SURFACE_PRESSURE above)? Maybe GFDL need this for stability because they're not coupling OM5 every timestep...?

Suggested change

MAX_P_SURF = 0.0 ! [Pa] default = -1.0

! The maximum surface pressure that can be exerted by the atmosphere and

! floating sea-ice or ice shelves. This is needed because the FMS coupling

! structure does not limit the water that can be frozen out of the ocean and the

! ice-ocean heat fluxes are treated explicitly. No limit is applied if a

! negative value is used.

USE_RIGID_SEA_ICE = True ! [Boolean] default = False

! If true, sea-ice is rigid enough to exert a nonhydrostatic pressure that

! resist vertical motion.

SEA_ICE_RIGID_MASS = 100.0 ! [kg m-2] default = 1000.0

! The mass of sea-ice per unit area at which the sea-ice starts to exhibit

! rigidity

aekiss · 2025-05-22T07:10:41Z

MOM_input

+DYNAMIC_SURFACE_PRESSURE = True !   [Boolean] default = False
+                                ! If true, add a dynamic pressure due to a viscous ice shelf, for instance.


Do we need this? It seems to be useful only for a dynamic ice shelf, which we don't have (yet).
I'm imagine "dynamic pressure" here is akin to that defined in Fig 1 here. I guess turning this off would also require USE_RIGID_SEA_ICE = False? See comment on lines +665 to +677 below.

Suggested change

DYNAMIC_SURFACE_PRESSURE = True ! [Boolean] default = False

! If true, add a dynamic pressure due to a viscous ice shelf, for instance.

@claireyung should the ice shelf be hydrostatic, rigid, or dynamically relax from rigid to hydrostatic over some timescale?

This parameter is for damping waves under ice shelves, even when a static ice shelf is used (it's in the barotropic solver part of the code). Of course it's not doing anything without an ice shelf but I will want it set to True when I add ice shelves. I haven't explored varying the lengthscales of this parameterisation.

OK, so assuming it has no stability or performance penalties we can just leave it on even in the configs without ice shelves.

aekiss · 2025-05-23T03:08:48Z

MOM_input

+                                ! scaled viscosity, the Smagorinsky and Leith viscosities, and AH.
+SMAGORINSKY_AH = True           !   [Boolean] default = False
+                                ! If true, use a biharmonic Smagorinsky nonlinear eddy viscosity.
+SMAG_BI_CONST = 0.06            !   [nondim] default = 0.0


access-om2 used k_smag_iso=2.0, but SMAG_BI_CONST is different notation specifying the same thing. k_smag_iso is C in Griffies & Hallberg 2000. Comparing MOM5 and MOM6 it seems that SMAGORINSKY_AH is equivalent (if we neglect grid anisotropy) to 0.125*(k_smag_iso/pi)**2, which equals 0.05066 in the access-om2 case with k_smag_iso=2.0. If this interpretation is correct, then our om3 viscosity is about the same (slightly larger) than in om2. So I'm happy with that.

aekiss · 2025-05-23T03:21:51Z

MOM_input

-                                ! The background biharmonic horizontal viscosity.
-LEITH_AH = True                 !   [Boolean] default = False
-                                ! If true, use a biharmonic Leith nonlinear eddy viscosity.
+AH_VEL_SCALE = 0.01             !   [m s-1] default = 0.0


I'm happy with this, but FYI ACCESS-OM2 used vel_micom_iso=0 for 0.25° and 0.1°, and 0.04 ms−1 for 1°

This may make OM3 0.25° more viscous than access-om2-025 in regions where this term dominates the viscosity (ie exceeds Smagorinsky).

aekiss · 2025-05-23T04:07:52Z

MOM_input

                                ! total tolerance for SSH is 4 times this value.  The default is
                                ! 0.5*NK*ANGSTROM, and this should not be set less than about
                                ! 10^-15*MAXIMUM_DEPTH.
-VELOCITY_TOLERANCE = 1.0E-04    !   [m s-1] default = 3.0E+08


Interesting that GFDL are only checking ETA_TOLERANCE (via default ETA_TOLERANCE = 1.0E-06), not VELOCITY_TOLERANCE. This seems potentially risky - is there any harm in retaining VELOCITY_TOLERANCE = 1.0E-04 ?

aekiss

Approving, on the assumption that these are an improvement over the previous parameters. But we should return to this and give a careful check that we're happy with these parameters.

minghangli-uni · 2025-06-04T06:31:17Z

!test repro commit

github-actions · 2025-06-04T06:37:26Z

❌ The Bitwise Reproducibility Check Failed ❌

When comparing:

555-params-updates (checksums created using commit f3b449b), against
dev-MC_25km_jra_ryf (checksums in commit 44d7453)

🔧 The new checksums will be committed to this PR, if they differ from what is on this branch.

Further information

The experiment can be found on Gadi at /scratch/tm70/repro-ci/experiments/access-om3-configs/f3b449b44696dbe79c84eaf73a0baea5fbe7c9de, and the test results at https://github.com/ACCESS-NRI/access-om3-configs/runs/43440599096.

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/15435266621/artifacts/3256357759.

The checksums compared against are found here https://github.com/ACCESS-NRI/access-om3-configs/tree/44d7453b92323f930131485cab8f43910f872c10/testing/checksum

github-actions · 2025-06-04T06:37:35Z

❌ !test Command Failed ❌ There was a problem committing the result of the reproducibility run. See https://github.com/ACCESS-NRI/access-om3-configs/actions/runs/15435266621

minghangli-uni · 2025-06-04T12:45:31Z

Here comes the error message

FATAL from PE 1299: ocean_model_restart was called with unused buoyancy fluxes. For conservation, the ocean restart files can only be created after the buoyancy forcing is applied.

This occurs during the restart write and is due to the current CI runlength is 3hrs, which is not an integer multiple of the tracer timestep (7200s). specifically 10800/7200=1.5, so the thermo cycle is only half way completed and its fluxes are still partially unapplied, hence MOM aborts this intentially to preserve conservation.

Changing the runlength to a multiple of DT_THERM, eg 2hrs, 4hrs resolves the issue.

minghangli-uni · 2025-06-05T02:31:57Z

!test repro commit

github-actions · 2025-06-05T02:41:57Z

❌ The Bitwise Reproducibility Check Failed ❌

When comparing:

555-params-updates (checksums created using commit f3b449b), against
dev-MC_25km_jra_ryf (checksums in commit 44d7453)

🔧 The new checksums will be committed to this PR, if they differ from what is on this branch.

Further information

The experiment can be found on Gadi at /scratch/tm70/repro-ci/experiments/access-om3-configs/f3b449b44696dbe79c84eaf73a0baea5fbe7c9de, and the test results at https://github.com/ACCESS-NRI/access-om3-configs/runs/43511491814.

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/15457063024/artifacts/3264267237.

The checksums compared against are found here https://github.com/ACCESS-NRI/access-om3-configs/tree/44d7453b92323f930131485cab8f43910f872c10/testing/checksum

minghangli-uni · 2025-06-05T02:54:53Z

Sorry @aekiss could you please review it again? I also ping @dougiesquire and @anton-seaice here if you are happy for a look?

aekiss

approved, assuming checksums are the only change since my last review

1. MOM_input: DT=900s, DT_THERM=7200s, DTBT=-0.9; 2. nuopc.runseq: 900s; 3. nuopc.runconfig: ocn_cpl_dt = 900.

Update MEKE module MOM_thickness_diffuse no GM parameterisation

Update horizontal mixing coefficients

Update background viscosity to 0

VELOCITY_TOLERANCE set to default

Disable Leith and enable biharmonic Smagorinsky eddy viscosity

Change to Bodner et al. 2023 formulation of the restratifying mixed-layer restratification parameterisation. module MOM_diabatic_driver Implicit energetics PBL scheme to determine diffusivity and viscosity in the PBL module MOM_CVMix_KPP Disabled module MOM_CVMix_conv Disabled module MOM_CVMix_shear Disabled module MOM_CVMix_ddiff Disabled module MOM_kappa_shear Jackson's shear-driven turbulence module MOM_energetic_PBL Enable ePBL and disable KPP

Add SIMPLE_TKE_TO_KD, but can not be used reliably with USE_REGRIDDING? Reduce BBL_EFFIC from default 0.2 to 0.01

Add background KD, KD_MIN, Henyey_igw_background latitude scaling for near surface background diffusivity Add double diffusion

Remove KHTR_MIN & Remove MAX_TR_DIFFUSION_CFL

Disable nuetral diffusion

Disable horizontal diffusion of tracers near boundaries

Reduce SEA_ICE_RIGID_MASS from 1000 to 100 Enable max limit of surface pressure Remove constant unresolved background gust for ustar

Disable surface pressure in EOS

Enable a dynamic pressure due to viscous ice shelves

…-configs/actions/runs/15457063024

minghangli-uni · 2025-06-05T04:48:20Z

My apologies @aekiss , I just made a small update to the comment for DTBT_RESET_PERIOD, changing it from 1.08E4 to 7200.0 to be consistent with *.short.

DTBT_RESET_PERIOD = 0.0 ! [s] default = 7200.0

minghangli-uni self-assigned this May 12, 2025

minghangli-uni force-pushed the 555-params-updates branch 3 times, most recently from bfc7fe0 to 9bb88ab Compare May 15, 2025 01:10

minghangli-uni marked this pull request as ready for review May 15, 2025 02:12

minghangli-uni requested a review from aekiss May 15, 2025 02:21

This was referenced May 15, 2025

Tidal dissipation parameterisation #562

Merged

CI repro check after #556 #569

Closed

aekiss added a commit to COSIMA/access-om3-doc that referenced this pull request May 21, 2025

add configs and tables for ACCESS-NRI/access-om3-configs#556

4c0b90f

minghangli-uni mentioned this pull request May 22, 2025

Restart repro check by each commit #570

Closed

aekiss reviewed May 22, 2025

View reviewed changes

aekiss reviewed May 23, 2025

View reviewed changes

aekiss previously approved these changes Jun 4, 2025

View reviewed changes

aekiss mentioned this pull request Jun 4, 2025

Parameter updates for improved stability in 25 km configuration #555

Closed

4 tasks

minghangli-uni force-pushed the 555-params-updates branch from 9bb88ab to f3b449b Compare June 4, 2025 06:27

dougiesquire mentioned this pull request Jun 4, 2025

Change default run length for ACCESS-OM3 ACCESS-NRI/model-config-tests#161

Closed

access-bot dismissed aekiss’s stale review via 548b621 June 5, 2025 02:42

chrisb13 requested review from claireyung and removed request for claireyung June 5, 2025 03:12

aekiss previously approved these changes Jun 5, 2025

View reviewed changes

minghangli-uni and others added 16 commits June 5, 2025 14:41

Timesteps,

03c2889

1. MOM_input: DT=900s, DT_THERM=7200s, DTBT=-0.9; 2. nuopc.runseq: 900s; 3. nuopc.runconfig: ocn_cpl_dt = 900.

module MOM_MEKE

d77a8b6

Update MEKE module MOM_thickness_diffuse no GM parameterisation

module MOM_lateral_mixing_coeffs

b63cb5f

Update horizontal mixing coefficients

module MOM_set_visc

88a7da6

Update background viscosity to 0

module MOM_continuity_PPM

d2b6119

VELOCITY_TOLERANCE set to default

module MOM_hor_visc

f3ffe73

Disable Leith and enable biharmonic Smagorinsky eddy viscosity

module MOM_set_diffusivity

fb9ecfc

Add SIMPLE_TKE_TO_KD, but can not be used reliably with USE_REGRIDDING? Reduce BBL_EFFIC from default 0.2 to 0.01

module MOM_bkgnd_mixing

900317d

Add background KD, KD_MIN, Henyey_igw_background latitude scaling for near surface background diffusivity Add double diffusion

module MOM_tracer_hor_diff

6968f60

Remove KHTR_MIN & Remove MAX_TR_DIFFUSION_CFL

module MOM_neutral_diffusion

8b5cfb4

Disable nuetral diffusion

module MOM_hor_bnd_diffusion

7dded09

Disable horizontal diffusion of tracers near boundaries

module MOM_surface_forcing_nuopc

1cec9fd

Reduce SEA_ICE_RIGID_MASS from 1000 to 100 Enable max limit of surface pressure Remove constant unresolved background gust for ustar

module MOM

b037c41

Disable surface pressure in EOS

module MOM_barotropic

2cf1a02

Enable a dynamic pressure due to viscous ice shelves

Updated checksums as part of https://github.com/ACCESS-NRI/access-om3…

7addf36

…-configs/actions/runs/15457063024

minghangli-uni dismissed aekiss’s stale review via 7addf36 June 5, 2025 04:44

minghangli-uni force-pushed the 555-params-updates branch from 548b621 to 7addf36 Compare June 5, 2025 04:44

aekiss approved these changes Jun 5, 2025

View reviewed changes

minghangli-uni merged commit fafee19 into dev-MC_25km_jra_ryf Jun 5, 2025
11 checks passed

minghangli-uni deleted the 555-params-updates branch June 5, 2025 05:19

dougiesquire mentioned this pull request Jul 9, 2025

Surface pressure limitation in MOM #634

Closed

dougiesquire mentioned this pull request Sep 19, 2025

What should the 100km mom config inherit from 25km / OM4-05? #771

Closed

5 tasks

anton-seaice mentioned this pull request Dec 9, 2025

CICE thermo energy conservation error #518

Closed

		DYNAMIC_SURFACE_PRESSURE = True ! [Boolean] default = False
		! If true, add a dynamic pressure due to a viscous ice shelf, for instance.

Parameter updates for improved stability in 25 km configuration #556

Parameter updates for improved stability in 25 km configuration #556

Uh oh!

Conversation

minghangli-uni commented May 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

minghangli-uni commented May 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

minghangli-uni commented May 15, 2025

Uh oh!

minghangli-uni commented May 15, 2025

Uh oh!

minghangli-uni commented May 15, 2025

Uh oh!

aekiss commented May 21, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aekiss left a comment

Choose a reason for hiding this comment

Uh oh!

minghangli-uni commented Jun 4, 2025

Uh oh!

github-actions bot commented Jun 4, 2025

Uh oh!

github-actions bot commented Jun 4, 2025

Uh oh!

minghangli-uni commented Jun 4, 2025

Uh oh!

minghangli-uni commented Jun 5, 2025

Uh oh!

github-actions bot commented Jun 5, 2025

Uh oh!

minghangli-uni commented Jun 5, 2025

Uh oh!

aekiss left a comment

Choose a reason for hiding this comment

Uh oh!

minghangli-uni commented Jun 5, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Parameter updates for improved stability in 25 km configuration #556

Parameter updates for improved stability in 25 km configuration #556

minghangli-uni commented May 12, 2025 •

edited

Loading

minghangli-uni commented May 15, 2025 •

edited

Loading