Reproducibility issue with MOM6 standalone between 0.3.0 and 0.4.0 builds

[minghangli-uni]

This is a separate issue from #266 to report repro inconsistencies in MOM6 standalone when comparing the 0.3.0 and 0.4.0 builds. The test case uses the current 025deg_jra55do_ryf configuration.

Below are the tests I’ve done. It seems that MOM6 from the latest 0.4.0 build does not reproduce the results from the 0.3.0 build.

— 0.3.0 build with default parameters -> porous true, RESOLN_N2_FILTER_DEPTH 2000m

  1   Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp,     Frac Mass Err,   Salin Err,    Temp Err
  2             [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]           [Nondim]        [PSU]        [degC]
  3      0,       0.000,     0, En 6.2897884199659999E-22, CFL  0.00000, SL  6.8277E-10, M 1.36283E+21, S 34.7205, T  3.62    37, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  4     64,       1.000,     0, En 1.1750935298777954E-03, CFL  0.14497, SL  3.1646E-06, M 1.36283E+21, S 34.7205, T  3.62    37, Me  8.71E-10, Se  2.59E-08, Te -1.53E-07

— 0.4.0 build with porous true, RESOLN_N2_FILTER_DEPTH 2000m, OM4_SUBCELLS related False

  1   Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp, Frac Mass Err,   Salin Err,    Temp Err
  2             [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]       [Nondim]        [PSU]        [degC]
  3      0,       0.000,     0, En 6.2897884199659999E-22, CFL  0.00000, SL  6.8277E-10, M 1.36283E+21, S 34.7205, T  3.6237, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  4     64,       1.000,     0, En 1.1753522694683549E-03, CFL  0.14498, SL  3.1848E-06, M 1.36283E+21, S 34.7205, T  3.6237, Me  8.77E-10, Se  2.60E-08, Te -1.55E-07

— 0.4.0 build with porous false, RESOLN_N2_FILTER_DEPTH -1, OM4_SUBCELLS related False

  1   Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp,     Frac Mass Err,   Salin Err,    Temp Err
  2             [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]           [Nondim]        [PSU]        [degC]
  3      0,       0.000,     0, En 6.2897884199659999E-22, CFL  0.00000, SL  6.8277E-10, M 1.36283E+21, S 34.7205, T  3.62    37, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  4     64,       1.000,     0, En 1.1750987892796693E-03, CFL  0.14497, SL  3.1687E-06, M 1.36283E+21, S 34.7205, T  3.62    37, Me  8.72E-10, Se  2.59E-08, Te -1.53E-07

— 0.4.0 build with default params -> porous false, RESOLN_N2_FILTER_DEPTH -1, OM4_SUBCELLS related True

  1   Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp,     Frac Mass Err,   Salin Err,    Temp Err
  2             [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]           [Nondim]        [PSU]        [degC]
  3      0,       0.000,     0, En 6.2897884199659999E-22, CFL  0.00000, SL  6.8277E-10, M 1.36283E+21, S 34.7205, T  3.62    37, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  4     64,       1.000,     0, En 1.1750931033948509E-03, CFL  0.14495, SL  3.1620E-06, M 1.36283E+21, S 34.7205, T  3.62    37, Me  8.70E-10, Se  2.58E-08, Te -1.53E-07

— 0.4.0 build with default params -> porous true, RESOLN_N2_FILTER_DEPTH 2000, OM4_SUBCELLS related True
  1   Step,       Day,  Truncs,      Energy/Mass,      Maximum CFL,  Mean Sea Level,  Total Mass,  Mean Salin, Mean Temp, F    rac Mass Err,   Salin Err,    Temp Err
  2             [days]                 [m2 s-2]           [Nondim]       [m]             [kg]         [PSU]      [degC]           [Nondim]        [PSU]        [degC]
  3      0,       0.000,     0, En 6.2897884199659999E-22, CFL  0.00000, SL  6.8277E-10, M 1.36283E+21, S 34.7205, T  3.623    7, Me  0.00E+00, Se  0.00E+00, Te  0.00E+00
  4     64,       1.000,     0, En 1.1753392797242994E-03, CFL  0.14501, SL  3.1829E-06, M 1.36283E+21, S 34.7205, T  3.623    7, Me  8.76E-10, Se  2.60E-08, Te -1.55E-07

[access-hive-bot]

This issue has been mentioned on ACCESS Hive Community Forum. There might be relevant details there:

https://forum.access-hive.org.au/t/cosima-twg-meeting-minutes-2025/4067/4

[dougiesquire]

One of the reasons for this was that our MOM6-standalone Cmake build system had a bug and many important compiler flags were not being set.

In addition, the following is the only parameter that needs to be added to MOM_input to preserve answers between the two version:

USE_HUYNH_STENCIL_BUG = True    !   [Boolean] default = False
                                ! If true, use a stencil width of 2 in PPM:H3 tracer advection. This is
                                ! incorrect and will produce regressions in certain configurations, but may be
                                ! required to reproduce results in legacy simulations.

I'm going to close this because with the above I can get the same answers from 0.3.0- and 0.4.0-era MOM6. Feel free to reopen if you don't think this is resolved