Adding OM5 Baltic test cases

[breichl]

• Adds two test cases, one Baltic grid with OM5 settings, and second with OM5 non-Boussinesq settings.
• These test cases can be used to test several OM5 settings that are not exercised in the current MOM6 test suite.
• The examples use the OM4 Baltic test case and forcing for simplicity.

[adcroft]

Testing at https://gitlab.gfdl.noaa.gov/ogrp/mom6ci/MOM6-examples/-/pipelines/27141

[adcroft]

It appears the checked-in layout files are inconsistent? The ocean ran with 32 cores in the CI but the ice is recorded with only 10?

$ git diff
diff --git a/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout b/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
index 6a4062941..07d03ed49 100644
--- a/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
+++ b/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
@@ -36,13 +36,13 @@ MASKTABLE = "MOM_mask_table"    ! default = "MOM_mask_table"
                                 !  4,6
                                 !  1,2
                                 !  3,6
-NIPROC = 2                      !
+NIPROC = 4                      !
                                 ! The number of processors in the x-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-NJPROC = 5                      !
+NJPROC = 8                      !
                                 ! The number of processors in the y-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-LAYOUT = 2, 5                   !
+LAYOUT = 4, 8                   !
                                 ! The processor layout that was actually used.
 IO_LAYOUT = 1, 1                ! default = 1, 1
                                 ! The processor layout to be used, or 0,0 to automatically set the io_layout to
diff --git a/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout b/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
index 6a4062941..07d03ed49 100644
--- a/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
+++ b/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
@@ -36,13 +36,13 @@ MASKTABLE = "MOM_mask_table"    ! default = "MOM_mask_table"
                                 !  4,6
                                 !  1,2
                                 !  3,6
-NIPROC = 2                      !
+NIPROC = 4                      !
                                 ! The number of processors in the x-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-NJPROC = 5                      !
+NJPROC = 8                      !
                                 ! The number of processors in the y-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-LAYOUT = 2, 5                   !
+LAYOUT = 4, 8                   !
                                 ! The processor layout that was actually used.
 IO_LAYOUT = 1, 1                ! default = 1, 1
                                 ! The processor layout to be used, or 0,0 to automatically set the io_layout to

[breichl]

It appears the checked-in layout files are inconsistent? The ocean ran with 32 cores in the CI but the ice is recorded with only 10?

$ git diff
diff --git a/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout b/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
index 6a4062941..07d03ed49 100644
--- a/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
+++ b/ice_ocean_SIS2/Baltic_OM5_025/SIS_parameter_doc.layout
@@ -36,13 +36,13 @@ MASKTABLE = "MOM_mask_table"    ! default = "MOM_mask_table"
                                 !  4,6
                                 !  1,2
                                 !  3,6
-NIPROC = 2                      !
+NIPROC = 4                      !
                                 ! The number of processors in the x-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-NJPROC = 5                      !
+NJPROC = 8                      !
                                 ! The number of processors in the y-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-LAYOUT = 2, 5                   !
+LAYOUT = 4, 8                   !
                                 ! The processor layout that was actually used.
 IO_LAYOUT = 1, 1                ! default = 1, 1
                                 ! The processor layout to be used, or 0,0 to automatically set the io_layout to
diff --git a/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout b/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
index 6a4062941..07d03ed49 100644
--- a/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
+++ b/ice_ocean_SIS2/Baltic_OM5nb_025/SIS_parameter_doc.layout
@@ -36,13 +36,13 @@ MASKTABLE = "MOM_mask_table"    ! default = "MOM_mask_table"
                                 !  4,6
                                 !  1,2
                                 !  3,6
-NIPROC = 2                      !
+NIPROC = 4                      !
                                 ! The number of processors in the x-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-NJPROC = 5                      !
+NJPROC = 8                      !
                                 ! The number of processors in the y-direction. With STATIC_MEMORY_ this is set
                                 ! in SIS2_memory.h at compile time.
-LAYOUT = 2, 5                   !
+LAYOUT = 4, 8                   !
                                 ! The processor layout that was actually used.
 IO_LAYOUT = 1, 1                ! default = 1, 1
                                 ! The processor layout to be used, or 0,0 to automatically set the io_layout to

Yep, thanks. Should be fixed now.

[adcroft]

Now passes (https://gitlab.gfdl.noaa.gov/ogrp/mom6ci/MOM6-examples/-/pipelines/27144)

[marshallward]

FATAL_UNUSED_PARAMS = True needs to be False for restart testing, since runtimes are configured through input.nml but still get stamped with a DAYMAX.

Few options:

• Remove FATAL_UNUSED_PARAMS from the config
• Explicitly set it to false for restart tests
• Check and don't set DAYMAX if runtime is set in input.nml

[marshallward]

Let's do option 2, it changes the least amount of code.

I ended up going with Option 3.