cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 #1029

gdicker1 · 2024-05-13T15:17:41Z

The new initial condition files contain fields that are expected by routines in cam_dev (cam7) physics. The new topo files were regenerated due to changes in the NCAR/Topo software. These files are only used when phys="cam_dev" to ensure previous compsets and/or tests aren't affected.

These files were checked for NaNs and runs were attempted for the F2000dev compset before making this PR.

Fixes #995 and #1094

The new IC files contain fields that are expected by routines in cam_dev (cam7) physics. The new topo files were regenerated due to changes in the NCAR/Topo software. These files are only used when 'phys="cam_dev"' to ensure previous compsets and/or tests aren't affected.

gdicker1 · 2024-05-13T15:24:37Z

NOTE: This PR branch is based on cam6_3_148 for compatibility with EarthWorks. This can be updated at any time, in whatever way reviewers would prefer.

Remaining work:

- Monitor and update branch based on review
- Work with others to add these files to CESM input data server¹
- Update branch w/ cam_development work
1. - (If not already) update references of cam_dev to cam7 when cam6_4_001: Change name of physics package 'cam_dev' to 'cam7' #1028 is merged
- Run test suite and record results

Until the files are added, Derecho users can copy my test input data within "/glade/derecho/scratch/gdicker/2024Apr04_CAMMPASv8Grids/test_inputdata2". This directory contains symlinks to my files, you may need to run something like cp -LR ${test_data} ${dest} ↩

This commit addresses issues that were seen when using ncdata with longitudes in the [-pi, pi] range in cases with full physics. This would cause the atmosphere to become unphysical and runs to fail. This change is made just after the longitude values are read to ensure they are correct for the remainder of execution.

NOTE: this is before the 'aux_cam' tests have been run

Previous versions of the lonCell_arr adjustment used an explicit loop. Without this loop, the nCells variable and the routine to fetch it are no longer needed.

Makes code consistent with CAM coding standards: 'All variables of type real must have a specified kind, including literals.'

bld/namelist_files/namelist_defaults_cam.xml

This commit also extends the comment that was with this code block, includes an if condition so valid values aren't modified, uses kinds and constants (pi) that are already used in the new file, and updates the ChangeLog so the correct file is referenced.

This aligns the files with changes in cam6_4_001 (PR ESCOMP#1028).

gdicker1 · 2024-07-26T17:56:56Z

Update branch w/ cam_development work

(If not already) update references of cam_dev to cam7 when cam6_4_001: Change name of physics package 'cam_dev' to 'cam7' #1028 is merged

Run test suite and record results

At this point 3.1 has been addressed by b608af1. I plan to proceed with step 3 and 4 once I have a tag number or directed otherwise.

Conform to CAM Coding Standards: 'Use symbolic numerical comparison operators (e.g., ==, /=, <, >=) not old character versions (e.g., .eq.).'

bld/namelist_files/namelist_defaults_cam.xml

The correct file for these physics schemes must include topography.

Enable no-topography, idealized test cases with the mpas dycore when running 58L or 93L on 480, 120, 60, and 30km grids.

NOTE: these files also support running with analytic_ic.

Duplicates an entry to ensure that the following two files are used together: - mpasa120_L32_topo_coords_c240508.nc - mpasa120_gmted2010_modis_bedmachine_nc3000_Laplace0100_noleak_20240507.nc Matches the physics specification on new real-data ICs, so the correct *_20240507.nc bnd_topo files are used with them.

gdicker1 · 2024-08-15T20:29:03Z

@adamrher could you take a look at this and see if it matches the files you expect? I want to make sure before I try to get these added to the CESM data repo.

This PR now:

Replaces the L32 analytic_ic files with updated versions
- the *_topo_*.nc file here uses the new bnd_topo
Adds notopo files for product {mpasa480, 120, 60, 30} x {58L, 93L}
Adds new real-data IC files for the product {mpasa480, 120} x {58L, 93L}
- these use the new bnd_topo below
- generated using the MPAS init_atmosphere core to add CSFR initial conditions from 2000-01-01T00:00:00Z¹
Adds 2 new bnd_topo files for mpasa480 and mpasa120

I couldn't get ctsm to use the correct files for mpasa480 and generate from analytic_ic. ↩

gdicker1 · 2024-08-16T15:34:40Z

@adamrher

I also wanted to see what you thought of this point:

[the real data ICs were] generated using the MPAS init_atmosphere core to add CSFR initial conditions from 2000-01-01T00:00:00Z

I've been told that some tracers and other fields would be missing if I didn't create the file from a year-long run (as you suggested before).

If this is a problem, I think it would be easier to drop mpasa480 cami*.nc and bnd_topo files from this PR than for me to figure out the CLM files needed for mpasa480. Then I could do what's needed to generate the correct mpasa120 ncdata.

adamrher · 2024-08-16T15:47:58Z

As long as the CFSR ncdata files work for a real world case, then I say move forward with the CFSR file. If the model doesn't complain, and some tracers are missing, then the model will probably initialize them to zero which is fine by me.

What's the issue with mpasa480? It looks like there is CTSM support for F2000climo/dev, but not FHIST.

gdicker1 · 2024-08-16T15:58:30Z

Moving forward sounds good, these files can always be replaced if they cause a problem.

CTSM support for F2000climo/dev, but not FHIST

This was my problem, I was trying to use FLTHIST and didn't consider the F2000 compsets.

jtruesdal · 2024-08-16T16:07:43Z

I do not know what the analytic ic topo cases are for and would agree that they seem to be unneeded. The analytic no topo (just vertical coordinates) are used and they also provide a way to run a new grid or start the model using a standard atmosphere initial condition.

adamrher · 2024-08-16T16:16:24Z

and they also provide a way to run a new grid or start the model using a standard atmosphere initial condition.

Is this true for mpasa coords files? I had thought that the vertical coordinate is a function of ncol and therefore must be consistent with the horizontal grid you're running? If this is not the case, than we only need one analytic_ic ncdata file for each vertical grid, as they would work for any arbitrary mpasaXXX grid.

gdicker1 · 2024-08-16T16:55:09Z

the vertical coordinate is a function of ncol and therefore must be consistent with the horizontal grid you're running

This is my understanding with MPAS. We get horizontal grid info from meshes generated by code not in the main MPAS-Model repo. It's easy to add vertical levels to that. I don't think that it would be efficient to do the Voronoi tessellation online (especially for high resolution) to run MPAS from only a set of vertical levels.

adamrher · 2024-08-16T17:12:31Z

Dylan I think what @jtruesdal is referring to is that, for example with the se dycore, when you want to run a new grid and you don't have an inic file, you can add a new grid entry in ccs_config/ pointing to a new ESMF mesh file (meaning you're supplying the new grid, not generating it in a run), and then run with analytic_ic. For the se dycore (and all but mpasa, maybe), the ncdata file only needs to have vertical level information, and the analytic_ic's are generated in the horizontal using the supplied mesh file and analytic_ic code in src/dynamics/. That's why there are only three coordinate files in the analytic_ic section (L26, L30, L32); they work for any arbitrary horizontal grid/dycore.

For no_topo mpasa, I think the zgrid is identical for every grid column. But I wouldn't be surprised if mpasa requires analytic_ic to have a ncdata file on the horizontal grid, which is indicated by the comments in the mpasa analytic_ic section of namelists_defaults.xml.

jtruesdal · 2024-08-16T17:20:38Z

Is this true for mpasa coords files? I had thought that the vertical coordinate is a function of ncol and therefore must be consistent with the horizontal grid you're running? If this is not the case, than we only need one analytic_ic ncdata file for each vertical grid, as they would work for any arbitrary mpasaXXX grid.

The MPAS vertical coordinate is height based so does have a dependence on the height of each ncol. The mpass coords files are very useful for aquaplanet cases where there is no topo. As you mentioned then we only need one coord file for each supported vertical resolution. I've also generated quick initial condition files from the aquaplanet runs that I've also used for non-aquaplanet cases.

gdicker1 · 2024-08-16T17:42:11Z

Thanks for explaining, my understanding of CAM/CESM was off. It seems to me the reason is:

MPAS separates the logic for vertical levels into the init_atmosphere core (may be easy to adapt in the CAM interface)
As they are created by mpas2esmf now, ESMF and SCRIP files in CESM don't have the horizontal grid info MPAS-A would need. (I think non-trivial to large amount of work to address)

Remove old L32 real-data files and all physics specififiers from new real-data files. The old mpasa bnd_topo is similarly replaced. One entry with the old mpasa120 bnd_topo file remains, it is still used by the mpas120km.waccm_fulltopo_c220818.nc ncdata. Specifiers are used to match these entries.

gdicker1 · 2024-08-19T22:52:34Z

After pushing 939a92d, 9afb3cd, and 24012df. The remaining work seems to be:

Generate an entry to replace mpas120km.waccm_fulltopo_c220818.nc
Replace mpasa120, 70L waccm ncdata with new file and remove mpas_120_nc3000_Co060_Fi001_MulG_PF_Nsw042_c200921.nc bnd_topo entry
Push approved files to CESM data repository
Prepare to test, merge in cam_development
Do aux_cam testing

The mpasa120km.waccm_fulltopo_c220818.nc was the last file using mpas_120_nc3000_Co060_Fi001_MulG_PF_Nsw042_c200921.nc, allowing it to be removed.

gdicker1 · 2024-09-05T23:19:27Z

@adamrher @jtruesdal @cacraigucar, I'm looking for another round of reviews or comments on this. Once any changes are made and this PR approved, I will ask for help getting the files below added to CESM data.

Files to add to atm/cam/inic/mpas (12):

mpasa120_L58_notopo_coords_c240814.nc
mpasa120_L93_notopo_coords_c240814.nc
mpasa30_L58_notopo_coords_c240814.nc
mpasa30_L93_notopo_coords_c240814.nc
mpasa480_L58_notopo_coords_c240814.nc
mpasa480_L93_notopo_coords_c240814.nc
mpasa60_L58_notopo_coords_c240814.nc
mpasa60_L93_notopo_coords_c240814.nc
cami_01-01-2000_00Z_mpasa120_L93_CFSR_c240814.nc
cami_01-01-2000_00Z_mpasa120_L58_CFSR_c240814.nc
cami_01-01-2000_00Z_mpasa480_L58_CFSR_c240814.nc
cami_01-01-2000_00Z_mpasa480_L93_CFSR_c240814.nc

Files to add to atm/waccm/ic (1):

mpasa120_L70.waccm_topography_SC_c240904.nc

Accessible on Derecho, within sub-dirs of: "/glade/derecho/scratch/gdicker/2024Jun04_CAMMPASv8Grids_2/files_for_cesmdata3"

adamrher

Looks great Dylan, thanks for doing this!

ESCOMP#1095) These changes address run-time failures that were seen when using MPAS ncdata with longitudes in the [-pi, pi] range in cases with non-simple physics. The atmospheric state would become unphysical and fail.

Part of catching up to cam_development

Catch PR branch up to current head of cam_development

gdicker1 · 2024-09-11T00:16:22Z

~~Commit a0db53f is a convenience for me as I bring in these changes to EarthWorksOrg/CAM as well. EarthWorksOrg/CAM is currently based on cam6_4_030.~~

Disregard this comment, EarthWorksOrg/CAM is actually based on cam6_4_019

These files currently can't be added to the CESM data repository due to size. Interested users may find these files within the CESM inputdata directory on Derecho and/or GLADE.

…w-develop (PR #33) Add updated ncdata and bnd_topo files for MPASv8. This PR brings in a feature branch that was contributed to ESCOMP/CAM ESCOMP#1029 (tagged cam6_4_033) and adds a few extra EarthWorks-specific files on top.

gdicker1 self-assigned this May 13, 2024

Update ChangeLog after creating PR

ef45d6b

gdicker1 mentioned this pull request Jul 18, 2024

Wrap MPAS-A longitudes to [0, 2pi) range #1094

Closed

gdicker1 mentioned this pull request Jul 18, 2024

part of cam6_4_033: Wrap MPAS longitude values to [0,2pi) range #1095

Merged

gdicker1 added 3 commits July 18, 2024 12:22

Add initial ChangeLog after making PR

7edc1db

NOTE: this is before the 'aux_cam' tests have been run

Remove unused nCells from cam_mpas_read_static

f3dfa51

Previous versions of the lonCell_arr adjustment used an explicit loop. Without this loop, the nCells variable and the routine to fetch it are no longer needed.

Add kind to number literal in lonCell_arr adjustment

288f3ed

Makes code consistent with CAM coding standards: 'All variables of type real must have a specified kind, including literals.'

cacraigucar requested changes Jul 25, 2024

View reviewed changes

bld/namelist_files/namelist_defaults_cam.xml Outdated Show resolved Hide resolved

gdicker1 added 2 commits July 26, 2024 11:29

Rename cam_dev to cam7 for files added in this branch

b608af1

This aligns the files with changes in cam6_4_001 (PR ESCOMP#1028).

gdicker1 marked this pull request as ready for review July 26, 2024 17:57

gdicker1 added 2 commits July 26, 2024 16:12

Replace .ge. with >= in lonCell loop condition

005e7f4

Conform to CAM Coding Standards: 'Use symbolic numerical comparison operators (e.g., ==, /=, <, >=) not old character versions (e.g., .eq.).'

Update comment on lonCell wrap code

1325911

adamrher reviewed Jul 30, 2024

View reviewed changes

bld/namelist_files/namelist_defaults_cam.xml Show resolved Hide resolved

cacraigucar approved these changes Aug 12, 2024

View reviewed changes

cacraigucar changed the title ~~Add updated ncdata and bnd_topo files for MPASv8~~ part of cam6_4_021?: Add updated ncdata and bnd_topo files for MPASv8 Aug 13, 2024

cacraigucar changed the title ~~part of cam6_4_021?: Add updated ncdata and bnd_topo files for MPASv8~~ Add updated ncdata and bnd_topo files for MPASv8 Aug 14, 2024

gdicker1 added 6 commits August 15, 2024 13:55

Add mpasa120_32L ncdata for analytic-ic with cam6 and cam7 physics

a396571

The correct file for these physics schemes must include topography.

Replace mpas L32 analytic IC files with v8 files

82fbdd0

Add mpasa 58L and 93L files for analytic ICs

6a75648

Enable no-topography, idealized test cases with the mpas dycore when running 58L or 93L on 480, 120, 60, and 30km grids.

Add real-data files {mpasa480,mpasa120} x {58L,93L}

c823836

NOTE: these files also support running with analytic_ic.

Update ChangeLog after adding L58 and L93 files

c45fd68

gdicker1 requested a review from adamrher August 15, 2024 20:29

Actual changes for last commit...

df17112

gdicker1 added 3 commits August 19, 2024 15:49

Remove mpasa120_L32_topo ncdata for analytic_ic

939a92d

Update ChangeLog after replacing bnd_topo and L32 ncdata

24012df

gdicker1 added 2 commits September 5, 2024 17:06

New mpasa120_L70 ncdata for waccm, remove now unused mpas120 bnd_topo

773a6f6

The mpasa120km.waccm_fulltopo_c220818.nc was the last file using mpas_120_nc3000_Co060_Fi001_MulG_PF_Nsw042_c200921.nc, allowing it to be removed.

Update ChangeLog after replacing waccm ncdata

1fcdd78

adamrher approved these changes Sep 10, 2024

View reviewed changes

gdicker1 added 3 commits September 10, 2024 17:59

Merge tag 'cam6_4_030' into add_v8mpasfiles

a0db53f

Part of catching up to cam_development

Merge tag 'cam6_4_032' into add_v8mpasfiles

8306f7f

Catch PR branch up to current head of cam_development

gdicker1 changed the title ~~Add updated ncdata and bnd_topo files for MPASv8~~ cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 Sep 11, 2024

gdicker1 added enhancement New feature or request SIMA labels Sep 11, 2024

gdicker1 mentioned this pull request Sep 11, 2024

EW: Add updated ncdata and bnd_topo files for MPASv8 EarthWorksOrg/CAM#33

Merged

gdicker1 added 3 commits September 12, 2024 14:54

Use the correct name for the mpasa480 notopo file

b36ac16

Remove mpasa30 L58 and L93 ncdata files

123d0fc

These files currently can't be added to the CESM data repository due to size. Interested users may find these files within the CESM inputdata directory on Derecho and/or GLADE.

Update ChangeLog with Derecho and Izumi tests

3c63318

gdicker1 merged commit fa94177 into ESCOMP:cam_development Sep 13, 2024
2 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 #1029

cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 #1029

gdicker1 commented May 13, 2024 •

edited

Loading

gdicker1 commented May 13, 2024 •

edited

Loading

gdicker1 commented Jul 26, 2024 •

edited

Loading

gdicker1 commented Aug 15, 2024 •

edited

Loading

gdicker1 commented Aug 16, 2024

adamrher commented Aug 16, 2024 •

edited

Loading

gdicker1 commented Aug 16, 2024

jtruesdal commented Aug 16, 2024

adamrher commented Aug 16, 2024

gdicker1 commented Aug 16, 2024

adamrher commented Aug 16, 2024

jtruesdal commented Aug 16, 2024

gdicker1 commented Aug 16, 2024 •

edited

Loading

gdicker1 commented Aug 19, 2024 •

edited

Loading

gdicker1 commented Sep 5, 2024

adamrher left a comment

gdicker1 commented Sep 11, 2024 •

edited

Loading

cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 #1029

cam6_4_033: Add updated ncdata and bnd_topo files for MPASv8 #1029

Conversation

gdicker1 commented May 13, 2024 • edited Loading

gdicker1 commented May 13, 2024 • edited Loading

Footnotes

gdicker1 commented Jul 26, 2024 • edited Loading

gdicker1 commented Aug 15, 2024 • edited Loading

Footnotes

gdicker1 commented Aug 16, 2024

adamrher commented Aug 16, 2024 • edited Loading

gdicker1 commented Aug 16, 2024

jtruesdal commented Aug 16, 2024

adamrher commented Aug 16, 2024

gdicker1 commented Aug 16, 2024

adamrher commented Aug 16, 2024

jtruesdal commented Aug 16, 2024

gdicker1 commented Aug 16, 2024 • edited Loading

gdicker1 commented Aug 19, 2024 • edited Loading

gdicker1 commented Sep 5, 2024

adamrher left a comment

Choose a reason for hiding this comment

gdicker1 commented Sep 11, 2024 • edited Loading

gdicker1 commented May 13, 2024 •

edited

Loading

gdicker1 commented May 13, 2024 •

edited

Loading

gdicker1 commented Jul 26, 2024 •

edited

Loading

gdicker1 commented Aug 15, 2024 •

edited

Loading

adamrher commented Aug 16, 2024 •

edited

Loading

gdicker1 commented Aug 16, 2024 •

edited

Loading

gdicker1 commented Aug 19, 2024 •

edited

Loading

gdicker1 commented Sep 11, 2024 •

edited

Loading