Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Workaround issues with MPASv8 ncdata files for EarthWorks #21

Merged

Conversation

gdicker1
Copy link

This PR implements a workaround for run failures caused by files added in EarthWorksOrg/CAM #16 . The offending ncdata and bnd_topo files are commented out and the previously existing MPAS ncdata files are made to work with cam_dev physics by adding use_gw_front = .false. to the default namelists.

@gdicker1 gdicker1 added the workaround An issue has a workaround or a pull request implements a workaround (instead of a fix) label Jul 10, 2024
@gdicker1 gdicker1 self-assigned this Jul 10, 2024
@gdicker1 gdicker1 marked this pull request as draft July 10, 2024 16:37
@gdicker1
Copy link
Author

Addresses (but does not completely fix) EarthWorksOrg/EarthWorks #55.

For now, the files added in #16 will remain available in the EarthWorks input data on Derecho and these lines can be uncommented if anyone wishes to make cases that use them.

@sherimickelson
Copy link

Thanks, @gdicker1 ! Could you add more details about your statement "Addresses (but does not completely fix) EarthWorksOrg/EarthWorks#55."? Does it fail in a spot further down the call tree?

@gdicker1 gdicker1 changed the title Workaround issues with MPASv8 ncdata files Workaround issues with MPASv8 ncdata files for EarthWorks Jul 10, 2024
@gdicker1
Copy link
Author

gdicker1 commented Jul 10, 2024

@sherimickelson

Could you add more details about your statement "Addresses (but does not completely fix) EarthWorksOrg/EarthWorks#55."?

I might be splitting hairs. I believe that a fix would add files that work with cam_dev physics without any modifications to the namelist or other config variables in CAM. The workaround lets EarthWorks move forward by using the pre-existing files and turning off a feature of the cam_dev physics package.

Does it fail in a spot further down the call tree?

I'm still testing this before moving out of draft, apologies for adding reviewers too soon. I'll answer this once I'm sure.

@gdicker1
Copy link
Author

gdicker1 commented Jul 10, 2024

The force push fa6a729 to 62939a9 makes it so use_gw_front is turned on for cam_dev when the dycore isn't MPAS. Also updates the commit message to match.

@gdicker1
Copy link
Author

This change is ready to go!

I tried testing with F2000dev and CHAOS2000dev compsets. At first these failed with "NaNs in the w field" from MPAS and a similar NaN/Inf issue from CLUBB. Once I switched MPAS to the same version as in cam6_3_136 (a near v7.3 version of MPAS), 30 day runs of these compsets were successful.

@gdicker1 gdicker1 marked this pull request as ready for review July 12, 2024 16:58
Copy link

@sherimickelson sherimickelson left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, @gdicker1

This commit comments out the real-data initial condition files and
corresponding topography files that were added in EWOrg/CAM PR#16. These
files cause the chemistry solvers to fail to converge, the dynamics to
become unstable (NaNs in the w field), and the runs of these cases to
fail.

By turning off gravity wave frontogenesis the other, existing MPAS
real-data files can be used with the cam_dev physics suite. This commit
also modifies logic in build-namelist so that frontogenesis is only
turned on with cam_dev if the MPAS-dycore isn't being used.
@gdicker1
Copy link
Author

Force pushed (62939a9 to 2ccf604) to correct the commit message. GitHub was turning the text EWOrg/CAM #16 into a link to ESCOMP/CAM PR#16.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
workaround An issue has a workaround or a pull request implements a workaround (instead of a fix)
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants