Skip to content

Conversation

@TomMelt
Copy link

@TomMelt TomMelt commented Jul 30, 2025

work in progress

This PR adds the UNet model into CAM and modifies the setup so that the single column, or UNet can be selected via appropriate namelist configuration (See README.md for more info)

TODO:

  • add routine to gather chunks
  • add routine to remap from phys grid to lonlat grid
  • gather data onto a single process
  • run through UNet
  • redistribute back to all procs
  • redistribute chunks
  • Compute tendencies
  • Use tendencies to update

This branch is based on Francis Vitt's work here

@TomMelt TomMelt self-assigned this Jul 30, 2025
! Create an ESMF_mesh for the physics decomposition
physics_grid_mesh = ESMF_MeshCreate(trim(grid_file), ESMF_FILEFORMAT_ESMFMESH, &
elementDistgrid=dist_grid_2d, rc=rc)
call check_esmf_error(rc, subname//'ESMF_MeshCreate')
Copy link
Collaborator

@ma595 ma595 Oct 31, 2025

Choose a reason for hiding this comment

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

For future README updates - this error is thrown as grid_file is unset.

As Tom suggests, the following is required in the user_nl_cam file.

! for regridding
cam_physics_mesh = '$ATM_DOMAIN_MESH'

Base automatically changed from nonlocal-gws to datawave_ml November 7, 2025 16:22
do col = 1, ncol
forcing(col,1) = -1*(flux(col,2) - flux(col,1))/(pmid(col,2) - pmid(col,1))
do level = 2, pver-1
forcing(col,level) = (flux(col,level+1) - flux(col,level-1)) / (pmid(col,level)*(log(pmid(col,level+1)) - log(pmid(col,level-1))))
Copy link
Collaborator

Choose a reason for hiding this comment

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

Looks like we might be missing a sign here @TomMelt. It's possible you haven't rebased off main?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants