Creating Vertical Diffusion Solver

[mwaxmonsky]

Originator(s): @mwaxmonsky

Summary (include the keyword ['closes', 'fixes', 'resolves'] and issue number):

Describe any changes made to the namelist: N\A

List all files eliminated and why: N\A

List all files added and what they do:
A to_be_ccppized/coords_1d.F90
A to_be_ccppized/linear_1d_operators.F90
A to_be_ccppized/vertical_diffusion_solver.F90

• Adds initial set of files to support solve step of vertical diffusion. Top level utilities files are copied directly from CAM and will be removed from CAM when integrating.

List all existing files that have been modified, and describe the changes: N\A
(Helpful git command: git diff --name-status development...<your_branch_name>)

List any test failures: None

Is this a science-changing update? New physics package, algorithm change, tuning changes, etc?
N\A

[jimmielin]

Thanks @mwaxmonsky! This looks good, I just had a few minor comments. The most major one is just to check the name of the to-be-ccpp(-)ized folder since it's not in ESCOMP/atmospheric_physics now.

[jimmielin]

Just confirming the directory is to be named to-be-ccpp-ized or to-be-ccppized (https://github.com/cacraigucar/atmospheric_physics/tree/atmos_phys_zmcleanup3/to_be_ccppized) since I've seen two variants around!

[mwaxmonsky]

Good catch @jimmielin!

@nusbaume, @cacraigucar, @peverwhee, we might have discussed this previously so my apologies for not taking better notes but do we have a preferred layout/naming convention for these non-ccppized files? Particularly with coords1d and linear_1d_operators being helper files (which might be needed by other files later) and the high level interface in the solver file.

I think having them at separate layers might be best for organization but not too sure on the naming conventions.

[nusbaume]

After discussing this with @jimmielin and @cacraigucar I believe we decided for the to_be_ccppized directory to be completely flat. This was to essentially make that directory ugly on purpose so that future users/developers remain motivated to get everything CCPP-ized. Of course if anyone disagrees with this plan please let me know!

@jimmielin @cacraigucar @peverwhee

[mwaxmonsky]

Flattened!

[jimmielin]

Thanks @mwaxmonsky for addressing my comments, I left the to-be-ccpp(-)ized and contiguous comments unresolved as there's some unfinished discussions on those. Once these are confirmed it should be good to go, I will approve the PR now.

[nusbaume]

Looks good! Just had some small code cleanup requests.

[nusbaume]

Everything looks great to me now. Thanks @mwaxmonsky!