EAMxx: Output P0 scalar variable alongside lev/ilev geometry data

[Copilot]

CF conventions require a scalar reference pressure variable (P0) to accompany hybrid sigma-pressure coordinate variables (lev/ilev). Its absence breaks downstream tools (e.g., ncclimo, vertical interpolation workflows) that work fine with EAM output.

Changes

• io_metadata.csv: Added P0 entry with long_name = "reference pressure".
• homme_grids_manager.cpp and mesh_free_grids_manager.cpp: add P0 as a 0-dim geo data field whenever lev/ilev geo data fields are added. Get value from physics::Constants<Real>. This ensures IO will automatically add it whenever the lev/ilev geo data fields are added to the output stream.

Luca: I edited this, as I changed the impl to a different strategy.

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds the CF-required scalar reference pressure variable (P0) alongside hybrid sigma-pressure vertical coordinate geometry (lev/ilev) so downstream tooling can correctly interpret hybrid coordinates.

Changes:

• Added P0 to io_metadata.csv with long_name = "reference pressure".
• Introduced FieldLayout::scalar() to represent 0-D (scalar) fields.
• Ensured P0 geometry data is created and populated from physics::Constants<Real>::P0.value when building/adding vertical coordinate geo data.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
components/eamxx/src/share/util/io_metadata/io_metadata.csv	Registers P0 metadata so IO can attach a descriptive long_name.
components/eamxx/src/share/field/field_layout.hpp	Adds a scalar (0-D) layout helper used to define P0 as a scalar field.
components/eamxx/src/share/data_managers/mesh_free_grids_manager.cpp	Creates and populates P0 in mesh-free grids to keep geo-data behavior consistent.
components/eamxx/src/dynamics/homme/homme_grids_manager.cpp	Creates and populates P0 in Homme physics grid and uses it to compute lev/ilev.

[bartgol]

@chengzhuzhang I don't know if you are familiar with the eamxx code, but I want to make sure that adding P0 is indeed enough to close the linked issue.

@AaronDonahue the PR adds P0 to the geo data of the grid, since we already have an automatic mechanism that writes geo data to file (including long name, units, etc).

[czender]

@bartgol I added myself as a reviewer because I do not see where in this PR the CF formula_terms attribute has been added to lev and ilev coordinates. Same for standard_name. Perhaps those are in a different PR? If so, my concerns have been addressed. If not, then this PR is insufficient to address the original problem. Writing those attributes is also necessary to resolve the problem.

[chengzhuzhang]

@bartgol thanks for getting to it so quickly. Sorry that I didn't make it clear that both formula_terms and standard_name attribute needs to be added, following the example from EAM. I just updated the issue description, and we may want to summon copilot for the additional work.

@czender thank you for catch this.

[bartgol]

Ah, thanks. I did not get that we needed the formula attribute as well. I can easily add it.

[bartgol]

Side q: none of our vars have the attribute positive = "down". I can add it to this PR if that's also needed.

[bartgol]

@czender @chengzhuzhang did I address all the concerns? Is there any other metadata you'd like me to add?

[czender]

@bartgol CF says "If the units attribute value is a valid pressure unit the default value of the positive attribute is down."
This means that positive = down is implicit for lev and ilev. It's your call whether to make that explicit by adding the attribute. In my opinion it would slightly improve the clarity of the EAMxx vertical coordinate. It's definitely a good idea to suggest adding this attribute to your MPAS Omega counterparts for their depth vertical coordinate.

[bartgol]

@bartgol CF says "If the units attribute value is a valid pressure unit the default value of the positive attribute is down." This means that positive = down is implicit for lev and ilev. It's your call whether to make that explicit by adding the attribute. In my opinion it would slightly improve the clarity of the EAMxx vertical coordinate. It's definitely a good idea to suggest adding this attribute to your MPAS Omega counterparts for their depth vertical coordinate.

Is the CF guidance to add this attribute only to the vert coord variable? Or to all variables with vertical levels?

Edit: nvm, the AI bot answered that for me

[bartgol]

Ok, I fixed the var name, and added the positive = down att too. Let me know if there's anything else we need to add.

[czender]

Looks good now. Thanks @bargol!

[chengzhuzhang]

The code change looks good to me! Thanks @bartgol .