Fix computation of angleEdge in NetCDF-C mesh converter #710
Merged
xylar merged 5 commits intoMPAS-Dev:masterfrom Apr 3, 2026
Merged
Fix computation of angleEdge in NetCDF-C mesh converter #710xylar merged 5 commits intoMPAS-Dev:masterfrom
angleEdge in NetCDF-C mesh converter #710xylar merged 5 commits intoMPAS-Dev:masterfrom
Conversation
Collaborator
Author
|
See E3SM-Project/polaris#479 for more discussion of this issue. |
xylar
force-pushed
the
fix-mesh-converter-angle-edge
branch
from
945708e to
6a007f9
Compare
xylar
changed the title
angleEdge in NetCDF-C mesh converter
Collaborator
Author
|
@cbegeman, no rush on this but have a look when you can. |
cbegeman
reviewed
Apr 1, 2026
| write_netcdf(dsMask, 'antarctic_mask.nc') | ||
|
|
||
|
|
||
| def test_conversion_angle_edge(): |
cbegeman
reviewed
Apr 1, 2026
cbegeman
approved these changes
Apr 2, 2026
Contributor
cbegeman
left a comment
cbegeman
left a comment
There was a problem hiding this comment.
Approving based on visual inspection. Did you test any meshes with finer resolutions than 1290km?
Collaborator
Author
|
No, I figured if it works for one mesh, it should work for all but happy to check something higher res before I merge. |
The old spherical `angleEdge` calculation in `MpasMeshConverter.x` inferred the edge angle from the latitude difference between edge vertices and then used a sign correction based on a nearby artificial "north pole" point. This approximation becomes inaccurate and unstable near the poles, where local east/north directions are poorly represented by latitude differences. This merge replaces that logic with a Cartesian geometric formulation: compute the edge normal from the vertex chord and edge position on the sphere, project it onto the local east/north basis at the edge, and compute `angleEdge` with `atan2`. This matches the downstream Polaris fix and gives robust spherical angles across the globe, including polar regions.
Add `mpas_tools.mesh.spherical` with utilities to compute local east/north vectors, edge-normal vectors, and recomputed spherical `angleEdge` values from mesh geometry. These routines mirror the corrected spherical formulation used for the C++ mesh converter and provide a reusable Python reference implementation.
Add a conversion test that runs `MpasMeshConverter.x` on the low-resolution QU test mesh, recomputes `angleEdge` in Python from spherical geometry, and verifies the converted mesh agrees to tight tolerance. This gives us a lightweight regression test for the polar `angleEdge` fix without introducing a dedicated C++ unit-test harness.
xylar
force-pushed
the
fix-mesh-converter-angle-edge
branch
from
6a007f9 to
9d2d739
Compare
Collaborator
Author
|
@cbegeman, I tested this in a similar way (C++ vs. python) on an EC30to60 mesh, again without any problems. I also added some fixes to the documentation I noticed when I was trying to install the conda package into my local pixi environment. |
Collaborator
Author
|
Just a note to say that I emailed @mgduda and he agreed that we could proceed here without his input. They use a different tool (not in MPAS-Tools) to compute |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The old spherical
angleEdgecalculation inMpasMeshConverter.xinferred theedge angle from the latitude difference between edge vertices and then used a
sign correction based on a nearby artificial "north pole" point. This
approximation becomes inaccurate and unstable near the poles, where local
east/north directions are poorly represented by latitude differences.
This merge replaces that logic with a Cartesian geometric formulation:
compute the edge normal from the vertex chord and edge position on the
sphere, project it onto the local east/north basis at the edge, and
compute
angleEdgewithatan2. This matches the downstream Polarisfix and gives robust spherical angles across the globe, including polar
regions.
This merge also adds
mpas_tools.mesh.sphericalwith utilities to compute local east/northvectors, edge-normal vectors, and recomputed spherical
angleEdgevalues frommesh geometry. These routines mirror the corrected spherical formulation used
for the C++ mesh converter and provide a reusable Python reference
implementation.
Finally, this merge adds a regression test to make sure a QU1920 mesh gets the same
angleEdgefrom the python and C++ codes.