Add new stretching transformation to mesh tools by cjohnson-pi · Pull Request #270 · MetOffice/lfric_core

cjohnson-pi · 2026-02-10T14:18:24Z

PR Summary

Sci/Tech Reviewer: James Kent (@jameskent-metoffice)
Code Reviewer: Ricky Wong (@mo-rickywong)

Documentation:
#270 (comment)

linked MetOffice/lfric_apps#344

Code Quality Checklist

I have performed a self-review of my own code
My code follows the project's style guidelines
Comments have been included that aid understanding and enhance the readability of the code
My changes generate no new warnings
All automated checks in the CI pipeline have completed successfully

Testing

I have tested this change locally, using the LFRic Core rose-stem suite
If required (e.g. API changes) I have also run the LFRic Apps test suite using this branch
If any tests fail (rose-stem or CI) the reason is understood and acceptable (e.g. kgo changes)
I have added tests to cover new functionality as appropriate (e.g. system tests, unit tests, etc.)
Any new tests have been assigned an appropriate amount of compute resource and have been allocated to an appropriate testing group (i.e. the developer tests are for jobs which use a small amount of compute resource and complete in a matter of minutes)

LFRic Apps Testing still to be completed.

trac.log

Test Suite Results - lfric_core - test_stretching/run11

Suite Information

Item	Value
Suite Name	test_stretching/run11
Suite User	christine.johnson
Workflow Start	2026-03-13T14:48:49
Groups Run	suite_default

Dependency	Reference	Main Like
lfric_core	cjohnson-pi/lfric_core@stretching	False
SimSys_Scripts	MetOffice/SimSys_Scripts@2025.12.1	True

Task Information

✅ succeeded tasks - 390

Security Considerations

I have reviewed my changes for potential security issues
Sensitive data is properly handled (if applicable)
Authentication and authorisation are properly implemented (if applicable)

Performance Impact

Performance of the code has been considered and, if applicable, suitable performance measurements have been conducted

AI Assistance and Attribution

Some of the content of this change has been produced with the assistance of Generative AI tool name (e.g., Met Office Github Copilot Enterprise, Github Copilot Personal, ChatGPT GPT-4, etc) and I have followed the Simulation Systems AI policy (including attribution labels)

Documentation

Where appropriate I have updated documentation related to this change and confirmed that it builds correctly

PSyclone Approval

If you have edited any PSyclone-related code (e.g. PSyKAl-lite, Kernel interface, optimisation scripts, LFRic data structure code) then please contact the TCD Team

Sci/Tech Review

I understand this area of code and the changes being added
The proposed changes correspond to the pull request description
Documentation is sufficient (do documentation papers need updating)
Sufficient testing has been completed

(Please alert the code reviewer via a tag when you have approved the SR)

Code Review

All dependencies have been resolved
Related Issues have been properly linked and addressed
CLA compliance has been confirmed
Code quality standards have been met
Tests are adequate and have passed
Documentation is complete and accurate
Security considerations have been addressed
Performance impact is acceptable

cjohnson-pi · 2026-02-18T16:07:08Z

cjohnson-pi · 2026-03-04T10:45:52Z

Document describing the new stretching function and its implementation stretch.pdf

cjohnson-pi · 2026-03-06T17:41:56Z

The coordinates and the cell spacing for the new method (with various polynomial powers) and the old 'inflation' method that is currently used in the UM (in black).

So the new method gives similar answers to the UM inflation method.

cjohnson-pi · 2026-03-09T09:28:58Z

The ratio of the cell spacing between neighbouring cells.

cjohnson-pi · 2026-03-16T09:30:51Z

A non-symmetrical mesh, where the centre of the high resolution region is at (30,10).

Relevant Configuration is:
edge_cells_x = 24
edge_cells_y = 24
domain_centre = 30.0, 10.0
cell_size_inner = 0.0135,0.0135
cell_size_outer = 0.036,0.036
n_cells_outer_nsew = 1, 8, 5, 1
n_cells_stretch_nsew = 1, 1, 1, 1
poly_power = 3

cjohnson-pi · 2026-03-16T11:08:43Z

Updated documentation (including non-symmetrical meshes) stretch.pdf

James Kent (jameskent-metoffice)

This PR adds a new stretching transform based on polynomial stretching. I've worked through the document and I'm happy that this stretching is correct, and it has been coded correctly. There are a number of places where the "&" need to be aligned, and there is one erroneous print statement. When these are corrected I am happy this science review is approved.

cjohnson-pi · 2026-03-19T10:19:48Z

Thank you for completing the science review. I have now completed the requested minor suggestions.

Ricky Wong (mo-rickywong) · 2026-06-15T15:45:11Z

Is the documentation, on the core repo. I suppose it should be for reference. Though do we have permission to do so?

Ricky Wong (mo-rickywong)

Quite a lot of comments to address, mostly stemming from aligning with existing mesh_tools code and conventions on ordering.

Implementation approach is not made clear and prominent, leading to confusion when interpreting the code. There is mention in the pdf, though the code should be able to be followed it's own without the stretch.pdf.

Mesh tools sphinx documentation has not been updated. The PR documentation paper (pdf) on the approach is useful information though doesn't necessarily translate to understanding the implemented code. At present there is no details on stretching in the mesh documentation so thats not a major issue, Though the generators namelist options and usage has been completed and should be updated with changes in this PR.

I've not commented fully on polynominal stretching mod as I suspect it will change as requested changes filter down.

There are changes to unrelated test cases, though there is no mention of these on the PR. have the changes been confirmed to be negligible? checksums don't give any indication of the size of the change.

Ricky Wong (mo-rickywong) · 2026-06-15T09:00:14Z

+  create_lbc_mesh  = .true.
+  lbc_rim_depth    = 6
+  lbc_parent_mesh  = 'primary'
+  stretch_function = 'uniform'


stretch_function doesn't seem appropriate as uniform is one function that applies to the whole domain, whereas polynominal and inflation have different functions in different regions. Suggest stretch_profile is a better fit.

Uniform, polynomial and inflation are all applied to the whole domain. I therefore suggest that keeping stretch_function is the appropriate name. Given input coordinates x, calculate new coordinates y=f(x) using the mathematical function f.

Ricky Wong (mo-rickywong) · 2026-06-15T09:06:22Z

Please do not change the example namelist except to add required entries. The example file is to try and demo a working config (with new functionality if required/possible). These changes look like they are left-over from development for this change.

Ricky Wong (mo-rickywong) · 2026-06-15T09:08:45Z

 sort-key=Panel-A10
 type=logical

+[namelist:planar_mesh=stretch_function]


As before, suggest alternative name &stretch_profile

Ricky Wong (mo-rickywong) · 2026-06-15T09:24:09Z

+    = interior, with a stretch region in between.
+    = Inflation: y_n = y_n-1 + dy * inflation ** stretch
+    = Polynomial: y = bx + ax^n where n is the power
+!kind=default


Suggested change

!kind=default

help=The function applied is:

= * "Uniform"

= Uniform resolution mesh

= y = ax

=

= * "Inflation"

= Low resolution outer and a high resolution

= interior, with a stretch region in between.

= y_n = y_n-1 + dy * inflation ** stretch

=

= * "Polynominal"

= Low resolution outer and a high resolution

= interior, with a stretch region in between.

= y = bx + ax^n

= where n is the power.

Format the text for better clarity, Minor duplication to keep it clean. Something like above, please check layout in Cylc-gui

Ricky Wong (mo-rickywong) · 2026-06-15T09:43:30Z

+help=[n_outer_n, n_outer_s, n_outer_e, n_outer_w]
+    =This is the number of cells in an outer region
+    =near to a domain edge for the North, South,
+    =East and West edges.


Suggested change

help=[n_outer_n, n_outer_s, n_outer_e, n_outer_w]

=This is the number of cells in an outer region

=near to a domain edge for the North, South,

=East and West edges.

help=[n_outer_w, n_outer_s, n_outer_e, n_outer_n]

=

=Number of cells in the outer region adjacent

=to domain boundary. Values are ordered for

=domain boundaries as West, South, East and

=North.

=

Instructure arrays index locations follow the reference element, for Faces/edges it goes [W,S,E,N] so this should follow accordingly. I accept there is a conflict in naming here with the original transform namelist options, though the ordering of arrays really shouldn't be encoded into variable names. please bear this in mind. Could possibly rename these later or remove the ordering in the name if the inflation profile is deprecated.

Surely the choice of index locations should be chosen to best help the user. The North and South boundaries will most likely have the same number of cells, as will the East and West. So grouping these together makes the most sense. And surely North always comes first. Furthermore, I think NSEW is the traditional ordering of compass directions. Therefore I strongly suggest we keep NSEW.

I have updated the rest of the text following your suggestion, but keeping the NSEW.

Ricky Wong (mo-rickywong) · 2026-06-15T20:37:17Z

+!> @param param_a   Parameter a
+!> @param param_b   Parameter b
+!> @param param_c   Parameter c


Doxygen comments add no extra information

Suggested change

!> @param param_a Parameter a

!> @param param_b Parameter b

!> @param param_c Parameter c

!> @param param_a Parameter a, for Inner region function

!> @param param_b Parameter b, for Stretch region function

!> @param param_c Parameter c, for Outer region function

I think this is clear from the definitions in @details

Ricky Wong (mo-rickywong) · 2026-06-15T21:00:05Z

+  ! This enables support meshes to be created with a variable
+  ! resolution stretching function.
+  if (configuration%namelist_exists('stretch_transform')) then


Still not sure this was the best way to do this, leaving it as it was there before, but may need to be restructured later on with some thought

Yes, I expect more restructuring can be done later.

Ricky Wong (mo-rickywong) · 2026-06-15T21:15:48Z

 sort-key=Panel-A04
 type=integer

+[namelist:stretch_transform=n_cells_outer_nsew]


Doesn't follow rest of infrastructure ordering [W,S,E,N] for edges/faces, [SW,SE,NE,NW] for nodes.

As explained in another comment, I would prefer to leave as is.

Ricky Wong (mo-rickywong) · 2026-06-15T21:20:14Z

 sort-key=Panel-A04
 type=integer

+[namelist:stretch_transform=n_cells_stretch_nsew]


Similar comments as earlier with respect to the ordering encoded in the variable name

I would prefer to keep the NSEW ordering

Ricky Wong (mo-rickywong) · 2026-06-15T21:29:42Z

+!> @brief Apply a polynomial stretching transformation to a given coordinate
+!> @details In inner y = b x, in stretch y = a (x - xi) ^n + b x
+!!          and in outer y = yo + c x
+!> @param param_a   Parameter a
+!> @param param_b   Parameter b
+!> @param param_c   Parameter c
+!> @param x_inner   Unit mesh coordinate betwen inner and stretch
+!> @param x_outer   Unit mesh coordinate between stretch and outer
+!> @param dx        Unit mesh cell size
+!> @param direction North-south or East-west


Dyoygen comments do not match, Similar comments to on doxygen comments in in polynominal parameters

cjohnson-pi · 2026-06-16T13:43:49Z

Is the documentation, on the core repo. I suppose it should be for reference. Though do we have permission to do so?

I'm not quite understanding this question. But regarding documentation, please let me know the location of any documentation that needs to be updated. I cannot find any Sphinx documentation. It may be that further documentation (in addition to the pdf that I have provided) needs to follow in a follow-on PR.

cjohnson-pi added 4 commits February 9, 2026 14:26

stretching

d41d463

example works

58fdb32

tidy

1d3aa30

rose

73a15d5

github-actions Bot added the cla-required The CLA has not yet been signed by the author of this PR - added by GA label Feb 10, 2026

github-actions Bot assigned cjohnson-pi Feb 10, 2026

cjohnson-pi added 9 commits February 13, 2026 12:20

Update polynomial calculation

0765224

update polynomial

0f37fbe

Correct for all powers

a32d1ba

tidy

77bd280

correct versions

41d57eb

use PR number

62467ad

correct rose meta

ba667ca

unit tests run

e6b680e

unit tests pass

c0d74c2

cjohnson-pi added the KGO This PR contains changes to KGO label Feb 18, 2026

whitespace

5e08d60

cjohnson-pi marked this pull request as ready for review March 4, 2026 10:46

cjohnson-pi requested review from a team and Ricky Wong (mo-rickywong) as code owners March 4, 2026 10:46

cjohnson-pi added this to the Summer 2026 milestone Mar 4, 2026

cjohnson-pi added 7 commits March 4, 2026 16:12

update domain_extents

ce41044

Merge remote-tracking branch 'origin/main' into stretching

12eeb80

minor improvements

32cb52d

kgo and whitespace

2cf1072

mesh kgo

c2549a2

Merge branch 'main' into stretching

75400a4

CLA

e53cb21

github-actions Bot added cla-signed The CLA has been signed as part of this PR - added by GA and removed cla-required The CLA has not yet been signed by the author of this PR - added by GA labels Mar 5, 2026

update versions

97d21eb

cjohnson-pi mentioned this pull request Mar 6, 2026

Mesh stretching - Linked PR for lfric core MetOffice/lfric_apps#344

Open

28 tasks

cjohnson-pi added the Linked Apps This PR is linked to a MetOffice/lfric_apps PR label Mar 6, 2026

cjohnson-pi requested a review from James Kent (jameskent-metoffice) March 6, 2026 11:55

cjohnson-pi added 2 commits March 13, 2026 11:39

non-symmetrical

38e2537

correct style, rose-meta and kgos

dfa0f47

James Kent (jameskent-metoffice) approved these changes Mar 18, 2026

View reviewed changes

cjohnson-pi added 3 commits March 19, 2026 09:44

science review response

7a7feee

More aligning

0cb4c83

correction

c14506e

Ricky Wong (mo-rickywong) requested changes Jun 15, 2026

View reviewed changes

-!kind=default
+help=The function applied is:
+    = * "Uniform"
+    =   Uniform resolution mesh
+    =        y = ax
+    =
+    = * "Inflation"
+    =   Low resolution outer and a high resolution
+    =   interior, with a stretch region in between.
+    =       y_n = y_n-1 + dy * inflation ** stretch
+    =
+    = * "Polynominal"
+    =    Low resolution outer and a high resolution
+    =    interior, with a stretch region in between.
+    =         y = bx + ax^n
+    =    where n is the power.

Conversation

cjohnson-pi commented Feb 10, 2026 • edited by jameskent-metoffice Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Code Quality Checklist

Testing

trac.log

Test Suite Results - lfric_core - test_stretching/run11

Suite Information

Task Information

Security Considerations

Performance Impact

AI Assistance and Attribution

Documentation

PSyclone Approval

Sci/Tech Review

Code Review

Uh oh!

cjohnson-pi commented Feb 18, 2026

Uh oh!

cjohnson-pi commented Mar 4, 2026

Uh oh!

cjohnson-pi commented Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cjohnson-pi commented Mar 9, 2026

Uh oh!

cjohnson-pi commented Mar 16, 2026

Uh oh!

cjohnson-pi commented Mar 16, 2026

Uh oh!

James Kent (jameskent-metoffice) left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cjohnson-pi commented Mar 19, 2026

Uh oh!

Ricky Wong (mo-rickywong) commented Jun 15, 2026

Uh oh!

Ricky Wong (mo-rickywong) left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Ricky Wong (mo-rickywong) Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

cjohnson-pi Jun 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

cjohnson-pi Jun 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

cjohnson-pi commented Feb 10, 2026 •

edited by jameskent-metoffice

Loading

cjohnson-pi commented Mar 6, 2026 •

edited

Loading

Ricky Wong (mo-rickywong) Jun 15, 2026 •

edited

Loading

cjohnson-pi Jun 16, 2026 •

edited

Loading

cjohnson-pi Jun 16, 2026 •

edited

Loading