Lfric2lfric w2 by ukmo-juan-castillo · Pull Request #390 · MetOffice/lfric_apps

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) · 2026-03-23T09:19:47Z

PR Summary

Sci/Tech Reviewer: Ricky Wong (@mo-rickywong)
Code Reviewer: mo-marqh

This is relatively simple change to allow lfric2lfric to regrid W2 fields. The accepted procedure to deal with W2 is to convert them to a set of two W3 and one Wtheta fields, process these fields, and then regenerate back a W2 field with the resulting W3 and Wtheta fields.

This changes will make use of the existing subroutines converting W2 to W3,Wtheta fields, and back.

When we merged this code to the trunk, we discovered that the results were not bit-comparable. After some investigation, it was discovered that the subroutines converting from W2 to W3 and the other way round have implicit dependencies on base_mesh, which have to be removed. This will be done as part of this new PR.

The dependency is in the kernel compute_sample_u_ops_code, which is called by interp_w3wth_to_w2_alg.

closes lfric2lfric: processing W2 fields #60
fixes lfric2lfric: processing W2 fields #60
Is related to Lfric2lfric: interpolate W2 fields #238, Remove base_mesh dependencies in compute_sample_u_ops kernel lfric_core#312
linked to Remove base mesh dependency in compute sample kernel lfric_core#313

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 Apps rose-stem suite
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)

Test Suite Results - lfric_apps - lfric2lfric_W2/run1

Suite Information

Item	Value
Suite Name	lfric2lfric_W2/run1
Suite User	juan.m.castillo
Workflow Start	2026-04-21T09:10:54
Groups Run	all

Dependency	Reference	Main Like
casim	MetOffice/casim@2026.03.2	True
jules	MetOffice/jules@2026.03.2	True
lfric_apps	ukmo-juan-castillo/lfric_apps@lfric2lfric_W2	False
lfric_core	ukmo-juan-castillo/lfric_core@b4eab04	True
moci	MetOffice/moci@2026.03.2	True
SimSys_Scripts	MetOffice/SimSys_Scripts@4387949	True
socrates	MetOffice/socrates@2026.03.2	True
socrates-spectral	MetOffice/socrates-spectral@2026.03.2	True
ukca	MetOffice/ukca@2026.03.2	True

Task Information

✅ succeeded tasks - 1511

⌛ waiting tasks - 1

Task	State
housekeep_azspice	waiting

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

transformation kernels

DrTVockerodtMO

Thanks for the changes. One minor issue regarding consistent style with the rest of the adjoint tests.

DanStoneMO · 2026-03-31T12:30:32Z

I tried running this against LFRic-JEDI, and the JEDI interface build failed with:
Generation Error: error: expected '5' arguments for the call to kernel 'compute_sample_u_ops_kernel_type' from invoke 'invoke_5' in the algorithm layer but found '7'. Expected '5' standard arguments, '0' stencil arguments, '0' qr_arguments and '0' halo-depth arguments.

From: https://cylchub/services/cylc-review/view/daniel.stone?&suite=mob-develop-next-mobbalt-3&no_fuzzy_time=0&path=log/job/1/build_jelf__exab_gnu/01/job.out

Any ideas what could be causing this? I've not yet ruled out an error on my part, and I will update if that is the case.

DanStoneMO · 2026-03-31T13:12:24Z

It looks like the build fail is due to the addition of two new arguments (geometry, topology) to the compute_sample_u_ops_kernel_type calls in interpolation_alg_mod and adj_interpolation_alg_mod.

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) · 2026-03-31T13:38:52Z

It looks like the build fail is due to the addition of two new arguments (geometry, topology) to the compute_sample_u_ops_kernel_type calls in interpolation_alg_mod and adj_interpolation_alg_mod.

Exactly, that is why there is a lfric_core linked PR. This lfric_core PR is the one modifying the kernel that you observed failing.

DanStoneMO · 2026-03-31T13:59:15Z

Ah, I see the branch for that now. Is there a PR for that one yet?

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) · 2026-03-31T14:09:32Z

Ah, I see the branch for that now. Is there a PR for that one yet?

Yes, there is a link in the PR description: MetOffice/lfric_core#312. I added a link to the PR in the issue.

DanStoneMO

Alright given it another shot and JEDI is all good with this. No linked PR will be needed for lfric-jedi.

DrTVockerodtMO

Thanks for making the changes, code owner review passed.

Lottie Turner (mo-lottieturner) · 2026-04-02T16:17:57Z

+          if (fs == W2) then
+            field_src_ptr => u_in_w3_src
+            field_dst_ptr => u_in_w3_dst
+            call lfric2lfric_map_regrid(field_dst_ptr, field_src_ptr)
+            field_src_ptr => v_in_w3_src
+            field_dst_ptr => v_in_w3_dst
+            call lfric2lfric_map_regrid(field_dst_ptr, field_src_ptr)
+            field_src_ptr => w_in_wth_src
+            field_dst_ptr => w_in_wth_dst
+            call lfric2lfric_map_regrid(field_dst_ptr, field_src_ptr)


What's the advantage of setting pointers and then calling lfric2lfric_map_regrid rather than just doing something like:

call lfric2lfric_map_regrid(u_in_w3_dst, u_in_w3_src) call lfric2lfric_map_regrid(v_in_w3_dst, v_in_w3_src) call lfric2lfric_map_regrid(w_in_w3_dst, w_in_w3_src)

Lottie Turner (mo-lottieturner) · 2026-04-02T16:23:14Z

+            field_src_ptr => u_in_w3_src
+            field_dst_ptr => u_in_w3_dst
+            call lfric2lfric_oasis_regrid(modeldb, oasis_clock, &
+                                  field_dst_ptr, field_src_ptr)
+            field_src_ptr => v_in_w3_src
+            field_dst_ptr => v_in_w3_dst
+            call lfric2lfric_oasis_regrid(modeldb, oasis_clock, &
+                       field_dst_ptr, field_src_ptr)
+            field_src_ptr => w_in_wth_src
+            field_dst_ptr => w_in_wth_dst
+            call lfric2lfric_oasis_regrid(modeldb, oasis_clock, &
+                                  field_dst_ptr, field_src_ptr)


Same question here as above - what's the advantage of setting pointers here?

Same question here as above - what's the advantage of setting pointers here?

The reason is that, when I tried to remove the pointers, I obtained segmentation faults. I investigated further and it turns out that one only needs to set the algorithm parameters to 'inout' rather than just 'out'. This is now fixed and I send it for a further review. I run all tests successfully and updated the trac.log.

Lottie Turner (mo-lottieturner) · 2026-04-02T17:10:41Z

Hi Juan - the linked section in the PR summary is meant for linked PRs - i.e. PRs in other repositories that need to go on together, rather than related PRs or PRs that need to go on in sequence. Can you move the #238 mention to the "is related to" section, and clarify whether lfric_core#312 is linked, part of a chain, or just related?

Lottie Turner (mo-lottieturner)

All looks good, thanks for fixing the pointers :)

Ricky Wong (mo-rickywong)

Kgos have changed, most likely due to the change in the geometry/topolgy enumerations used. However, other than a change in KGOs, no evidence is posted on the correctness of the output, Please post plots of the relevant fields that have changed (original and after change would be best).

Other than that, the code appears to fall in line with the PR description

Ricky Wong (mo-rickywong) · 2026-04-20T17:45:18Z

+    class(field_parent_type),  pointer :: field     => null()
+    type(field_type),          pointer :: field_src => null()
+    type(field_type),          pointer :: field_dst => null()


Don't nullify on declaration, nullify variables after declaration section of code.

Ricky Wong (mo-rickywong) · 2026-04-20T17:47:38Z

+                          element_order_h, element_order_v, Wtheta)
+
+    ! Get the geometry and topology of the destination mesh
+    mesh_dst      => mesh_collection%get_mesh(trim(mesh_names(dst)))


Suggested change

mesh_dst => mesh_collection%get_mesh(trim(mesh_names(dst)))

mesh_dst => mesh_collection%get_mesh(trim(mesh_names(dst)))

Ricky Wong (mo-rickywong) · 2026-04-20T17:52:49Z

+      fs = field_src%which_function_space()
+      if (fs == W2) then


Suggested change

fs = field_src%which_function_space()

if (fs == W2) then

fs_id = field_src%which_function_space()

if (fs_id == W2) then

Use fs_id instead as fsis sometimes to refer to the function space object, strictly speaking W0,W1,W3 etc are just integers which identify a space configuration defined by the project.

Ricky Wong (mo-rickywong) · 2026-04-20T17:53:24Z

      select case (regrid_method)
        case (regrid_method_map)
-          call lfric2lfric_map_regrid(field_dst, field_src)
+          if (fs == W2) then


Suggested change

if (fs == W2) then

if (fs_id == W2) then

Ricky Wong (mo-rickywong) · 2026-04-20T17:54:20Z

+    type(field_type),          target  :: v_in_w3_src,  v_in_w3_dst
+    type(field_type),          target  :: w_in_wth_src, w_in_wth_dst
+
+    integer(kind=i_def)                :: fs


Suggested change

integer(kind=i_def) :: fs

integer(kind=i_def) :: fs_id

Ricky Wong (mo-rickywong) · 2026-04-20T17:56:22Z

 #ifdef MCT
-          call lfric2lfric_oasis_regrid(modeldb, oasis_clock, &
-                                        field_dst, field_src)
+          if (fs == W2) then


Suggested change

if (fs == W2) then

if (fs_id == W2) then

Ricky Wong (mo-rickywong) · 2026-04-20T17:57:02Z

      end select
+
+      ! Rebuild the W2 fields from a set of W3 and Wtheta fields
+      if (fs == W2) then


Suggested change

if (fs == W2) then

if (fs_id == W2) then

Ricky Wong (mo-rickywong) · 2026-04-20T18:00:02Z

+  !> @param[in]      geometry Mesh geometry
+  !> @param[in]      topology Mesh topology


Suggested change

!> @param[in] geometry Mesh geometry

!> @param[in] topology Mesh topology

!> @param[in] geometry Mesh geometry enumeration value

!> @param[in] topology Mesh topology enumeration value

Ricky Wong (mo-rickywong) · 2026-04-20T18:00:40Z

+  !> @param[in]      geometry Mesh geometry
+  !> @param[in]      topology Mesh topology


Suggested change

!> @param[in] geometry Mesh geometry

!> @param[in] topology Mesh topology

!> @param[in] geometry Mesh geometry enumeration value

!> @param[in] topology Mesh topology enumeration value

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) · 2026-04-21T08:32:23Z

All suggested changes have been addressed, I am running again all the tests just in case and will update the trac.log once it is finished. KGOs changes respect to the trunk because before this PR the W2 fields were not processed in lfric2lfric (they just write zeros to the field output), and now they are. I am afraid I do not have any script to plot W2 fields, so if you could provide them (it could take me a good while to write one myself), I would be happy to generate W2 plots.

Ricky Wong (mo-rickywong) · 2026-04-23T08:19:21Z

All suggested changes have been addressed, I am running again all the tests just in case and will update the trac.log once it is finished. KGOs changes respect to the trunk because before this PR the W2 fields were not processed in lfric2lfric (they just write zeros to the field output), and now they are. I am afraid I do not have any script to plot W2 fields, so if you could provide them (it could take me a good while to write one myself), I would be happy to generate W2 plots.

I do not have any scripts to plot output, which is why I have requested it on this PR

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) · 2026-04-23T09:26:09Z

I have asked around and nobody has successfully plotted the output W2 field, I will need some time to write my own plotting scripts.

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) and others added 18 commits January 30, 2026 16:17

Issue 192: update partitioning namelist in lfric2lfric

e7ca8f2

Revert previous change, intended for another branch

5d310ab

Issue 192: update partitioning namelist in lfric2lfric

1127c9b

Revert previous change, intended for another branch

7c61c2d

Merge branch 'main' into lfric2lfric_W2

aba975a

Issue 60: lfric2lfric interpolation of W2 fields

150e94e

Issue 60: update missing KGOs

7e1ef9c

Issue 60: update KGOs

5b22236

Issue 60: update KGO

0e7f71a

Issue 192: update partitioning namelist in lfric2lfric

780587d

Revert previous change, intended for another branch

55ebdd2

Merge remote-tracking branch 'refs/remotes/origin/main'

ec2a136

Merge branch 'main' into lfric2lfric_W2

8909db6

MetOffice#238 Update KGOs

2f5da8a

Issue 192: update partitioning namelist in lfric2lfric

2aa3304

Revert previous change, intended for another branch

d94a998

Merge branch 'main' into lfric2lfric_W2

771f692

Update KGOs

2686de2

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) requested a review from Lottie Turner (mo-lottieturner) as a code owner March 23, 2026 09:19

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) marked this pull request as draft March 23, 2026 09:19

github-actions Bot assigned Juan Manuel Castillo Sanchez (ukmo-juan-castillo) Mar 23, 2026

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) and others added 2 commits March 23, 2026 09:41

Merge branch 'MetOffice:main' into main

3e63dfa

Merge branch 'main' into lfric2lfric_W2

7028ee1

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) mentioned this pull request Mar 23, 2026

Remove base_mesh dependencies in compute_sample_u_ops kernel MetOffice/lfric_core#312

Open

Remove base_mesh geometry and topology dependencies in W3 to W2

81d8bb2

transformation kernels

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) mentioned this pull request Mar 24, 2026

Remove base mesh dependency in compute sample kernel MetOffice/lfric_core#313

Open

29 tasks

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) and others added 2 commits March 24, 2026 13:43

Update KGOs and point to latest version of lfric_core branch

33a4052

Merge branch 'MetOffice:main' into main

985ea8e

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) requested a review from Ricky Wong (mo-rickywong) March 24, 2026 15:55

Changes due to update of linked lfric_core branch

77e8913

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) added 3 commits March 30, 2026 11:09

Update KGOs

815e7d7

Remove unnecessary variables, point to the linked lfric_core branch

42d206d

Properly set the lfric2lfric geometry and update KGOs

317dd70

Ricky Wong (mo-rickywong) marked this pull request as ready for review March 31, 2026 09:20

Ricky Wong (mo-rickywong) requested review from a team, DrTVockerodtMO, Matt Shin (matthewrmshin), Steven Sandbach (ss421) and thomasmelvin as code owners March 31, 2026 09:20

DrTVockerodtMO requested changes Mar 31, 2026

View reviewed changes

Comment thread applications/adjoint_tests/source/algorithm/interpolation/adjt_interpolation_alg_mod.x90 Outdated

Changing the location of one import line after a reviewer's comment

0a70553

DanStoneMO approved these changes Mar 31, 2026

View reviewed changes

DrTVockerodtMO approved these changes Apr 1, 2026

View reviewed changes

Lottie Turner (mo-lottieturner) reviewed Apr 2, 2026

View reviewed changes

Remove unnecessary pointers

12dd1a0

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) requested a review from Lottie Turner (mo-lottieturner) April 8, 2026 08:53

Lottie Turner (mo-lottieturner) approved these changes Apr 14, 2026

View reviewed changes

Ricky Wong (mo-rickywong) requested changes Apr 20, 2026

View reviewed changes

Changes suggested by the reviewer

f51f629

Update link to lfric_core

bfcbed5

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) mentioned this pull request Apr 28, 2026

Lfric2lfric multi cpu test #392

Open

28 tasks

	mesh_dst => mesh_collection%get_mesh(trim(mesh_names(dst)))
	mesh_dst => mesh_collection%get_mesh(trim(mesh_names(dst)))

		!> @param[in] geometry Mesh geometry
		!> @param[in] topology Mesh topology

Conversation

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Code Quality Checklist

Testing

Test Suite Results - lfric_apps - lfric2lfric_W2/run1

Suite Information

Task Information

Security Considerations

Performance Impact

AI Assistance and Attribution

Documentation

PSyclone Approval

Sci/Tech Review

Code Review

Uh oh!

DrTVockerodtMO left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

DanStoneMO commented Mar 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

DanStoneMO commented Mar 31, 2026

Uh oh!

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

DanStoneMO commented Mar 31, 2026

Uh oh!

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

DanStoneMO left a comment

Choose a reason for hiding this comment

Uh oh!

DrTVockerodtMO 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!

Choose a reason for hiding this comment

Uh oh!

Lottie Turner (mo-lottieturner) commented Apr 2, 2026

Uh oh!

Lottie Turner (mo-lottieturner) left a comment

Choose a reason for hiding this comment

Uh oh!

Ricky Wong (mo-rickywong) left a comment • 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!

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!

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Apr 21, 2026

Uh oh!

Ricky Wong (mo-rickywong) commented Apr 23, 2026

Uh oh!

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Apr 23, 2026

Uh oh!

Reviewers

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 23, 2026 •

edited

Loading

DanStoneMO commented Mar 31, 2026 •

edited

Loading

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 31, 2026 •

edited

Loading

Juan Manuel Castillo Sanchez (ukmo-juan-castillo) commented Mar 31, 2026 •

edited

Loading

Ricky Wong (mo-rickywong) left a comment •

edited

Loading