Merge rrfs.v1.0.0 to develop by ShunLiu-NOAA · Pull Request #982 · NOAA-EMC/GSI

ShunLiu-NOAA · 2026-02-19T17:07:22Z

Description

The rrfsv1.0.0 branch was created in March 2024 when the RRFS code was frozen for pre-implementation testing. Since then, several updates have been applied to this branch:

Fixed a bug in read_radar. (#744)
Integrated GDIT's performance optimizations for RRFS. (#841)
Initialized variables in observer_fv3_reg.f90.
Resolved issues with CTest when running in debug [mode.] (#862)
Updated bufr_d to bufr_4 for compatibility with BUFR v12.1.0. (#963)

Type of change

Please delete options that are not relevant.

Bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to not work as expected)
This change requires a documentation update

How Has This Been Tested?
These changes were tested in Realtime RRFS workflow by EMC and NCO.

Checklist

My code follows the style guidelines of this project
I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
New and existing tests pass with my changes
Any dependent changes have been merged and published

This PR is to bring read_radar bug fix (NOAA-EMC#738) committed in the develop branch to RRFS release branch.

…age when compiling

NOAA-EMC#841) add GDIT's performance optimization for RRFS to release branch initialize variables in observer_fv3reg.f90 to avoid the warning message when compiling

…OAA-EMC#862) This is a draft PR to show fix for problems when running ctests using debug mode built GSI at rrfs.1.0.0.

This makes small changes needed for this branch to work with BUFR/12.2, which is the version being used for the RRFS implementation.

Add the GOES-19 satellite ID to the read-in list so that RRFSv1 can properly use abi_g19 data.

merge EMC repo/Release/rrfs.v1.0.0 to my fork

merging rrfs.v1.0.0 into test

ShunLiu-NOAA · 2026-02-20T04:58:07Z

ctest on wcoss2
Test project /lfs/h2/emc/lam/noscrub/emc.lam/Shun.Liu/gsi_regression/GSI/build
Start 1: global_4denvar
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 735.48 sec
2/6 Test #6: global_enkf ...................... Passed 979.93 sec
3/6 Test #2: rtma ............................. Passed 1156.43 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 1223.76 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 1404.88 sec
6/6 Test #1: global_4denvar ................... Passed 1925.21 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1925.22 sec

ShunLiu-NOAA · 2026-02-20T18:07:44Z

ctest on ursa:
Test project /scratch4/NCEPDEV/fv3-cam/Shun.Liu/GSI/build
Start 1: global_4denvar
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 487.77 sec
2/6 Test #2: rtma ............................. Passed 728.79 sec
3/6 Test #6: global_enkf ...................... Passed 740.77 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 793.08 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 852.80 sec
6/6 Test #1: global_4denvar ................... Passed 1230.19 sec

100% tests passed, 0 tests failed out of 6

HuiLiu-NOAA

Thanks.
I tested these modifications to the GSI develop/ branch in a retro DA test of conventional + AMV data, and the obs space diagnostics are identical to the RRFS release version/.

Copilot

Pull request overview

This PR merges the RRFS v1.0.0 release branch into develop, incorporating bug fixes, performance optimizations, and compatibility updates accumulated since the branch was frozen in March 2024. The changes include GDIT's performance optimizations using OpenMP parallelization and file-based caching, variable initialization fixes for debug mode, radar observation bug fix, and BUFR v12.1.0 compatibility updates.

Changes:

Performance optimizations with OpenMP parallelization and node-local file caching for grid coefficient computation
Variable initialization fixes to resolve debug mode CTest issues
Integration of BUFR v12.1.0 compatibility changes
Addition of timing diagnostics for ensemble member reading
New CRC32 module for portable node identification

Reviewed changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated 12 comments.

Show a summary per file

File	Description
src/gsi/read_iasi.f90	Initializes data_all array to zero
src/gsi/mod_fv3_lola.f90	Major performance optimization with OpenMP, file caching, and MPI node-aware broadcasting
src/gsi/hybrid_ensemble_isotropic.F90	OpenMP parallelization and file caching optimization for normalization factors
src/gsi/gsi_files.cmake	Adds crc32 source files to build
src/gsi/get_gefs_for_regional.f90	Adds MPI timing diagnostics
src/gsi/crc32_c.c	New C implementation of CRC32 using zlib
src/gsi/crc32.F90	New Fortran module wrapping CRC32 for node identification
src/gsi/cplr_get_fv3_regional_ensperts.f90	Adds timing diagnostics and OpenMP parallelization
src/gsi/combine_radobs.f90	Initializes data_all_in array to zero
src/gsi/anisofilter.f90	Moves operations inside mype==0 conditional to avoid unnecessary computation
src/gsi/CMakeLists.txt	Adds ZLIB dependency
src/enkf/loadbal.f90	Initializes arrays to zero with commented debugging MPI barrier
src/enkf/enkf.f90	Improves OpenMP thread safety using firstprivate to avoid false sharing
modulefiles/gsi_wcoss2.intel.lua	Updates module versions including BUFR and adds zlib

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

…sz' and deallocate this varaible

ShunLiu-NOAA · 2026-02-21T04:58:33Z

Reran CTest with the above modifications on wcoss2.
Test project /lfs/h2/emc/lam/noscrub/emc.lam/Shun.Liu/gsi_regression/GSI/build
Start 1: global_4denvar
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 733.71 sec
2/6 Test #6: global_enkf ...................... Passed 920.30 sec
3/6 Test #2: rtma ............................. Passed 1158.91 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 1284.53 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 1405.24 sec
6/6 Test #1: global_4denvar ................... Passed 1926.22 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1926.24 sec

ShunLiu-NOAA · 2026-02-23T01:20:16Z

(base) [Shun.Liu@ufe01 build] ctest -j 6
Test project /scratch4/NCEPDEV/fv3-cam/Shun.Liu/GSI/build
Start 1: global_4denvar
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 726.02 sec
2/6 Test #2: rtma ............................. Passed 793.51 sec
3/6 Test #5: hafs_3denvar_hybens .............. Passed 1031.38 sec
4/6 Test #6: global_enkf ...................... Passed 1042.00 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 1096.35 sec
6/6 Test #1: global_4denvar ................... Passed 1288.11 sec

100% tests passed, 0 tests failed out of 6

RussTreadon-NOAA · 2026-02-23T11:26:11Z

It seems the Copilot review caught some real issues which should be addressed. Other suggestions are for improving the robustness of the code - a good thing for operational codes. We need approvals from two reviewers not including members of the GSI Handling Review Committee, to move this PR forward. Hui approved. @TingLei-NOAA would you please review the changes in this PR?

TingLei-NOAA · 2026-02-23T13:26:59Z

I think @daniel.kokron (the developers of this important optimization session) could give us a quick clarification/explanation on this issue .
@daniel.kokron, would you please review those issues raised by Copilot and let us know what your opinions are? especially about this one #982 (comment). Thanks.

ShunLiu-NOAA · 2026-02-23T17:44:28Z

Test project /lfs/h2/emc/lam/noscrub/emc.lam/Shun.Liu/gsi_regression/GSI/build
Start 1: global_4denvar
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 740.43 sec
2/6 Test #6: global_enkf ...................... Passed 918.74 sec
3/6 Test #2: rtma ............................. Passed 1155.83 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 1223.00 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 1402.62 sec
6/6 Test #1: global_4denvar ................... Passed 1925.52 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1925.61 sec

RussTreadon-NOAA · 2026-02-23T17:50:03Z

Thank you @ShunLiu-NOAA for rerunning ctests. To save you time, ctests can wait until all conversations are resolved.

ShunLiu-NOAA · 2026-02-23T19:06:35Z

Thank you @ShunLiu-NOAA for rerunning ctests. To save you time, ctests can wait until all conversations are resolved.

Thank you, Russ. I just wanted to ensure that these minor changes I made have no impact on the results.

RussTreadon-NOAA · 2026-02-24T12:00:55Z

@ShunLiu-NOAA : We can move this PR forward once we get feedback from Dan regarding the two remaining open conversations. Would you please follow up with Dan?

TingLei-NOAA · 2026-02-24T19:23:08Z

 ! assume work load proportional to number of 'nearby' obs
 call estimate_work_enkf1(numobs) ! fill numobs array with number of obs per horiz point
 ! distribute the results of estimate_work to all processors.
+! added mpi_barrier for debug mode, on wcoss2.


@ShunLiu-NOAA Is this comment line still relevant? seems it should be removed to avoid confusing.

TingLei-NOAA

Now looks good to me!

RussTreadon-NOAA

Two peer reviews and approvals plus input from Dan Kokron.

Approve so @ShunLiu-NOAA can merge.

ShunLiu-NOAA · 2026-02-25T17:04:04Z

HERA ctest:
(base) [Shun.Liu@ufe01 build] ctest -j 6
Test project /scratch4/NCEPDEV/fv3-cam/Shun.Liu/GSI/build
Start 1: global_4denvar
Start 2: rtma
Start 3: rrfs_3denvar_rdasens
Start 4: hafs_4denvar_glbens
Start 5: hafs_3denvar_hybens
Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 487.56 sec
2/6 Test #2: rtma ............................. Passed 728.71 sec
3/6 Test #6: global_enkf ...................... Passed 741.85 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 792.25 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 911.57 sec
6/6 Test #1: global_4denvar ................... Passed 1290.51 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1290.52 sec

WCOSS2 ctest:
Test project /lfs/h2/emc/lam/noscrub/emc.lam/Shun.Liu/gsi_regression/GSI/build
Start 4: hafs_4denvar_glbens
Start 1: global_4denvar
Start 5: hafs_3denvar_hybens
Start 2: rtma
Start 6: global_enkf
Start 3: rrfs_3denvar_rdasens
1/6 Test #3: rrfs_3denvar_rdasens ............. Passed 732.93 sec
2/6 Test #6: global_enkf ...................... Passed 857.42 sec
3/6 Test #2: rtma ............................. Passed 1095.93 sec
4/6 Test #5: hafs_3denvar_hybens .............. Passed 1286.59 sec
5/6 Test #4: hafs_4denvar_glbens .............. Passed 1462.09 sec
6/6 Test #1: global_4denvar ................... Passed 1926.72 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1926.91 sec

RussTreadon-NOAA · 2026-03-03T16:12:53Z

@ShunLiu-NOAA : I think this PR closes some GSI issues.

I will close #981.

What about #860. Does this PR also close this issue?

Any other GSI issues we can close now that this PR has been merged?

ShunLiu-NOAA · 2026-03-03T16:48:48Z

@RussTreadon-NOAA We can also close issue #860 (#860), as it should be resolved by the rrfs release branch.

hu5970 and others added 18 commits May 7, 2024 11:25

Bring read_radar bug fix (PR 738) to RRFS release branch. (NOAA-EMC#744)

529b6ea

This PR is to bring read_radar bug fix (NOAA-EMC#738) committed in the develop branch to RRFS release branch.

add GDIT's performance optimizaiton for RRFS

d0facd7

add GDIT's performance optimization for RRFS

1ae752d

initialize variables in observer_fv3reg.f90 to avoid the warning mess…

0b21bad

…age when compiling

update github ci yamls

ffabb40

add gdit's modification

c814d94

use indent format with three spaces

5a9f28b

rrfs.v1.0.0 release branch update with GDIT's performance optimization (

0893efc

NOAA-EMC#841) add GDIT's performance optimization for RRFS to release branch initialize variables in observer_fv3reg.f90 to avoid the warning message when compiling

fix for problems in ctests using debug mode of GSI release/rrfs1.0.0 (N…

abf57f6

…OAA-EMC#862) This is a draft PR to show fix for problems when running ctests using debug mode built GSI at rrfs.1.0.0.

Update bufr_d to bufr_4 for bufr 12.1.0 version. (NOAA-EMC#941)

ed2b482

[release/rrfs.v1.0.0] Adds changes needed for BUFR/12.2 (NOAA-EMC#963)

fdfa520

This makes small changes needed for this branch to work with BUFR/12.2, which is the version being used for the RRFS implementation.

add sat ID of goes19 to read-in list (NOAA-EMC#967)

bcc8117

Add the GOES-19 satellite ID to the read-in list so that RRFSv1 can properly use abi_g19 data.

Merge branch 'rrfs.v1.0.0' into release/rrfs.v1.0.0

844f48e

Merge pull request #1 from NOAA-EMC/release/rrfs.v1.0.0

24539f5

merge EMC repo/Release/rrfs.v1.0.0 to my fork

Merge branch 'test' into rrfs.v1.0.0

982ad0f

Merge pull request #2 from ShunLiu-NOAA/rrfs.v1.0.0

0a76605

merging rrfs.v1.0.0 into test

revert pcgsoi to the version in develop branch

ba59e76

Merge branch 'NOAA-EMC:develop' into test

bba747d

ShunLiu-NOAA marked this pull request as ready for review February 20, 2026 18:07

ShunLiu-NOAA requested review from HuiLiu-NOAA and TingLei-NOAA February 20, 2026 18:08

RussTreadon-NOAA self-requested a review February 20, 2026 18:09

RussTreadon-NOAA assigned ShunLiu-NOAA Feb 20, 2026

HuiLiu-NOAA approved these changes Feb 20, 2026

View reviewed changes

RussTreadon-NOAA reviewed Feb 20, 2026

View reviewed changes

Comment thread modulefiles/gsi_wcoss2.intel.lua

RussTreadon-NOAA reviewed Feb 20, 2026

View reviewed changes

Comment thread modulefiles/gsi_wcoss2.intel.lua Outdated

RussTreadon-NOAA reviewed Feb 20, 2026

View reviewed changes

Comment thread src/enkf/loadbal.f90 Outdated

RussTreadon-NOAA reviewed Feb 20, 2026

View reviewed changes

Comment thread src/enkf/enkf.f90

remove the extra blank line

acd493d

RussTreadon-NOAA requested a review from Copilot February 20, 2026 18:53

Copilot started reviewing on behalf of RussTreadon-NOAA February 20, 2026 18:53 View session

Copilot AI reviewed Feb 20, 2026

View reviewed changes

remove commentout lines and fix the incorrect type for the variable '…

388a8bc

…sz' and deallocate this varaible

RussTreadon-NOAA reviewed Feb 20, 2026

View reviewed changes

Comment thread src/gsi/hybrid_ensemble_isotropic.F90 Outdated

ShunLiu-NOAA added 2 commits February 21, 2026 02:57

typo fix

8a80927

remove timers in the code

79f8af1

remove duplicate bufr_ver

40a0a01

after confirming from dkokron, the issues found by copilot are addressed

fbbe5f9

RussTreadon-NOAA self-requested a review February 24, 2026 16:00

TingLei-NOAA reviewed Feb 24, 2026

View reviewed changes

remove an unnecessary comment line.

a01d768

TingLei-NOAA approved these changes Feb 24, 2026

View reviewed changes

RussTreadon-NOAA approved these changes Feb 25, 2026

View reviewed changes

ShunLiu-NOAA merged commit b857ade into NOAA-EMC:develop Feb 25, 2026

RussTreadon-NOAA mentioned this pull request Feb 27, 2026

hafs_3denvar_hybens and hafs_4denvar_glbens seg fault after PR #982 #987

Closed

RussTreadon-NOAA mentioned this pull request Mar 3, 2026

Merge rrfsv1.0.0 to develop #981

Closed

RussTreadon-NOAA mentioned this pull request Mar 3, 2026

problems running debug mode GSI on wcoss2 for release/rrfs1.0.0 #860

Closed

Conversation

ShunLiu-NOAA commented Feb 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ShunLiu-NOAA commented Feb 20, 2026

Uh oh!

ShunLiu-NOAA commented Feb 20, 2026

Uh oh!

HuiLiu-NOAA left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

ShunLiu-NOAA commented Feb 21, 2026

Uh oh!

ShunLiu-NOAA commented Feb 23, 2026

Uh oh!

RussTreadon-NOAA commented Feb 23, 2026

Uh oh!

TingLei-NOAA commented Feb 23, 2026

Uh oh!

ShunLiu-NOAA commented Feb 23, 2026

Uh oh!

RussTreadon-NOAA commented Feb 23, 2026

Uh oh!

ShunLiu-NOAA commented Feb 23, 2026

Uh oh!

RussTreadon-NOAA commented Feb 24, 2026

Uh oh!

TingLei-NOAA Feb 24, 2026

Choose a reason for hiding this comment

Uh oh!

ShunLiu-NOAA Feb 24, 2026

Choose a reason for hiding this comment

Uh oh!

TingLei-NOAA left a comment

Choose a reason for hiding this comment

Uh oh!

RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

Uh oh!

ShunLiu-NOAA commented Feb 25, 2026

Uh oh!

RussTreadon-NOAA commented Mar 3, 2026

Uh oh!

ShunLiu-NOAA commented Mar 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

ShunLiu-NOAA commented Feb 19, 2026 •

edited

Loading