Add Changes Related to HAFS TDR SuperOb by JingCheng-NOAA · Pull Request #1007 · NOAA-EMC/GSI

JingCheng-NOAA · 2026-05-13T20:44:49Z

Description
The upcoming HAFSv2.2 will assimilate super-obbed TDR observations. To support this, the existing along-the-beam thinning function will be bypassed if the input TDR BUFR data is already super-obbed. This PR introduces a new namelist option, l_tdr_thin_alongbeam, allowing users to explicityly toggle along-the-beam thinning on or off.

Type of change

Please delete options that are not relevant.

Bug fix (non-breaking change which fixes an issue)
[x ] 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?

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

…sed in upcoming HAFSv2.2

JingCheng-NOAA · 2026-05-13T20:46:26Z

Regression Test performed on Hera:
1/6 Test #1: global_4denvar ................... Passed 1964.67 sec
2/6 Test #2: rtma ............................. Passed 1092.61 sec
3/6 Test #3: rrfs_3denvar_rdasens ............. Passed 670.17 sec
4/6 Test #4: hafs_4denvar_glbens .............. Passed 1155.10 sec
5/6 Test #5: hafs_3denvar_hybens .............. Passed 1156.24 sec
6/6 Test #6: global_enkf ...................... Passed 1159.08 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 7197.93 sec

ShunLiu-NOAA · 2026-05-19T13:25:34Z

@JingCheng-NOAA could you start a ctest on WCOSS?

JingCheng-NOAA · 2026-05-19T17:37:24Z

@JingCheng-NOAA could you start a ctest on WCOSS?

I tried many times on WCOSS2 but seems like I don't have right permission to all the files related to regression tests.
For all the 6 test cases, only hafs related ctest cases passed without issue.
This is what I found in one of the failed test cases (global_enkf) :

 running on           32  processors ...


* . * . * . * . * . * . * . * . * . * . * . * . * . * . * . * . * . * . * . * .
     PROGRAM ENKF_ANL HAS BEGUN. COMPILED 2011319.55     ORG: NP25
     STARTING DATE-TIME  MAY 19,2026  16:17:35.316  139  TUE   2461180


 number of satellite radiance files used          81
 number of satellite ozone files used          16
 error: vlocal_eig.dat does not exist
 ****STOP2****  ABORTING EXECUTION w/code=          19

ShunLiu-NOAA · 2026-05-21T13:00:19Z

@yonghuiweng or @XuLu-NOAA could you help run a ctest on WCOSS2?

RussTreadon-NOAA · 2026-05-21T15:42:35Z

WCOSS2 Ctests

Install JingCheng-NOAA:feature/hafs_tdr at e157e33 and develop at 55faa5e on Cactus. Run ctests with following results

Test project /lfs/h2/emc/da/noscrub/russ.treadon/git/gsi/pr1007/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  731.62 sec
2/6 Test #6: global_enkf ......................   Passed  860.05 sec
3/6 Test #2: rtma .............................   Passed  1155.54 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1223.75 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1343.08 sec
6/6 Test #1: global_4denvar ...................   Passed  1925.83 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1926.16 sec

All tests Passed on Cactus.

Question: Do either of the HAFS tests exercise the code modified by this PR? We need confirmation that the changes in the PR function as intended.

Recommendation: Echo the contents of namelist obs_input to stdout. Currently only select elements of namelist obs_input are written to stdout. Variable l_tdr_thin_alongbeam is not written to stdout. The GSI log file (e.g., stdout) should capture the full GSI configuration. This both documents how the GSI was configured as well as provides useful information in case we need to debug a GSI run.

XuLu-NOAA · 2026-05-21T15:45:47Z

Regression performed on WCOSS2:
/lfs/h2/emc/hur/noscrub/xu.lu/save/develop/build]$ ctest
Test project /lfs/h2/emc/hur/noscrub/xu.lu/save/develop/build
Start 1: global_4denvar
1/6 Test #1: global_4denvar ................... Passed 2109.08 sec
Start 2: rtma
2/6 Test #2: rtma ............................. Passed 1157.54 sec
Start 3: rrfs_3denvar_rdasens
3/6 Test #3: rrfs_3denvar_rdasens ............. Passed 915.69 sec
Start 4: hafs_4denvar_glbens
4/6 Test #4: hafs_4denvar_glbens .............. Passed 1465.61 sec
Start 5: hafs_3denvar_hybens
5/6 Test #5: hafs_3denvar_hybens .............. Passed 1225.12 sec
Start 6: global_enkf
6/6 Test #6: global_enkf ...................... Passed 865.98 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 7739.26 sec

JingCheng-NOAA · 2026-05-21T15:53:32Z

Thank you Russ and Xu.
And to Russ, for now the regression test cases for HAFS does not contain this change. I can modify the regression test settings to reflect it. But it will also require the observation dataset update, because the input TDR data will be superobed.
This may require another PR to update the HAFS regression test.

RussTreadon-NOAA

Minor comments.

ShunLiu-NOAA · 2026-05-21T15:55:21Z

@RussTreadon-NOAA and @XuLu-NOAA Thank you for completing ctest on WCOSS2.

Copilot

Pull request overview

This PR adds a configurable switch to control whether Tail Doppler Radar (TDR) radial-velocity observations are thinned along the radar beam, enabling users to bypass the existing hard-coded along-beam thinning when ingesting already super-obbed TDR BUFR input (as expected for HAFSv2.2).

Changes:

Introduces a new obs_input namelist option l_tdr_thin_alongbeam (default .true.) to toggle along-the-beam thinning.
Updates TDR processing in read_radar to conditionally apply the 3 km along-beam thinning logic based on l_tdr_thin_alongbeam.
Wires the new flag through obsmod and gsimod so it can be set via the existing namelist path.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
`src/gsi/read_radar.f90`	Applies conditional logic to bypass hard-coded along-beam thinning for TDR Vr when configured.
`src/gsi/obsmod.F90`	Adds the new public module logical and default initialization for `l_tdr_thin_alongbeam`.
`src/gsi/gsimod.F90`	Exposes `l_tdr_thin_alongbeam` via the `obs_input` namelist and updates inline namelist documentation.

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

RussTreadon-NOAA · 2026-05-25T12:51:43Z

+              if(l_tdr_thin_alongbeam) then
+!                Select data every 3 km along each beam
+                 if(MOD(INT(tdr_obs(1,k)-tdr_obs(1,1)),3000) < 100)then
+                    if(tdr_obs(3,k) >= 800.) then
+                       nmissing=nmissing+1     !xx
+                    else
+                       ii=ii+1
+                       dopbin(ii)=tdr_obs(3,k)
+                       thisrange=tdr_obs(1,k)
+
+                       call getvrlocalinfo(thisrange,thisazimuth,this_stahgt,aactual,a43,selev0,celev0, &
+                                  rlon0,clat0,slat0,r8,r89_5,nsubzero,ii,z(ii),elev(ii),elat8(ii), &
+                                  elon8(ii),glob_azimuth8(ii))
+                    end if
                 else
+                    ntdrvr_thin1=ntdrvr_thin1+1
+                 endif
+              else
+                 if(tdr_obs(3,k) >= 800.) nmissing=nmissing+1     !xx
+                 if(tdr_obs(3,k) < 800.) then


@JingCheng-NOAA : What do you think about this comment from Copilot? The intent of the suggested change is to improve readability of the run time log.

RussTreadon-NOAA · 2026-05-25T12:52:36Z

@@ -1101,7 +1102,7 @@ module gsimod
 !                         allows use of archived prepbufr files)

  namelist/obs_input/dmesh,time_window_max,time_window_rad, &
-       ext_sonde,l_foreaft_thin,hofx_2m_sfcfile, ignore_2mQM
+       ext_sonde,l_foreaft_thin,l_tdr_thin_alongbeam,hofx_2m_sfcfile, ignore_2mQM


@JingCheng-NOAA : Another Copilot suggestion to improve readability. What do you think?

RussTreadon-NOAA · 2026-05-21T15:58:01Z

@JingCheng-NOAA : Agreed. We do not need to change the HAFS GSI ctests ith this PR. A separate issue and PR is sufficient. Failure to keep GSI ctests in sync with added or modified GSI functionality runs the risk of these changes being unknowingly broken by other PRs.

RussTreadon-NOAA · 2026-05-21T16:01:46Z

@JingCheng-NOAA : Good suggestions from copilot. Please consider making the suggested changes.

JingCheng-NOAA · 2026-05-21T16:07:18Z

Thanks! will update the code accordingly.

…

On Thu, May 21, 2026 at 12:04 PM RussTreadon-NOAA ***@***.***> wrote: *RussTreadon-NOAA* left a comment (NOAA-EMC/GSI#1007) <#1007 (comment)> @JingCheng-NOAA <https://github.com/JingCheng-NOAA> : Good suggestions from copilot. Please consider making the suggested changes. — Reply to this email directly, view it on GitHub <#1007 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BAHEWIP5WMOQYLSSDLEDAYL434SGZAVCNFSM6AAAAACY5B6O6WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMJQGEYTQMBZGU> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

…statement related to along the beam thinning. Rephrase the namelist documentation for l_tdr_thin_alongbeam

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

RussTreadon-NOAA · 2026-05-25T12:54:04Z

+  if(l_tdr_thin_alongbeam) then
+    write(6,*)'READ_RADAR: # data removed by thinning along the beam ntdrvr_thin1=', ntdrvr_thin1
+  else
+    write(6,*) 'READ_RADAR: # offline superob applied on TDR, thinning along the beam is disabled'


@JingCheng-NOAA : Another improve readability suggestion from Copilot. What do you think?

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

RussTreadon-NOAA · 2026-05-25T12:49:24Z

@JingCheng-NOAA : 8ac349a adds my recommended change (i.e., echo obs_input to the run time log).

As expected, WCOSS2 (Cactus) ctests still pass after making change 8ac349a

Test project /lfs/h2/emc/da/noscrub/russ.treadon/git/gsi/pr1007/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  733.34 sec
2/6 Test #6: global_enkf ......................   Passed  855.98 sec
3/6 Test #2: rtma .............................   Passed  1100.31 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1234.52 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1344.58 sec
6/6 Test #1: global_4denvar ...................   Passed  1924.50 sec

100% tests passed, 0 tests failed out of 6

…to feature/hafs_tdr

JingCheng-NOAA · 2026-05-26T13:33:54Z

Accepted, and the corresponding change has been made

ShunLiu-NOAA · 2026-05-27T12:42:53Z

@JingCheng-NOAA could you run a ctest for a quick sanity check? otherwise I think the PR is in good shape to merge.

RussTreadon-NOAA · 2026-05-27T13:49:03Z

GSI ctests

Install JingCheng-NOAA:feature/hafs_tdr at f50ab69 as updat and NOAA-EMC:develop at 55faa5e as contrl on Gaea C6, Ursa, and WCOSS2 (Dogwood). Run GSI ctests with the following results

Gaea C6

Test project /gpfs/f6/ira-sti/scratch/Russ.Treadon/git/gsi/pr1007/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  482.07 sec
2/6 Test #6: global_enkf ......................   Passed  482.11 sec
3/6 Test #2: rtma .............................   Passed  722.54 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  903.47 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  963.50 sec
6/6 Test #1: global_4denvar ...................   Passed  1321.37 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1321.37 sec

Ursa

Test project /scratch3/NCEPDEV/da/Russ.Treadon/git/gsi/ursa/pr1007/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  1028.58 sec
2/6 Test #6: global_enkf ......................   Passed  1032.57 sec
3/6 Test #2: rtma .............................   Passed  1269.33 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1273.24 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1393.14 sec
6/6 Test #1: global_4denvar ...................   Passed  1744.21 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1744.23 sec

WCOSS2 (Dogwood)

Test project /lfs/h2/emc/da/noscrub/russ.treadon/git/gsi/pr1007/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  730.20 sec
2/6 Test #6: global_enkf ......................   Passed  854.87 sec
3/6 Test #2: rtma .............................   Passed  1033.75 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1220.57 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1221.04 sec
6/6 Test #1: global_4denvar ...................   Passed  1923.34 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1923.35 sec

All tests Passed on Gaea C6, Ursa, and WCOSS2 (Dogwood)

ShunLiu-NOAA · 2026-05-27T13:51:06Z

@RussTreadon-NOAA Thank you for the ctest.

JingCheng-NOAA · 2026-05-27T13:51:30Z

Ctests passed on Hera

Test project /scratch4/NCEPDEV/hwrf/save/Jing.Cheng/GSIversions/GSI_TDRupdate/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  487.27 sec
2/6 Test #2: rtma .............................   Passed  1028.56 sec
3/6 Test #6: global_enkf ......................   Passed  1087.83 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1091.26 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1211.16 sec
6/6 Test #1: global_4denvar ...................   Passed  1841.87 sec

100% tests passed, 0 tests failed out of 6

Total Test time (real) = 1841.88 sec

Added changes related to TDR thining selection. This change will be u…

04a319f

…sed in upcoming HAFSv2.2

ShunLiu-NOAA assigned JingCheng-NOAA and ShunLiu-NOAA May 13, 2026

ShunLiu-NOAA requested review from XuLu-NOAA and delippi May 13, 2026 20:58

XuLu-NOAA approved these changes May 13, 2026

View reviewed changes

JingCheng-NOAA mentioned this pull request May 14, 2026

Track Changes in HAFS GSI for V2.2 hafs-community/GSI#20

Open

ShunLiu-NOAA requested a review from yonghuiweng May 14, 2026 21:12

yonghuiweng approved these changes May 16, 2026

View reviewed changes

Merge branch 'NOAA-EMC:develop' into feature/hafs_tdr

e157e33

RussTreadon-NOAA reviewed May 21, 2026

View reviewed changes

Comment thread src/gsi/read_radar.f90 Outdated

RussTreadon-NOAA reviewed May 21, 2026

View reviewed changes

Comment thread src/gsi/read_radar.f90 Outdated

RussTreadon-NOAA reviewed May 21, 2026

View reviewed changes

Comment thread src/gsi/read_radar.f90 Outdated

RussTreadon-NOAA reviewed May 21, 2026

View reviewed changes

RussTreadon-NOAA requested a review from Copilot May 21, 2026 15:54

Copilot started reviewing on behalf of RussTreadon-NOAA May 21, 2026 15:55 View session

Copilot AI reviewed May 21, 2026

View reviewed changes

Replace 800. with 800.0_r_double for tdr_obs; Update the diagnostic …

f4f0144

…statement related to along the beam thinning. Rephrase the namelist documentation for l_tdr_thin_alongbeam

RussTreadon-NOAA requested a review from Copilot May 22, 2026 17:31

Copilot started reviewing on behalf of RussTreadon-NOAA May 22, 2026 17:32 View session

Copilot AI reviewed May 22, 2026

View reviewed changes

RussTreadon-NOAA and others added 2 commits May 22, 2026 13:44

correct typo in comment

4b81945

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

write namelist obs_input to stdout

8ac349a

JingCheng-NOAA added 2 commits May 26, 2026 13:26

Modified the output info with the l_tdr_thin_alongbeam = .false.

769b74e

Merge branch 'feature/hafs_tdr' of github.com:JingCheng-NOAA/GSI-1 in…

f50ab69

…to feature/hafs_tdr

ShunLiu-NOAA approved these changes May 27, 2026

View reviewed changes

ShunLiu-NOAA merged commit 860d137 into NOAA-EMC:develop May 27, 2026

Conversation

JingCheng-NOAA commented May 13, 2026

Uh oh!

JingCheng-NOAA commented May 13, 2026

Uh oh!

ShunLiu-NOAA commented May 19, 2026

Uh oh!

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

Uh oh!

Uh oh!

ShunLiu-NOAA commented May 21, 2026

Uh oh!

RussTreadon-NOAA commented May 21, 2026

Uh oh!

XuLu-NOAA commented May 21, 2026

Uh oh!

Uh oh!

JingCheng-NOAA commented May 21, 2026

Uh oh!

Uh oh!

Uh oh!

RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

Uh oh!

ShunLiu-NOAA commented May 21, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

RussTreadon-NOAA May 25, 2026

Choose a reason for hiding this comment

Uh oh!

RussTreadon-NOAA May 25, 2026

Choose a reason for hiding this comment

Uh oh!

RussTreadon-NOAA commented May 21, 2026

Uh oh!

RussTreadon-NOAA commented May 21, 2026

Uh oh!

JingCheng-NOAA commented May 21, 2026 via email

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

RussTreadon-NOAA May 25, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

RussTreadon-NOAA commented May 25, 2026

Uh oh!

JingCheng-NOAA commented May 26, 2026

Uh oh!

ShunLiu-NOAA commented May 27, 2026

Uh oh!

RussTreadon-NOAA commented May 27, 2026

Uh oh!

ShunLiu-NOAA commented May 27, 2026

Uh oh!

JingCheng-NOAA commented May 27, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

JingCheng-NOAA commented May 19, 2026 •

edited

Loading