Update select JEDI hashes

[RussTreadon-NOAA]

This PR updates select jedi hashes in sorc with recent JCSDA updates.

Resolves: #1313
Resolves: #1407
Resolves: #1419

Depends

• jcb-gdas #56
• DA-utils #7

[emcbot]

Automated GW-GDASApp Testing Results:
Machine: hera

Start: Fri Dec 20 15:42:50 UTC 2024 on hfe08
---------------------------------------------------
Build:                                 *SUCCESS*
Build: Completed at Fri Dec 20 16:37:12 UTC 2024
---------------------------------------------------
Tests: ctest -j12 -R gdasapp -E atm_jjob|C96C48_ufs_hybatmDA|C96C48_hybatmaerosnowDA|C48mx500_3DVarAOWCDA|C48mx500_hybAOWCDA|C96C48_hybatmDA
Tests:                                 *SUCCESS*
Tests: Completed at Fri Dec 20 16:38:01 UTC 2024
Tests: 100% tests passed, 0 tests failed out of 32

[RussTreadon-NOAA]

Running full suite of gdasapp ctests, both GDASApp and g-w based, using g-w develop at e684944) and feature/stable-nightly at 5ea120e.

Based on last night's stable-nightly results all tests should pass. So far 110/134 tests have run and passed. Several g-w based jobs are PENDING in the queue. I'll update this PR with the full ctest results once all tests finish.

[CoryMartin-NOAA]

approved pending successful completion of tests

[RussTreadon-NOAA]

jcb-algorithm PR #9 removes Jb from the test reference output file. This caused test_gdasapp_atm_jjob_var_run to fail

0: OOPS_STATS Run end                                  - Runtime:     53.83 sec,  Memory: total:     9.52 Gb, per task: min =     1.54 Gb, max =     1.64 Gb
0: Run: Finishing oops::Variational<FV3JEDI, UFO and IODA observations> with status = 0
0: terminate called after throwing an instance of 'oops::TestReferenceTextMismatchError'
0:   what():  Test reference Text mismatch @ Line:1
0: Test: 'CostJo   : Nonlinear Jo(AMSUA N19) = 3.9490015258314117e+04, nobs = 73667, Jo/n = 5.3606112992675303e-01, err = 6.2697697603634124e+00'
0: Ref:  'CostJb   : Nonlinear Jb = 0.0000000000000000e+00'
srun: error: h35m34: task 0: Aborted (core dumped)

I suspect other ctests will fail the reference check. I will update the references in feature/stable-nightly once the updated reference files from all failures are available.

Previous tests did not catch this because feature/stable-nightly was using an earlier hash of jcb-algorithms.

[danholdaway]

Thanks @RussTreadon-NOAA. Apologies, I let Guillaume remove the backward compatibility there. I didn't realize that would be included in the ref.

[RussTreadon-NOAA]

No problem. The reference check is doing its job.

[RussTreadon-NOAA]

Ugh, updating the reference file dismissed the approvals!

@CoryMartin-NOAA, @danholdaway , or @DavidNew-NOAA : If you see this message would you please approve this PR.

g-w based reference checks did not fail because g-w PR #3177 turned off reference checks in g-w based GDASApp ctests.

We still need to update select reference files for g-w based ctests given jcb-algorithm PR #9, but we don't need to do so for this PR.

[RussTreadon-NOAA]

Thank you @CoryMartin-NOAA ! I'll go ahead and merge.