Turn off anlstat job for GSI based experiments by default

[Copilot]

Description

The anlstat job was unconditionally running for GSI-based experiments, causing issues on some platforms. This PR turns it off by default for GSI experiments and adds an explicit opt-in flag, while preserving the existing behavior that enables DO_ANLSTAT for JEDI-based experiments.

Changes:

• dev/parm/config/gfs/config.base.j2: Reverts DO_ANLSTAT default back to "NO" (as it was in e8d51af); adds new DO_GSI_ANLSTAT="NO" flag; restores the conditional logic from e8d51af that turns DO_ANLSTAT back to "YES" for JEDI-based experiments:
  • When DO_JEDIATMVAR="YES" (JEDI atmospheric variational DA)
  • When any of DO_AERO="YES", DO_JEDIOCNVAR="YES", or DO_JEDISNOWDA="YES"
• dev/scripts/exglobal_analysis_stats.py: Replaces the unconditional else: append('atmos_gsi') with elif config.DO_GSI_ANLSTAT: — GSI analysis stats now only run when explicitly opted in
• docs/source/configure.rst: Added documentation for DO_ANLSTAT and DO_GSI_ANLSTAT options

# Before: atmos_gsi always added when JEDI not active
if config.DO_JEDIATMVAR:
    config.STAT_ANALYSES.append('atmos')
else:
    config.STAT_ANALYSES.append('atmos_gsi')

# After: atmos_gsi only added when DO_GSI_ANLSTAT=YES
if config.DO_JEDIATMVAR:
    config.STAT_ANALYSES.append('atmos')
elif config.DO_GSI_ANLSTAT:
    config.STAT_ANALYSES.append('atmos_gsi')

To re-enable GSI analysis statistics, set DO_GSI_ANLSTAT="YES" in the experiment configuration. JEDI-based experiments are unaffected — DO_ANLSTAT is still automatically enabled for them via the conditional logic in config.base.j2.

Type of change

• Bug fix (fixes something broken)
• New feature (adds functionality)
• Maintenance (code refactor, clean-up, new CI test, etc.)

Change characteristics

• Is this change expected to change outputs (e.g. value changes to existing outputs, new files stored in COM, files removed from COM, filename changes, additions/subtractions to archives)? NO
  • GFS
  • GEFS
  • SFS
  • GCAFS
• Is this a breaking change (a change in existing functionality)? YES — GSI experiments that relied on anlstat running by default will no longer run it unless DO_GSI_ANLSTAT="YES" is set
• Does this change require a documentation update? YES — documentation for DO_ANLSTAT and DO_GSI_ANLSTAT has been added to docs/source/configure.rst
• Does this change require an update to any of the following submodules? NO
  • EMC verif-global
  • GDAS
  • GFS-utils
  • GSI
  • GSI-monitor
  • GSI-utils
  • UFS-utils
  • UFS-weather-model
  • wxflow

How has this been tested?

Code review only; no runtime tests conducted.

Checklist

• Any dependent changes have been merged and published
• 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
• I have documented my code, including function, input, and output descriptions
• My changes generate no new warnings
• New and existing tests pass with my changes
• This change is covered by an existing CI test or a new one has been added
• Any new scripts have been added to the .github/CODEOWNERS file with owners
• I have made corresponding changes to the system documentation if necessary

---

⌨️ Start Copilot coding agent tasks without leaving your editor — available in VS Code, Visual Studio, JetBrains IDEs and Eclipse.

[CoryMartin-NOAA]

@copilot this isn't quite right, please look at the entire config.base file in that hash, there was logic towards the end that turned it back to YES depending on other settings

[Copilot]

@copilot this isn't quite right, please look at the entire config.base file in that hash, there was logic towards the end that turned it back to YES depending on other settings

Fixed in aa11aae. Restored the conditional block from e8d51af that sets DO_ANLSTAT="YES" for JEDI-based experiments — when DO_JEDIATMVAR="YES", or when any of DO_AERO, DO_JEDIOCNVAR, or DO_JEDISNOWDA are YES. GSI-only experiments remain off by default (controllable via DO_GSI_ANLSTAT).

[DavidHuber-NOAA]

@CoryMartin-NOAA @DavidNew-NOAA I think this would also resolve #4450 and could supersede #4465. Could we close #4465 in favor of this approach?

[CoryMartin-NOAA]

Probably. @DavidNew-NOAA thoughts? I somehow missed that this was open

[DavidNew-NOAA]

@CoryMartin-NOAA @DavidHuber-NOAA Yes, agree

[DavidHuber-NOAA]

Launching CI on Hercules.

@DavidHuber-NOAA The build on Hercules Slurm timed out building gdas ~T.McG

| gfs_ww3prepost     | SUCCEEDED  |
| gdas               | DEAD       |
| gsi_enkf           | SUCCEEDED  |

[2026-03-26T12:35:19.015] error: *** JOB 8391847 ON hercules-02-03 CANCELLED AT 2026-03-26T12:35:19 DUE TO TIME LIMIT **

[DavidHuber-NOAA]

@copilot please add documentation to docs/source/configure.rst to illustrate what the DO_ANLSTAT and DO_GSI_ANLSTAT options do.

[DavidHuber-NOAA]

I didn't know I could hijack someone else's copilot session 😮

[Copilot]

@copilot please add documentation to docs/source/configure.rst to illustrate what the DO_ANLSTAT and DO_GSI_ANLSTAT options do.

Added in 12a8990. The two new entries appear after DO_VMINMON in the monitoring section of the table:

• DO_ANLSTAT — defaults NO, adjusts XML; notes it's automatically set to YES for JEDI-based experiments
• DO_GSI_ANLSTAT — defaults NO, does not adjust XML; explains when it's relevant (GSI + DO_ANLSTAT=YES)

[TravisElless-NOAA]

Hercules CI failure here appears to be due to the build job timing out. I'll rekick it and see if it gets going again

[TerrenceMcGuinness-NOAA]

@TravisElless-NOAA Yes gdas timed out its build on Hercules. BTW when restarting a build job in GitLab it will start over again because it has to re-clone unlike all the other jobs that pick up where they left off.