Update global_4denvar and global_enkf ctests to reflect GFS v17

[RussTreadon-NOAA]

The global_4denvar (gsi.x) and global_enkf (enkf.x) scripts and namelists

• regression/global_4denvar.sh
• regression/global_enkf.sh
• regression/regression_namelists.sh
• regression/regression_namelists_db.sh

should reflect how we intend to run gsi.x and enkf.x in GFS v17. This issue is opened to document these changes.

[RussTreadon-NOAA]

@CatherineThomas-NOAA , @emilyhcliu : Any input regarding what needs to change? Whom else should we tag for input?

[CatherineThomas-NOAA]

@RussTreadon-NOAA : One that immediately comes to my mind is that reducedgrid needs to be set to .false. for the EnKF. I'll take a closer look tomorrow and circle back to you for the rest.

[RussTreadon-NOAA]

@CatherineThomas-NOAA , if you have log files from a GFS v17 parallel, I can get the namelist changes from these. No need to add to your workload.

[RussTreadon-NOAA]

Work for this issue will be done in RussTreadon-NOAA:feature/global_ctests

[CatherineThomas-NOAA]

@RussTreadon-NOAA : Here is what I saw in a quick run through of some logs, ignoring known differences in resolution/iterations

deterministic (regression v g-w develop)

• IMP_PHYSICS = 11 v 8
• TA2TB = F v T
• OPTCONV = 0.05 v 0.0
• CRIS_CADS/IASI_CADS = T v F

ensemble (regression v g-w develop)

• READIN_LOCALIZATION = T v F
• REDUCEDGRID = T v F
• IMP_PHYSICS = 11 v 8

Comments

• imp_physics is the switch to from GFDL (11) to Thompson (8)
• The CADS parameters were set to F for now by Jim J in the workflow until further testing was completed.
• I know optconv was added for improved minimization, but I'm unsure of the tested/desired value.
• I believe ta2tb should be true, @emilyhcliu can confirm.
• readin_localization should be set to F due to the model-space localization LETKF
• reducedgrid should be set to F

I did not look at the obs bufrs list. I have a recent g-w test here if anyone wants to look at the namelists more closely:
/scratch2/NCEPDEV/stmp1/Catherine.Thomas/ROTDIRS/archive_restartb/

[RussTreadon-NOAA]

Thank you @CatherineThomas-NOAA . I'll digest your findings, look at some log files, and update global_4denvar and global_enkf namelists accordingly.