add two band sdl to global_4denvar namelist

[RussTreadon-NOAA]

Description
GFS v17 runs gsi.x with scale-dependent localization (SDL). Update the global_4denvar ctest to include SDL.

Fixes #756

Type of change

• New feature (non-breaking change which adds functionality)

How Has This Been Tested?
Install feature/sdl on Dogwood and run global_4denvar ctest. Confirm correct namelist parameters are set to activate SDL.

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code

[RussTreadon-NOAA]

global_4denvar ctests have been run on Dogwood with acceptable results.

[RussTreadon-NOAA]

@CatherineThomas-NOAA and @TravisElless-NOAA : This PR activates two-band SDL in the global_4denvar ctest. Do either of you have time to review the changes in this PR?

[CatherineThomas-NOAA]

@RussTreadon-NOAA: I can review these changes today.

[RussTreadon-NOAA]

Thank you @CatherineThomas-NOAA

[TravisElless-NOAA]

Potential newbie question. Should regression_namelists_db.sh also be updated with these values given this if block?

[CatherineThomas-NOAA]

My ctests ran on Hera without issue. Looking at the log for the ctests, I see that the parameters are set appropriately:

 NSCLGRP =           2,
 NGVARLOC        =           1,
 NAENSLOC        =           3,
 R_ENSLOCCOV4TIM =   1.00000000000000     ,
 R_ENSLOCCOV4VAR =   1.00000000000000     ,
 R_ENSLOCCOV4SCL =   1.00000000000000     ,
 S_ENS_H =   1000.00000000000     ,   300.000000000000     ,   685.000000000000     , 17*2828.00000000000        ,
 S_ENS_V = 2*-0.500000000000000      ,  0.000000000000000E+000, 17*30.0000000000000        ,
 READIN_LOCALIZATION     = F,
 GLOBAL_SPECTRAL_FILTER_SD       = F,

The only snag is that due to the decreased resolution of the ctest, our chosen second scale is too small:

( pi*rearth*.001/jcap_in <= s_ens_hv <= 5500 )

Our second scale localization value of 300 km is reset automatically:

  s_ens_hv(           1 ) =    300.000000000000       km--too small, min value =
     416.994064886485       km.
  s_ens_hv(           1 ) reset to min value

Once it's reset, the scales are showing as expected:

 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48   1000.00    0.4608313731E-11
 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48    416.99    0.4824141087E-11
 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48    685.00    0.4678923915E-11

@RussTreadon-NOAA, @TravisElless-NOAA: Should we adjust our second scale so these warnings aren't set off? 450 km instead of 300?

[RussTreadon-NOAA]

Good suggestion @TravisElless-NOAA: I will add SDL to the global_4denvar case in regression_namelists_db.sh

Good catch @CatherineThomas-NOAA: I will increase the 300 km scale to 450 km.

[RussTreadon-NOAA]

Rerun global_4denvar on Dogwood with 450 km scale. No more warning messages

  reading time level            7
 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48   1000.00    0.4608313731E-11
 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48    450.00    0.4799716180E-11
 in init_sf_xy, jcap,s_ens_hv(    1 -  127), max diff(f0-f)=        48    685.00    0.4678923915E-11
GLBSOI: jiter,jiterstart,jiterlast,jiterend=    1    1    2    1

ctest still passes (it should!)

Test project /lfs/h2/emc/da/noscrub/russ.treadon/git/gsi/sdl/build
    Start 1: global_4denvar
1/1 Test #1: global_4denvar ...................   Passed  2344.44 sec

100% tests passed, 0 tests failed out of 1

Total Test time (real) = 2344.45 sec

[CatherineThomas-NOAA]

Looks great. Thanks @RussTreadon-NOAA!

[RussTreadon-NOAA]

Update modulefiles/gsi_orion.intel.lua in order to compile feature/sdl on Orion and run ctest. Done at aab4a06.

Sorry @CatherineThomas-NOAA - this commit dismissed your approval.

[CatherineThomas-NOAA]

That's fine @RussTreadon-NOAA. I'll wait until the Orion tests complete before approving again.

[RussTreadon-NOAA]

WARNING

Updating to Orion Rocky 9 is not as simple as updating a few files. The Orion and Hercules Rocky 9 crtm-fix/2.4.0.1_emc module do not contain the same coefficient files.

Orion crtm-fix/2.4.0.1_emc has

setenv("CRTM_FIX","/work/noaa/epic/role-epic/spack-stack/orion/spack-stack-1.6.0/envs/unified-env-rocky9/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-qls55kd/fix")

whereas Hercules crtm-fix/2.4.0.1_emc has

setenv("CRTM_FIX","/work/noaa/epic/role-epic/spack-stack/hercules/spack-stack-1.6.0/envs/unified-env/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-2os2hw2/fix")

A diff -r of these two CRTM_FIX directories finds 447 files differ. This is not correct.

hercules-login-3:~$ diff -r /work/noaa/epic/role-epic/spack-stack/orion/spack-stack-1.6.0/envs/unified-env-rocky9/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-qls55kd/fix /work/noaa/epic/role-epic/spack-stack/hercules/spack-stack-1.6.0/envs/unified-env/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-2os2hw2/fix |wc
    447    2680  159396

For example,

hercules-login-3:~$ ls -l /work/noaa/epic/role-epic/spack-stack/orion/spack-stack-1.6.0/envs/unified-env-rocky9/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-qls55kd/fix/abi_gr.TauCoeff.bin /work/noaa/epic/role-epic/spack-stack/hercules/spack-stack-1.6.0/envs/unified-env/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-2os2hw2/fix/abi_gr.TauCoeff.bin
-rw-r--r-- 1 role-epic epic  10972 Dec 20  2023 /work/noaa/epic/role-epic/spack-stack/hercules/spack-stack-1.6.0/envs/unified-env/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-2os2hw2/fix/abi_gr.TauCoeff.bin
-rw-r--r-- 1 role-epic epic 184588 Jun 14 16:06 /work/noaa/epic/role-epic/spack-stack/orion/spack-stack-1.6.0/envs/unified-env-rocky9/install/intel/2021.9.0/crtm-fix-2.4.0.1_emc-qls55kd/fix/abi_gr.TauCoeff.bin

Given this, revert commits 086e273 and aab4a06 from feature/sdl.

[RussTreadon-NOAA]

@CatherineThomas-NOAA : I abandoned the Orion ctest attempt. While updating the module is easy, the CRTM coefficients when loading the Orion crtm-fix/2.4.0.1_emc differ from those one gets when loading the Hercules crtm-fix/2.4.0.1_emc. This is not correct. Leave this to be sorted out in GSI issue #754

I am done making changes to this PR.

[TravisElless-NOAA]

All looks good to me. Thanks @RussTreadon-NOAA!