b4b-dev: Move nml parameters (changes paramfile)

[swensosc]

Description of changes

move namelist parameters to parameter file. parameters moved are: baseflow_scalar, precip_repartition_nonglc_all_rain_t, precip_repartition_nonglc_all_snow_t, precip_repartition_glc_all_rain_t, precip_repartition_glc_all_snow_t

Specific notes

Contributors other than yourself, if any:

CTSM Issues Fixed (include github issue #):

Are answers expected to change (and if so in what way)? No

Any User Interface Changes (namelist or namelist defaults changes)? Yes

Does this create a need to change or add documentation? Did you do so? No No

Testing performed, if any:

@slevis-lmwg working locally in /glade/work/slevis/git/b4b-dev

[ekluzek]

I did a quick skim over this and this looks good to me. I'll rebase this PR to go to b4b-dev, which should be fine without any trouble. Sometimes you have to rebase the branch when that's done, but I don't expect that now.

The one thing noticeably missing is an updated parameter file with the parameters moved to it.

[ekluzek]

Adding next to talk about who will shepherd this in.

[swensosc]

parameter file location: /fs/cgd/csm/inputdata/lnd/clm2/paramdata/ctsm5.3.041.Nfix_params.v13.c250221_upplim250_add_nml.nc

[ekluzek]

We talked about this and will bring it in on b4b-dev.

[ekluzek]

I had looked at this when this first came in and thought I had already approved. So approving now.

I skimmed through again and didn't notice anything that warrants attention.

[slevis-lmwg]

Paused testing because need to bring the new params (see post above) to the latest paramfiles:

OK ./build-namelist_test.pl
./run_sys_tests -s aux_clm -c ctsm5.4.001 --skip-generate
./run_sys_tests -s aux_clm -c ctsm5.4.001 --skip-generate --skip-git-status

[slevis-lmwg]

Before I add the new params to paramfiles:

• list the latest paramfiles on master and b4b-dev
  master currently ctsm5.4.004 has the latest and Erik tells me that we do not expect other paramfile updates:

<paramfile phys="clm6_0" >lnd/clm2/paramdata/ctsm60_params.c251211.nc</paramfile>
<paramfile phys="clm5_0" >lnd/clm2/paramdata/clm50_params.c250311.nc</paramfile>
<paramfile phys="clm4_5" >lnd/clm2/paramdata/clm45_params.c250311.nc</paramfile>

ctsm60 and clm50 are the paramfiles to update, not clm45.

Consider the new tool in #3622 if needed.

[slevis-lmwg]

I generated two new paramfiles with the updates:

/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/clm50_params.c251222.nc
/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60_params.c251222.nc

• Repeat with current paramfile
• After testing, ./rimport and relink.py

[ekluzek]

Ran testing on Derecho last night. Things worked as expected with the exception of the following:

ERS_Ld3.f10_f10_mg37.I2000Clm60Bgc.derecho_intel.clm-ciso_cwd_hr        (NLCOMP RUN)
ERS_Ld3.f10_f10_mg37.I2000Clm60Bgc.derecho_intel.clm-ciso_cwd_hr--clm-matrixcnOn        (NLCOMP RUN)
ERS_Ld3.f10_f10_mg37.I2000Clm60BgcCrujra.derecho_intel.clm-ciso_cwd_hr  (NLCOMP RUN)
FUNITCTSM_P1x1.f10_f10_mg37.I2000Clm50Sp.derecho_intel  (NLCOMP RUN)
SMS_D.f10_f10_mg37.I2000Clm60BgcCrop.derecho_nvhpc.clm-crop     (NLCOMP RUN)            EXPECTED (RUN)
SMS_D_Mmpi-serial_Ld5.5x5_amazon.I2000Clm60Bgc.derecho_gnu.clm-HillslopeC       (NLCOMP RUN)
SMS_Lm1.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-clm60_monthly_matrixcn_soilCN30    (NLCOMP RUN)
SMS_Lm1.f10_f10_mg37.IHistClm60Bgc.derecho_intel.clm-leafcn_t_evolving  (NLCOMP RUN)

I think this is mostly because of tests that have custom parameter files.

[swensosc]

the original code mixes units, and I thought it was more clear to add the suffix to the variable name.

…

On Thu, Mar 5, 2026 at 11:30 AM Erik Kluzek ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In src/main/atm2lndType.F90 <#3391 (comment)>: > - ' ERROR: For repartition_rain_snow true, precip_repartition_glc_all_snow_t must be provided') - end if - if (.not. present(precip_repartition_glc_all_rain_t)) then - call endrun(subname // & - ' ERROR: For repartition_rain_snow true, precip_repartition_glc_all_rain_t must be provided') - end if - if (.not. present(precip_repartition_nonglc_all_snow_t)) then - call endrun(subname // & - ' ERROR: For repartition_rain_snow true, precip_repartition_nonglc_all_snow_t must be provided') - end if - if (.not. present(precip_repartition_nonglc_all_rain_t)) then - call endrun(subname // & - ' ERROR: For repartition_rain_snow true, precip_repartition_nonglc_all_rain_t must be provided') - end if + ! non-glacier all rain temperature (degrees C) + call readNcdioScalar(ncid, 'precip_repartition_nonglc_all_rain_t', subname, precip_repartition_nonglc_all_rain_t_celsius) @swensosc <https://github.com/swensosc> question here. This is reading in the variable from the params file without the suffix "_celsius" in the name -- but putting it into a variable with that suffix. This is something that can be confusing when variables have a different name than the variable name on input files. — Reply to this email directly, view it on GitHub <#3391 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGRN57GKF3BRTUOPV2BNZOT4PHBVLAVCNFSM6AAAAACDLEK2KCVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTQOJYG43DSNZWHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[slevis-lmwg]

I have picked this up again hoping to merge this week if possible.

I made changes to the constructor as I understood them from the conversation above.
I submitted these tests from the short list of failures that I saw above:

FAIL SMS_Lm1.f10_f10_mg37.IHistClm60Bgc.derecho_intel.clm-leafcn_t_evolving.C.20260310_154544_rur8jp DIFF
FAIL ERS_Ld3.f10_f10_mg37.I2000Clm60Bgc.derecho_intel.clm-ciso_cwd_hr.C.20260310_161644_a4bfhi DIFF
FAIL FUNITCTSM_P1x1.f10_f10_mg37.I2000Clm50Sp.derecho_intel.C.20260310_162208_8z7rve RUN
FAIL SMS_D_Mmpi-serial_Ld5.5x5_amazon.I2000Clm60Bgc.derecho_gnu.clm-HillslopeC.C.20260310_165833_cbz4lj DIFF
FAIL SMS_Lm1.f10_f10_mg37.I1850Clm60Bgc.derecho_intel.clm-clm60_monthly_matrixcn_soilCN30.C.20260310_170039_nb7q0s DIFF

comparing to baseline ctsm5.4.021.

FUNIT still fails at RUN.
The others now fail with DIFFs from the baseline, unlike the 4th in the list when I tested it earlier: #3391 (comment)

@ekluzek I will push my mods to get your feedback.

UPDATE 2026/3/11 after the mods discussed below:

PASS FUNITCTSM_P1x1.f10_f10_mg37.I2000Clm50Sp.derecho_intel.C.20260311_174252_r947vm
PASS SMS_D_Mmpi-serial_Ld5.5x5_amazon.I2000Clm60Bgc.derecho_gnu.clm-HillslopeC.C.20260311_173515_tmxfun

So I will push my mods and submit aux_clm.

[slevis-lmwg]

./run_sys_tests -s aux_clm -c ctsm5.4.021 --skip-generate

• derecho
• izumi

[slevis-lmwg]

@ekluzek please look if you want to go over my latest changes. Tests are passing now, including FUNIT, so I submitted aux_clm.

I'm not confident about one variable naming choice: I kept Sean's _celsius suffix in subr. ReadParams and removed it elsewhere. Would you rather I kept it throughout?