Repository navigation
Description
Description:
Baseline for PPE tests with 5.4 surface data, clone of #100
- 2 degree no-crop
- Sturm used over veg and lake column, Jordan over glacier
@linniahawkins and @djk2120, feel free to suggest a different baseline for PPE tests you'd like to run. This will build on what we have in 98-100 which are 1 degree runs designed to evaluate 5.4 surface data and the snow changes being requested for CESM3.
Case directory:
Locally:
/glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_AD
./create_newcase --case /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_AD --compset I1850Clm60BgcCrujra --res f19_g17 --run-unsupported
./create_clone --clone /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_AD --case /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_SASU --keepexe
./create_clone --clone /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_SASU --case /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_pSASU --keepexe
./create_newcase --case /glade/work/wwieder/ctsm5.3.0/ctsm53041_54surfdata_PPEbaseline_101_HIST --compset HIST_DATM%CRUJRA2024_CLM60%BGC_SICE_SOCN_MOSART_SGLC_SWAV_SESP --res f19_g17 --run-unsupported
Sandbox:
Locally:
/glade/u/home/wwieder/CTSM/
branch = JordanGlacier, even with tag == 5.3.041
- AD
user_nl_ changes:
As in CTSM5.4 tests: snowTherm_100: ctsm53041_54surfdata_snowTherm_100_[AD, SASU, pSASU, hist] #100, turn on Sturm over lake and Jordan on glaciers
fsurdat = '/glade/derecho/scratch/wwieder/mksrf_5.4/surfdata_1.9x2.5_hist_1850_78pfts_c250429.nc'
paramfile = '/glade/work/wwieder/modified_inputs/lhc0000_53041_compat.nc'
snow_thermal_cond_glc_method = 'Jordan1991'
snow_thermal_cond_lake_method = 'Sturm1997'
hist_fincl1 = 'TOTECOSYSC', 'TOTECOSYSN', 'TOTSOMC', 'TOTSOMN', 'TOTVEGC', 'TOTVEGN', 'TLAI', 'GPP', 'CPOOL', 'NPP', 'TWS','H2OSNO','NFIX'
do_grossunrep=.false.
During the spin-up, we also set the following xml changes
./xmlchange MOSART_MODE=NULL
./xmlchange STOP_N=40,STOP_OPTION=nyears
./xmlchange CLM_ACCELERATED_SPINUP=on,CLM_FORCE_COLDSTART=on
./xmlchange CONTINUE_RUN=FALSE,RESUBMIT=8
- SASU
./xmlchange RUN_TYPE=hybrid
./xmlchange GET_REFCASE=TRUE
./xmlchange RUN_REFDATE=0361-01-01
./xmlchange DRV_RESTART_POINTER=rpointer.cpl.0361-01-01-00000
./xmlchange CLM_ACCELERATED_SPINUP=sasu,CLM_FORCE_COLDSTART=off
./xmlchange CONTINUE_RUN=FALSE,RESUBMIT=7
./xmlchange RUN_REFDIR=/glade/derecho/scratch/wwieder/archive/ctsm53041_54surfdata_PPEbaseline_101_AD/rest/0361-01-01-00000
./xmlchange RUN_REFCASE=ctsm53041_54surfdata_PPEbaseline_101_AD
echo "reseed_dead_plants = .true." >> user_nl_clm
echo "finidat = 'ctsm53041_54surfdata_PPEbaseline_101_AD.clm2.r.0361-01-01-00000.nc '" >> user_nl_clm
- pSASU
./xmlchange RUN_REFDIR=/glade/derecho/scratch/wwieder/archive/ctsm53041_54surfdata_PPEbaseline_101_SASU/rest/0321-01-01-00000
./xmlchange RUN_REFCASE=ctsm53041_54surfdata_PPEbaseline_101_SASU
./xmlchange RUN_REFDATE=0321-01-01
./xmlchange DRV_RESTART_POINTER=rpointer.cpl.0321-01-01-00000
./xmlchange STOP_N=40,STOP_OPTION=nyears
./xmlchange CLM_ACCELERATED_SPINUP=off,CLM_FORCE_COLDSTART=off
./xmlchange CONTINUE_RUN=FALSE,RESUBMIT=3
echo "finidat = 'ctsm53041_54surfdata_PPEbaseline_101_SASU.clm2.r.0321-01-01-00000.nc'" >> user_nl_clm
- HIST
Keith had to modify initial conditions to get this to work...
1850-1871
./xmlchange RUN_TYPE=startup
./xmlchange RUN_STARTDATE=1850-01-01
./xmlchange STOP_OPTION=nyears
./xmlchange STOP_N=21
./xmlchange CONTINUE_RUN=FALSE
./xmlchange RESUBMIT=0
./xmlchange DATM_YR_ALIGN=1901
./xmlchange DATM_YR_START=1901
./xmlchange DATM_YR_END=1920
1871-1901
./xmlchange STOP_N=30
./xmlchange CONTINUE_RUN=TRUE
1901-2023
./xmlchange STOP_N=31
./xmlchange DATM_YR_END=2023
./xmlchange RESUBMIT=3
./xmlchange DATM_SKIP_RESTART_READ=TRUE
- user_nl_clm
SourceMods:
None, using JordanGlacier branch, CTSM PR#3072
Diagnostics:
Diags (if available)
ILAMB
https://webext.cgd.ucar.edu/I20TR/ctsm53041_54surfdata_PPEbaseline_101_HIST/lnd/_build_PPEbaseline_101/
Output:
Output (if still available):
/glade/derecho/scratch/wwieder/archive/ctsm53041_54surfdata_PPEbaseline_101_HIST
Historical output now at:
/glade/campaign/cgd/tss/common/Land_Only_Simulations/CTSM53_DEV/ctsm53041_54surfdata_PPEbaseline_101_HIST
Restarts from AD, SASU, pSASU are at:
/glade/campaign/cgd/tss/people/oleson/CLM5_restarts/
Contacts:
@wwieder, @linniahawkins
Extra details:
AD startup 360 yrs, 2.16% disequilibrium
SASU hybrid 320 yrs, 12.75% disequilibrium
pSASU hybrid 160 yrs, 2.87% disequilibrium
Activity
linniahawkins commentedon May 5, 2025
Given the other 1-degree tests that are running; I think the most useful thing for us would be to run this simulation with this parameter file:
/glade/work/linnia/CLM6-PPE/ctsm6_lhc/paramfiles/lhc0000.nc
So we can evaluate surface dataset changes against our "default" case.
wwieder commentedon May 6, 2025
OK, would it be best to run a 2 degree default case that's identical to #100 and then this test case with your
lhc0000.ncparameter file or do you just want to use thelhc0000as your PPE baseline? (I think our last 2 degree case was #88 )wwieder commentedon May 6, 2025
Oh, now I see that this is the same parameter file you had in #88? So we ARE looking looking 2 degree clone of #100, using the #88 parameter file. I may include
upplim_destruct_metamorph = 250if that's OK?linniahawkins commentedon May 6, 2025
Correct, 2-degree clone of #100 but with lhc0000.nc paramfile. I would prefer if we kept upplim_destruct_metamorph=175 so the comparison is equivalent.
Since upplim_destruct_metamorph was included in the CLM6 PPE we are performing the calibration with it set to 250. So, we do not need to evaluate the impact of that change in this test (the emulators have already learned it). Thanks!
[-]PPE 5.4 baseline [AD, SASU, pSASU, HIST][/-][+]PPE 5.4 baseline: ctsm53041_54surfdata_PPEbaseline_101_ [AD, SASU, pSASU, HIST][/+][-]PPE 5.4 baseline: ctsm53041_54surfdata_PPEbaseline_101_ [AD, SASU, pSASU, HIST][/-][+]PPE 5.4 baseline: ctsm53041_54surfdata_PPEbaseline_101_[AD, SASU, pSASU, HIST][/+][-]PPE 5.4 baseline: ctsm53041_54surfdata_PPEbaseline_101_[AD, SASU, pSASU, HIST][/-][+]PPE 5.4 baseline: ctsm53041_54surfdata_PPEbaseline_101_[AD,SASU,pSASU,HIST][/+]wwieder commentedon May 6, 2025
Minor issue, but we've added a new parameter to the default parameter files (
mimics_fi) that's missing on the parameter file you suggested, @linniahawkins. Not to worry, I copied it onto a new file, but at some point it would be helpful to have on the parameter sets suggested from the PPE. e.g.:wwieder commentedon May 9, 2025
Is there any pft-level data you want on the historical simulation, @linniahawkins? I can repeat what we have for #98, without crop data?
linniahawkins commentedon May 9, 2025
wwieder commentedon May 9, 2025
@olyson, can you work tall building magic on this case so I can do a historical run, NOTE it's 2 degree, no crop.
/glade/derecho/scratch/wwieder/archive/ctsm53041_54surfdata_PPEbaseline_101_pSASU/rest/0161-01-01-00000/ctsm53041_54surfdata_PPEbaseline_101_pSASU.clm2.r.0161-01-01-00000.ncolyson commentedon May 9, 2025
@wwieder , try this:
/glade/derecho/scratch/oleson/ANALYSIS/ctsm53041_54surfdata_PPEbaseline_101_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
wwieder commentedon May 10, 2025
@linniahawkins historical result are here
/glade/derecho/scratch/wwieder/archive/ctsm53041_54surfdata_PPEbaseline_101_HIST/lnd/histLet us know if you want daigs on these vs. previous 2 degree runs (#88), or when you're ready to suggest alternative parameter sets to test out.
linniahawkins commentedon May 16, 2025
Thanks, this was very helpful. It looks like there are slight differences in the region of Brazil where 5.4 surfdat has less BETT and more grass than 5.3 surfdat.
Figure 1. differences in LAI, GPP, and TOTVEGC for 2001-2020 in CLM 5.3 surfdat - 5.4 surfdat.
Our biome means are slightly affected in the tropical savanna, subtropical savannna and boreal shrubland. But this is much smaller than the error due to sparse grid versus full grid.
Figure 2. Biome mean LAI from full grid simulations with 5.3 surfdat (x-axis) and 5.4 surfdat (y-axis).
Overall, the error is good to be aware of but not a concern for PPE calibration.
wwieder commentedon Aug 13, 2025
@olyson can we move this to campaign store, it's the CTSM5.3 PPE baseline that uses 5.4 surface datasets