b.e30_alpha08i.B1850C_LTso.ne30_t232_wgx3.315 (spatially distributed params)

[slevis-lmwg]

Description:

• Discuss this with Erik on March 9
• Run with alpha08b or 08g: Erik recommended 08i since available
• See email titled "Made a branch tag for distributed parameters code" for more info
• Hybrid startup from Cecile's simulation 315 year 100 (or 316 or similar)
• Simulation similar to Cecile's simulation 315, including namelist changes for all components
• Clm will
  • point to branch_tags/distparams.n02.ctsm5.4.022 and
  • new spatially distributed paramfile from Sean

./create_newcase --compset 1850C_CAM70%LT_CLM60%BGC-CROP_CICE_MOM6_MOSART_DGLC%NOEVOLVE_WW3_SESP --res ne30pg3_t232_wg37 --case ~slevis/cases_ctsm5.4/b.e30_alpha08i.B1850C_LTso.ne30_t232_wgx3.315 --run-unsupported

---

Case directory:
Locally:
~slevis/cases_ctsm5.4/b.e30_alpha08i.B1850C_LTso.ne30_t232_wgx3.315

---

Sandbox:
Locally:
/glade/work/slevis/git_cesm/cesm3
git describe for cesm and clm returns:
cesm3_0_alpha08i
branch_tags/distparams.n02.ctsm5.4.022 (git branch: distributed_parameters, PR ESCOMP/CTSM#3405, I had to use git fetch --tags before checking out this branch tag)

---

user_nl_clm:

use_distributed_parameters = .true.
! The distparam file MUST match the model resolution
stream_fldfilename_distparams = '$DIN_LOC_ROOT/lnd/clm2/paramdata/paramdata_ctsm60_params_c260204_ne30np4.pg3_hist_1850_78pfts_c251022.nc'
stream_meshfile_distparams = '$LND_DOMAIN_MESH'

paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60-cam70_params.c260305.nc'

reseed_dead_plants = .true.

For future runs, Sean asked that we change from mapalgo_distparams = bilinear to mapalgo_distparams = nn

user_nl_cam:

seasalt_emis_scale = 1.6

tracer_cnst_datapath = '/glade/derecho/scratch/tilmes/archive/b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001/atm/proc/TracerCnstCyclical1850'
tracer_cnst_file = 'tracer_cnst_MTt4s_halons_3Dmonthly_1850climo251_c20251209.nc'

xs_long_file = '/glade/campaign/cesm/cesmdata/inputdata/atm/waccm/phot/temp_prs_GT200nm_c221005mm.nc'

cldfrc_dp1 = 0.1

clubb_c8 = 4.5

clubb_c_k10h = 0.20

dust_emis_fact = 3.2

se_statefreq = 144

micro_mg_dcs = 500.D-6

micro_mg_vtrmi_factor = 1.0

microp_aero_wsub_scale = 0.8

stream_ndep_data_filename = '/glade/derecho/scratch/tilmes/archive/b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001/atm/proc/DepositionCyclical1850/fndep_clm_B1850C_MT4s.1850monthly.ne30_c251220.nc'

prescribed_ozone_datapath = '/glade/derecho/scratch/tilmes/archive/b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001/atm/proc/ozone_strataero_day_5'
prescribed_ozone_file = 'ozone_strataero_b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001_1850climo_zm_5day_c251204.nc'

prescribed_strataero_datapath = '/glade/derecho/scratch/tilmes/archive/b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001/atm/proc/ozone_strataero_day_5'
prescribed_strataero_file = 'ozone_strataero_b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001_1850climo_zm_5day_c251204.nc'

zmconv_c0_lnd                = 0.0035D0

user_nl_cice:

restart_fsd = .true.
 
tscale_pnd_drain = 0.5d0
 
distribution_type = "spacecurve"

user_nl_mom:
1st attempt used:

TAPER_LEITHY = True
LEITHY_DEPTH = 1000.0
LEITHY_WIDTH = 500.0
VISC_RES_FN_POWER = 6 

2nd attempt as explained in #148 (comment)

---

SourceMods:
1st_attempt no SourceMods
2nd_attempt as explained in
#148 (comment)
#148 (comment)

---

Diagnostics:
vs the 315 baseline

• Initially (years 1-3 vs 80-99)

---

Output:
Output (if still available):

/glade/scratch/slevis/archive/$CASE/1st_attempt
/glade/scratch/slevis/archive/$CASE

---

Contacts:
@slevis-lmwg

---

Extra details:
1st_attempt I got the thing to run
2nd_attempt includes corrections from @gustavo-marques @ekluzek

[slevis-lmwg]

I got this to start after a few failed attempts. A few concerns to raise:

1. Am I running the correct grid?
2. @gustavo-marques does this seem ok?
   diff /glade/derecho/scratch/slevis/b.e30_alpha08i.B1850C_LTso.ne30_t232_wgx3.315/run/MOM_override /glade/derecho/scratch/hannay/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.315/run/. returns:

---
> USE_KH_BG_2D = True
> KH_BG_2D_FILENAME = "KH_BG_2D.nc"
> KH_BG_2D_VARNAME = "KH"

3. @ekluzek I found this default paramfile used
   paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60_params.c260303.nc'
   instead of
   paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60-cam70_params.c260305.nc'
   though I had worked on that a few weeks ago and the setting seems correct in namelist_defaults_ctsm.xml...

Erik and Gustavo, feel free to check the rest of my work above, in case you see problems. Thanks!

[gustavo-marques]

@slevis-lmwg, you need to copy the *F90 files in
/glade/p/cesmdata/cseg/runs/cesm2_0/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.315/SourceMods/src.mom/
and then recopile.
Also, please use the following user_nl_mom:
/glade/p/cesmdata/cseg/runs/cesm2_0/b.e30_alpha08b.B1850C_LTso.ne30_t232_wgx3.315/user_nl_mom

[gustavo-marques]

There is also a SourceMods for cice
SourceMods/src.cice:
icepack_wavefracspec_alt.F90

[ekluzek]

I got this to start after a few failed attempts. A few concerns to raise:
3. @ekluzek I found this default paramfile used
paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60_params.c260303.nc'
instead of
paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60-cam70_params.c260305.nc'
though I had worked on that a few weeks ago and the setting seems correct in namelist_defaults_ctsm.xml...

Yes, there's a bug in the picking of the parameter file. So setting it explicitly in the user_nl_clm as you have done is the right thing to do.

Here's the bug it's a simple fix...

ESCOMP/CTSM#3799

Something to add:

In looking over the changes it looks like you are missing this from #315:

stream_fldfilename_ndep = '/glade/derecho/scratch/tilmes/archive/b.e30_alpha07c_cesm.B1850C_MTt4s.ne30_t232_wgx3.251.001/atm/proc/fndep_clm_B1850C_MT4s.1850monthly.ne30_c251209.nc'
stream_year_first_ndep = 11

stream_year_last_ndep = 11

Something to remove:

You are also setting stream_fldfilename_urbantv and stream_fldfilename_prigentroughness which you don't really need to. The new datasets are b4b with the old ones you are pointing to, and the new ones are needed to work with the new ifx compiler which we will be moving to soon. So you might as well use the new datasets.

Everything else looks fine though.

[slevis-lmwg]

Thank you @gustavo-marques @ekluzek for your input.

@ekluzek I tried adding the _ndep settings and got this error, so I think these are not needed anymore:
When ndep isn't required or coming from the CPL none of: stream_year_first_ndep,stream_year_last_ndep, model_year_align_ndep, ndep_taxmod, ndep_varlist, nor stream_fldfilename_ndep can be set!
Pls let me know if I'm mistaken.

[ekluzek]

Thank you @gustavo-marques @ekluzek for your input.

@ekluzek I tried adding the _ndep settings and got this error, so I think these are not needed anymore: When ndep isn't required or coming from the CPL none of: stream_year_first_ndep,stream_year_last_ndep, model_year_align_ndep, ndep_taxmod, ndep_varlist, nor stream_fldfilename_ndep can be set! Pls let me know if I'm mistaken.

Ahhh, yes that's correct. When coupled to CAM we get whatever CAM is sending. And the user_nl_cam is sending ndep from the same case, that I pointed to for CLM (but an updated dataset). #315 had it set in CTSM also, because it wasn't hooked up correctly. But, now that we have it hooked up correctly you don't need that, and as you point out it tells you "you can't set that there". So the model is telling the correct thing to do -- this is good!