Initial WIEMIP usermods

[wwieder]

Initial draft of usermods I'm using in WIEMIP spinups (currently for AD, only)

[wwieder]

@lawrencepj1 can you suggest additional history output we need for WIEMIP simulations?

@swensosc the list of variables I have for hillslope output is below. Does this seem appropriate?

!---------------------
! History output
!---------------------

hist_nhtfrq = 0,0,0
hist_mfilt = 1,1,1
hist_dov2xy = .true., .false., .false.
hist_type1d_pertape = 'GRID','COLS','PFTS'

hist_fincl1 = 'STREAM_WATER_DEPTH','STREAM_WATER_VOLUME','FROST_TABLE',
   'CWDC_HR','LIT_MET_HR','LIT_CEL_HR','LIT_LIG_HR',
   'SOM_ACT_HR_S2','SOM_ACT_HR_S3','SOM_SLO_HR_S1',
   'SOM_SLO_HR_S3','SOM_PAS_HR','FCO2','FROST_TABLE',
   'LIT_MET_N_TO_SOM_ACT_N','LIT_CEL_N_TO_SOM_ACT_N',
   'LIT_LIG_N_TO_SOM_SLO','TAF','LITFIRE',
   'M_DEADROOTC_STORAGE_TO_FIRE','M_DEADROOTC_TO_FIRE',
   'M_DEADROOTC_XFER_TO_FIRE','M_DEADSTEMC_STORAGE_TO_FIRE',
   'M_DEADSTEMC_TO_FIRE','M_DEADSTEMC_XFER_TO_FIRE',
   'M_FROOTC_STORAGE_TO_FIRE','M_FROOTC_TO_FIRE',
   'M_FROOTC_XFER_TO_FIRE','M_LEAFC_STORAGE_TO_FIRE',
   'M_LEAFC_TO_FIRE','M_LEAFC_XFER_TO_FIRE',
   'M_LIVEROOTC_STORAGE_TO_FIRE','M_LIVEROOTC_TO_FIRE',
   'M_LIVEROOTC_XFER_TO_FIRE','M_LIVESTEMC_STORAGE_TO_FIRE',
   'M_LIVESTEMC_TO_FIRE','M_LIVESTEMC_XFER_TO_FIRE',
   'M_GRESP_STORAGE_TO_FIRE','M_GRESP_XFER_TO_FIRE'

hist_fincl2 = 'H2OSFC','H2OSNO','ZWT','ZWT_PERCH','FROST_TABLE',
   'TWS','SOILLIQ','SOILICE','FSNO','FH2OSFC','ALT','QLATFLOWOUT',
   'QDRAI','QOVER','QVEGT','QVEGE','QSOIL','TSOI','FINUNDATED',
   'CH4_SURF_DIFF_SAT','CH4_SURF_DIFF_UNSAT','CH4_EBUL_TOTAL_SAT',
   'CH4_EBUL_TOTAL_UNSAT','CH4_SURF_EBUL_SAT','CH4_SURF_EBUL_UNSAT',
   'CH4_SURF_AERE_SAT','CH4_SURF_AERE_UNSAT','TOTCOLCH4','CONC_CH4_SAT',
   'CONC_CH4_UNSAT','CH4_PROD_COL','CH4_OXID_COL',
   'CH4_OXID_DEPTH_SAT','CH4_OXID_DEPTH_UNSAT',
   'CH4_PROD_DEPTH_SAT','CH4_PROD_DEPTH_UNSAT',
   'SNOWDP', 'SNOW_DEPTH','ALTMAX','BTRANMN'

hist_fincl3 = 'FCTR','FGEV','FROOTC_ALLOC','FSH',
   'FSRND','FSRNI','FSRVD','FSRVI','FSDSND',
   'FSDSNI','FSDSVD','FSDSVI',
   'GPP','HTOP','LEAFC_ALLOC','NPP','BTRANMN',
   'MEG_acetone','MEG_isoprene','MEG_methanol',
   'NFIX','FFIX_TO_SMINN','PFT_FIRE_CLOSS',
   'Rnet','TOTVEGC','TOTVEGN','WOODC_ALLOC','TLAI'

[swensosc]

you could add 'CH4_PROD_COL','CH4_OXID_COL', which are new history outputs I recently added to look at production and oxidation at the column level (the original outputs are gridcell only variables)

…

On Tue, Apr 28, 2026 at 8:59 AM will wieder ***@***.***> wrote: *wwieder* left a comment (ESCOMP/CTSM#3950) <#3950 (comment)> @lawrencepj1 <https://github.com/lawrencepj1> can you suggest additional history output we need for WIEMIP simulations? @swensosc <https://github.com/swensosc> the list of variables I have for hillslope output is below. Does this seem appropriate? !!! extra history from Sean's wiemip test case hist_nhtfrq = 0,0 hist_mfilt = 1,1 hist_dov2xy = .true., .false. ! average to column level for vector output hist_type1d_pertape = 'GRID','COLS' hist_fincl1 = 'STREAM_WATER_DEPTH','STREAM_WATER_VOLUME','FROST_TABLE' hist_fincl2 = 'H2OSFC','H2OSNO','ZWT','ZWT_PERCH','FROST_TABLE', 'TWS','SOILLIQ','SOILICE','FSNO','FH2OSFC','ALT','QFLOOD','QLATFLOWOUT', 'QDRAI','QOVER','QVEGT','QVEGE','QSOIL','TSOI','FINUNDATED', 'CH4_SURF_DIFF_SAT','CH4_SURF_DIFF_UNSAT','CH4_EBUL_TOTAL_SAT', 'CH4_EBUL_TOTAL_UNSAT','CH4_SURF_EBUL_SAT','CH4_SURF_EBUL_UNSAT', 'CH4_SURF_AERE_SAT','CH4_SURF_AERE_UNSAT','TOTCOLCH4','CONC_CH4_SAT', 'CONC_CH4_UNSAT' — Reply to this email directly, view it on GitHub <#3950 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGRN57CSFZBDA47EFBKWBST4YDBOPAVCNFSM6AAAAACYDFJCSSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DGMZWGQ3TCNRSGM> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[lawrencepj1]

Hi @wwieder and @swensosc

This is directly from Adrianna's user_nl_clm files used for TRENDY 2025 S3 (full transient) simulation:

hist_mfilt = 1,1,1
hist_nhtfrq = 0,0,-24

! monthly global gridded
hist_fincl1 = 'AR', 'COL_FIRE_CLOSS', 'CROPPROD1C_LOSS', 'CWD_C', 'CWD_C_TO_LIT_CEL_C',
'CWD_C_TO_LIT_LIG_C', 'DEADCROOTC', 'DEADCROOTC_STORAGE', 'DEADCROOTC_XFER',
'DEADCROOTN_STORAGE', 'DEADCROOTN_XFER', 'DEADSTEMN_STORAGE', 'DEADSTEMN_XFER',
'LIVECROOTN_STORAGE', 'LIVECROOTN_XFER', 'LIVESTEMN_STORAGE', 'LIVESTEMN_XFER',
'DEADCROOTN', 'DEADSTEMN', 'DENIT', 'DWT_CONV_CFLUX', 'F_N2O_DENIT',
'F_N2O_NIT', 'FAREA_BURNED', 'FFIX_TO_SMINN', 'FROOTC', 'FROOTC_STORAGE',
'FROOTC_TO_LITTER', 'FROOTC_XFER', 'FROOTN', 'FSDS', 'GPP', 'HR', 'H2OSNO',
'FSR', 'FGEV', 'FCTR', 'FSRNI', 'FSRVI', 'FSDSNI', 'FSDSVI', 'FSH',
'FLDS', 'FIRE',
'LEAFC', 'LEAFC_STORAGE', 'LEAFC_TO_LITTER', 'LEAFC_XFER', 'LEAFN',
'LEAFN_STORAGE', 'LEAFN_XFER', 'LIT_CEL_C_TO_SOM_ACT_C',
'LIT_LIG_C_TO_SOM_SLO_C', 'LIT_MET_C_TO_SOM_ACT_C', 'LITFALL', 'LITFIRE',
'LIVECROOTC', 'LIVECROOTC_STORAGE', 'LIVECROOTC_XFER', 'LIVECROOTN',
'LIVESTEMN', 'M_DEADROOTC_STORAGE_TO_FIRE', 'M_DEADROOTC_TO_FIRE',
'M_DEADROOTC_XFER_TO_FIRE', 'M_DEADSTEMC_STORAGE_TO_FIRE',
'M_DEADSTEMC_TO_FIRE', 'M_DEADSTEMC_XFER_TO_FIRE',
'M_FROOTC_STORAGE_TO_FIRE', 'M_FROOTC_TO_FIRE', 'M_FROOTC_XFER_TO_FIRE',
'M_GRESP_STORAGE_TO_FIRE', 'M_GRESP_XFER_TO_FIRE', 'M_LEAFC_STORAGE_TO_FIRE',
'M_LEAFC_TO_FIRE', 'M_LEAFC_XFER_TO_FIRE', 'M_LIVEROOTC_STORAGE_TO_FIRE',
'M_LIVEROOTC_TO_FIRE', 'M_LIVEROOTC_XFER_TO_FIRE',
'M_LIVESTEMC_STORAGE_TO_FIRE', 'M_LIVESTEMC_TO_FIRE',
'M_LIVESTEMC_XFER_TO_FIRE', 'NBP', 'NDEP_TO_SMINN', 'NET_NMIN', 'NFIX',
'NPP', 'QRUNOFF', 'QSOIL', 'QVEGE', 'QVEGT', 'RAIN', 'RR', 'QIRRIG_DRIP',
'SMIN_NO3_LEACHED', 'SMIN_NO3_RUNOFF', 'SMINN', 'SMINN_TO_PLANT', 'SNOW',
'SOILICE', 'SOILLIQ', 'SOM_ACT_C', 'SOM_ACT_C_1m', 'SOM_ACT_C_TO_SOM_PAS_C',
'SOM_ACT_C_TO_SOM_SLO_C', 'SOM_PAS_C', 'SOM_PAS_C_1m',
'SOM_PAS_C_TO_SOM_ACT_C', 'SOM_SLO_C', 'SOM_SLO_C_1m',
'SOM_SLO_C_TO_SOM_ACT_C', 'SOM_SLO_C_TO_SOM_PAS_C', 'SOMC_FIRE', 'SOMFIRE',
'TLAI', 'TOT_WOODPRODC', 'TOT_WOODPRODC_LOSS', 'TOT_WOODPRODN', 'TOTLITC',
'TOTLITN', 'TOTSOILICE', 'TOTSOILLIQ', 'TOTSOMC', 'TOTSOMN', 'TOTVEGC',
'TOTVEGN', 'TSA', 'TSOI', 'WOODC',
'FROOTN_STORAGE', 'FROOTN_XFER', "FSRND", "FSRVD", "FSDSND", "FSDSVD"

! monthly pft
hist_fincl2 = 'AGNPP', 'FCEV', 'FCTR', 'FGEV', 'FIRE', 'FLDS', 'FROOTC_ALLOC', 'FSDS', 'FSDSND',
'FSDSNI', 'FSDSVD', 'FSDSVI', 'FSH', 'FSR', 'FSRND', 'FSRNI', 'FSRVD',
'FSRVI', 'GPP', 'HR', 'HTOP', 'LEAFC_ALLOC', 'NPP', 'QIRRIG_DRIP', 'TLAI',
'TOTVEGC', 'TV', 'WOODC_ALLOC', 'GRAINC_TO_FOOD', 'QVEGE', 'QVEGT', 'QSOIL',
'Rnet'

! daily global gridded
hist_fincl3 = 'TLAI'

hist_dov2xy = .true., .false., .true.

[lawrencepj1]

@wwieder

Can you point me to the default history files being produced in your flat spin up?

Thanks
Peter

[wwieder]

@lawrencepj1, sample default h0 files available are here:
/glade/derecho/scratch/wwieder/archive/wetlands.n03.ctsm5.4.029_f09noCrop_HH_165_testCTRL

[wwieder]

@slevis-lmwg noted in this comment. I'm migrating this thread to the CTSM-side PR here.

@wwieder I noticed the following (may or may not be of concern):

1. In BGC and RAD, should co2 align be 1850 and last be 1869 or 2000?

co2tseries.cmip6_20tr:year_first=1850
co2tseries.cmip6_20tr:year_last=1850
co2tseries.cmip6_20tr:year_align=1

2. I have one or two more comments but glade has frozen on me... Adding now:
   Two paramfiles and two hillslope files appear in user_nl_clm. Which one of each is correct?

3. We may have already discussed the topic of transient vs. fixed ndep, and we decided that fixed made more sense? In any case, that's what you're using.

[wwieder]

Thanks for looking at these, Sam.

I'm constantly baffled by year_align. Can you help me on this?

CO2 year_last should be:

• 2000 for BGC (where we cycle climate 1850-1869) and
• 1850 for RAD (constant CO2, but time evolving climate).

[olyson]

@wwieder I think you had a question about the urbantv streams (not sure where the question was but I'm posting here). Currently, the urbantv streams are just a way to turn off air conditioning (AC) in the 1850-1949 part of the transient. For 1950-2100 AC is on but the values of the variables on the file are the same for every year.
For an 1850 spinup the settings should show up explicitly in the namelist as:

stream_year_first_urbantv = 1850
stream_year_last_urbantv = 1850

AC will thus be off.
For a historical the settings should show up explicitly in the namelist as:

model_year_align_urbantv = 1850
stream_year_first_urbantv = 1850
stream_year_last_urbantv = 2106

Happy to talk about what settings might be appropriate for your particular simulations.

[slevis-lmwg]

@wwieder I added a couple more comments to my post above for you to look at.

Regarding year_align, it confused me, too, for a while. Now I think it's simpler than it appears:

• year_first and last give the range of stream data to use in the simulation
• year_align represents the simulation year that you want aligned with the data in year_first, and it really just matters most in transient cases; examples:

1. In transient simulations from 1850 to 1920 where we cycle stream data from 1901 to 1920, the most intuitive year_align = 1901 because I want the same data year as model year. For the sake of argument, you could less intuitively offset year_align by 20 yrs (1881, 1921, etc.), and you would have the same outcome.
2. In 1850 simulations that start with model year 1 and cycle stream data from 1901 to 1920, year_align doesn't really matter. If you set it to 1, then the model will align the first year of stream data (1901) with the first year of the simulation (1).
3. A workflow that would likely still confuse me is Peter Thornton's. His logic makes sense, but I do not think that we do this anymore. Peter wanted the stream data at the end of AD to line up with the stream data at the beginning of postAD and similarly the stream data at the end of postAD to line up with the stream data at the beginning of the next simulation, and so forth. In other words if AD ended with stream data year 1907, then I should make sure to start postAD with stream data year 1908. I sometimes spent hours trying to figure out how things would line up correctly...

[wwieder]

@lawrencepj1, sorry to nag, but have you made any progress on going through the WIEMIP data request to make sure we're writing out the history needed form the experiments? (if it's easier, see also #3949)

[slevis-lmwg]

From meeting with @lawrencepj1 and @ekluzek
Based on the time this will take, Peter expects to resolve this by next week.

Notes:

• WIEMIP variable request
• https://github.com/adrifoster/trendyclm and doc with corresponding instructions

[rosiealice]

What are y'all thinking we should report for 'surface wind' here? I found U10 - wind at 10m, which is the standard output, and U_S_MEAN, which is in the dust model and is the wind at 0.1m above the surface. The latter seems to correspond more to the actual surface wind but I don't know if it has any dust model calibration terms baked into it? Can you advise, @dmleung ? :)

CTSM/src/biogeochem/DustEmisLeung2023.F90

Line 716 in 7731d83

u_mean_slt(p) = (wnd_frc_slt/k) * log(hgt_sal / z0a_glob) ! translating from ustar (velocity scale) to actual wind

tagging @JessicaNeedham too...

[rosiealice]

On a similar note, what are you using for:

• frozen layer thickness? ( "Freeze depth; depth to zero centigrade isotherm in non-permafrost"). I wondered if this was FROST_TABLE ?

• Total soil wetness (Vertically integrated soil moisture divided by maximum allowable soil moisture above wilting point) . This seems like a sketchy version of BTRAN, but what is 'allowable' here? Is it the maximum possible SWC or field capacity? I don't think that moisture above field capacity is super interesting functionally?

Fractional Ocean Cover? I guess this is probably a surface dataset variable?

FWIW, here is the table where we are filling in the output fields:
https://docs.google.com/spreadsheets/d/1MG_x52Uf5T9TSANtw34VqDi2WBr-57-gZrEJaBzl_ok/edit?usp=sharing
The highlighting scheme is:
Blue: Things I am still unsure of.
Orange: Things CLM-FATES could output in theory but can't without some code mods
Green: things I think we can output on the CLM PFT machinery (but aren't doing yet by default).

If you see anything odd then please say!

[dmleung]

What are y'all thinking we should report for 'surface wind' here? I found U10 - wind at 10m, which is the standard output, and U_S_MEAN, which is in the dust model and is the wind at 0.1m above the surface. The latter seems to correspond more to the actual surface wind but I don't know if it has any dust model calibration terms baked into it? Can you advise, @dmleung ? :)

CTSM/src/biogeochem/DustEmisLeung2023.F90

Line 716 in 7731d83

u_mean_slt(p) = (wnd_frc_slt/k) * log(hgt_sal / z0a_glob) ! translating from ustar (velocity scale) to actual wind

tagging @JessicaNeedham too...

Hi @rosiealice :) yes this is the 0.1 m wind speed calculated from friction velocity fv. The z0a_glob term you see is the soil roughness, assumed to be a constant of 10^-4 m for dust emission modeling. Apart from this there is no other dust model calibration in this line. Does this help?

[ekluzek]

@wwieder I don't want to mess up what you are doing here. Especially because the basic purpose is for you to have something that preserves how the WEIMIP simulations are done. And much of how that's done is going to be temporary as it's unique to WIEMIP. But, I would like to peel out of this any parts that we do want to preserve outside of WIEMIP. And I get the impression that there is some of that hidden in here.

There's also a lot of duplication that could be handled differently for maintainability. And I'm pretty sure there are things you are setting to be explicit that really don't need to be set because they are determined by the compset or something else. Because this is pretty large that duplication likely means it's harder for you to maintain. So it might be worth working on it, even if it only helps you.

There's a bunch of files that are on scratch, which might be OK for the forcing data since it's fairly large. But, I think everything else we should have in inputdata and rimported so that you could at least run something similar (but with different forcing) elsewhere.

My impression is that for HillSlope compsets, these settings need to be done:

use_hillslope = .true.
use_hillslope_routing = .true.
finundation_method = 'h2osfc'
hillslope_fsat_equals_zero = .true.

and then there are surface datasets that are picked for Hillslope. We should have those preserved in the regular mechanism for when a Hillslope compset is chosen you get it out of the box for at least f09. Are the WIEMIP versions going to be used for the final CESM3.0 Hillslope simualtions or something else?

This is long enough that we should meet and talk it over. Thanks.

[wwieder]

@ekluzek, thanks for these thoughts.

I wonder if your question actually pertains to this draft PR, or something else? My reading of your question is that you're wondering what should go into a supported hillslope compset? If that's true, your impression seems correct:

use_hillslope = .true.
use_hillslope_routing = .true.
finundation_method = 'h2osfc'
hillslope_fsat_equals_zero = .true.

Plus, I'll add this to the list.

paramfile = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/paramdata/ctsm60-HH_params.c260518.nc'

! 16 column hillslope and vegetation data
!hillslope_file = '/glade/derecho/scratch/swensosc/WIEMIP/hillslope_geo_params_trapezoid_worldcover_0.9x1.25_Global_FAB.nc'

! 9 column hillslope file
hillslope_file = '/glade/work/swensosc/HillslopeHydrology/InputData/Global/hillslope_geo_params_3_col_3_aspect_trapezoid_worldcover_0.9x1.25_Global_FAB.nc'

n_dom_pfts = 2

I'll also note that the hillslope file and number of dominant PFTs are somewhat subjective, but we should have a default compset defined that we want to support. Irrigation options and history ouptup are also relevant to the compset topic.