questions about CLM6.0

[jinmuluo]

Hi,

I recently updated my model to CLM6.0 (ctsm5.2.dev.005), I found many things about the Nitrogen part have changed compared to the CLM5.1 (ctsm5.1.dev.163).
(1) N fertilisers usage has been increased a lot: maybe because the CLM6.0 uses new land cover data maps?
(2) N deposition cycle over 1850 (below) when I created the compset "HIST_DATM%GSWP3v1_CLM60%BGC-CROP_SICE_SOCN_MOSART_SGLC_SWAV" , no sure why this happened on my case.

&ndepdyn_nml
ndep_taxmode = 'cycle'
ndep_varlist = 'NDEP_month'
ndepmapalgo = 'bilinear'
stream_fldfilename_ndep = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/ndepdata/fndep_clm_hist_b.e21.BWHIST.f09_g17.CMIP6-historical-WACCM.ensmean_1849-2015_monthly_0.9x1.25_c180926.nc'
stream_meshfile_ndep = '/glade/campaign/cesm/cesmdata/inputdata/share/meshes/fv0.9x1.25_141008_polemod_ESMFmesh.nc'
stream_year_first_ndep = 1850
stream_year_last_ndep = 1850

Appreciate any comments on the questions mentioned above!

Best,

Jinmu

 

[oleson]

For your first question, what history variable are you looking at to determine N fertilizer usage? Maybe I can look at some simulations that have been done with CLM5 and CTSM52.
For your second question, I setup a case as follows with ctsm5.2.005 like this:

./create_newcase --case ctsm52d005_test_historical --res f09_g17 --compset HIST_DATM%GSWP3v1_CLM60%BGC-CROP_SICE_SOCN_MOSART_SGLC_SWAV --run-unsupported

and got this ndep stream:

&ndepdyn_nml
model_year_align_ndep = 1850
ndep_taxmode = 'cycle'
ndep_varlist = 'NDEP_month'
ndepmapalgo = 'bilinear'
stream_fldfilename_ndep = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/ndepdata/fndep_clm_hist_b.e21.BWHIST.f09_g17.CMIP6-historical-WACCM.ensmean_1849-2015_monthly_0.9x1.25_c180926.nc'
stream_meshfile_ndep = '/glade/campaign/cesm/cesmdata/inputdata/share/meshes/fv0.9x1.25_141008_polemod_ESMFmesh.nc'
stream_year_first_ndep = 1850
stream_year_last_ndep = 2015
/

which seems to be correct.

 

[jinmuluo]

For your first question, what history variable are you looking at to determine N fertilizer usage? Maybe I can look at some simulations that have been done with CLM5 and CTSM52.
For your second question, I setup a case as follows with ctsm5.2.005 like this:

./create_newcase --case ctsm52d005_test_historical --res f09_g17 --compset HIST_DATM%GSWP3v1_CLM60%BGC-CROP_SICE_SOCN_MOSART_SGLC_SWAV --run-unsupported

and got this ndep stream:

&ndepdyn_nml
model_year_align_ndep = 1850
ndep_taxmode = 'cycle'
ndep_varlist = 'NDEP_month'
ndepmapalgo = 'bilinear'
stream_fldfilename_ndep = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/ndepdata/fndep_clm_hist_b.e21.BWHIST.f09_g17.CMIP6-historical-WACCM.ensmean_1849-2015_monthly_0.9x1.25_c180926.nc'
stream_meshfile_ndep = '/glade/campaign/cesm/cesmdata/inputdata/share/meshes/fv0.9x1.25_141008_polemod_ESMFmesh.nc'
stream_year_first_ndep = 1850
stream_year_last_ndep = 2015
/

which seems to be correct.

Hi,

I take "FERT_TO_SMINN" as the fertilisation rate to soil and I guess it includes two different parts, synthetic fertilisers and fixed manure application rate.

 

[jinmuluo]

For your first question, what history variable are you looking at to determine N fertilizer usage? Maybe I can look at some simulations that have been done with CLM5 and CTSM52.
For your second question, I setup a case as follows with ctsm5.2.005 like this:

./create_newcase --case ctsm52d005_test_historical --res f09_g17 --compset HIST_DATM%GSWP3v1_CLM60%BGC-CROP_SICE_SOCN_MOSART_SGLC_SWAV --run-unsupported

and got this ndep stream:

&ndepdyn_nml
model_year_align_ndep = 1850
ndep_taxmode = 'cycle'
ndep_varlist = 'NDEP_month'
ndepmapalgo = 'bilinear'
stream_fldfilename_ndep = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/ndepdata/fndep_clm_hist_b.e21.BWHIST.f09_g17.CMIP6-historical-WACCM.ensmean_1849-2015_monthly_0.9x1.25_c180926.nc'
stream_meshfile_ndep = '/glade/campaign/cesm/cesmdata/inputdata/share/meshes/fv0.9x1.25_141008_polemod_ESMFmesh.nc'
stream_year_first_ndep = 1850
stream_year_last_ndep = 2015
/

which seems to be correct.

and I have tried several times, my case still cycles 1850. I can easily change the stream year in CaseDocs but I do want to solve this question. Could you please tell me where I can start to detect this problem? Thanks.

In case of if you need to take a look at the case.
/glade/work/jinmuluo/cases/VerTranGlobal/CaseDocs

 

[oleson]

It doesn't look like you are actually using ctsm5.2.005, you are using ctsm5.2.004. I see this when I do a git describe in /glade/u/home/jinmuluo/CLM6.0-FANv3

ctsm5.2.004-219-ge058e201b

There were some bugs in that tag associated with CLM6 compsets.
Try updated your code to ctsm5.2.005.

 

[oleson]

CLM6 does use new surface datasets. Do you see big differences in historical simulations?

 

[oleson]

This is a comparison of new versus old datasets. It's a bit old but I don't think things have changed that much since then.

https://webext.cgd.ucar.edu/I20TR/ihist_bgccrop/lnd/ihist_bgccrop.1995_2014-ihist_bgccrop_old_fsurdat.1995_2014/set1/set1_NFERTILIZATION.png

 

[jinmuluo]

This is a comparison of new versus old datasets. It's a bit old but I don't think things have changed that much since then.

https://webext.cgd.ucar.edu/I20TR/ihist_bgccrop/lnd/ihist_bgccrop.1995_2014-ihist_bgccrop_old_fsurdat.1995_2014/set1/set1_NFERTILIZATION.png

Thank you Keith,

Does this number include synthetic fertilizer and manure usage?

Best,

 

[oleson]

It is the NFERTILIZATION history variable. It is reporting the fert_patch variable in the code. This seems to be derived in biogeochem/CNPhenologyMod.F90 as

fert => cnveg_nitrogenflux_inst%fert_patch

fert(p) = (manunitro(ivt(p)) * 1000._r8 + fertnitro(p))/ fert_counter(p)

So I think it includes both.