An error when I set "use_soil_mositure_stream=.ture. "

[kwzhang]

Hi everyone,
I would like to investigate the role of soil moisture in the influence of the Atlantic Multidecadal Variability (AMV) on the Eurasian climate. Therefore, I would like to design an experiment with prescribed postive-AMO SST pattern in Northern Atlantic and climatology of soil moisture over the land. The experiment will run for 100 years in F2000climo in CESM2. The resoluton is set as f09_f09_mg17. The experiment is just similar to the LS3MIP prescribed soil moisture simulation in CMIP6 (f.e21.FHIST_SSP126.f09_f09_mg17.CMIP6-lfmip-pdLC.001), except I also set the prescribed postive-AMO SST pattern in Northern Atlantic.
Therefore, I downloaded LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc from the NCAR website. I think this file is used by the LS3MIP prescribed soil moisture simulation which is just similar to my experiment? Before buiding the case, I set the prescribed postive-AMO SST pattern in Northern Atlantic and also set "use_soil_moisture_streams = .true. ". The settings in the file user_nl_clm is shown below:

use_soil_moisture_streams = .true.
hist_empty_htapes = .true.
hist_avgflag_pertape = 'A','A'
hist_nhtfrq = 0,-24
hist_mfilt = 1,1
hist_fincl1 = 'FSNO','FSM','QSOIL','QVEGE','QVEGT','TSKIN','TSOI','SOILICE','SOILLIQ','SOILWATER_10CM','RH2M','TREFMNAV','TREFMXAV','TLAI','PCT_NAT_PFT'
hist_fincl2 = 'FSNO','FSM','QSOIL','QVEGE','QVEGT','TSKIN','TSOI','SOILICE','SOILLIQ','SOILWATER_10CM','RH2M','TREFMNAV','TREFMXAV','TLAI','PCT_NAT_PFT'
The settings in the file lnd_in is shown below:

The settings about the soil_moisture_streams in the file lnd_in is shown below:

soilm_offset = 0
soilm_tintalgo = 'linear'
stream_fldfilename_soilm = '/BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc'
stream_year_first_soilm = 1997
stream_year_last_soilm = 1997

After case submitting, I got the following error in the file lnd.log.7104048.231027-181639:

(shr_dmodel_readstrm) open : /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc
(shr_dmodel_readstrm) file lb: /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc 12
(shr_dmodel_readstrm) file ub: /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc 1
Input soil moisture dataset is not vegetated as expected: gridcell= 1
active = T
ENDRUN:
ERROR:
('PrescribedSoilMoistureInterp') ERROR:: The input soil moisture stream is NOT
vegetated for one of the land points

Any hints would be helpful! Thanks in advance!
Best,
Kaiwen

 

[oleson]

Yes, that is the same file that was used in the LS3MIP prescribed soil moisture simulation.
It is compatible with the surface datasets and code that was used in that LS3MIP simulation. In particular, the surface dataset and landuse timeseries that was used is:
fsurdat = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc'
flanduse_timeseries = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/release-clm5.0.18/landuse.timeseries_0.9x1.25_SSP1-2.6_16pfts_Irrig_CMIP6_simyr1850-2100_c190214.nc'

Maybe you are using something different?

 

[kwzhang]

Yes, that is the same file that was used in the LS3MIP prescribed soil moisture simulation.
It is compatible with the surface datasets and code that was used in that LS3MIP simulation. In particular, the surface dataset and landuse timeseries that was used is:
fsurdat = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc'
flanduse_timeseries = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/release-clm5.0.18/landuse.timeseries_0.9x1.25_SSP1-2.6_16pfts_Irrig_CMIP6_simyr1850-2100_c190214.nc'

Maybe you are using something different?

Dear Oleson,
Thank you so much for your reply. I have checked the the surface dataset and landuse timeseries that was used in the F2000climo. The flanduse_timeseries is not used in this compset. I think this is because the F2000climo is not a transient experiment like FHIST? The surface dataset used in F2000climo is:
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
fsurdat = '/BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/surfdata_map/release-clm5.0.18/surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr2000_c190214.nc'.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
I downloaded the file 'surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc' from the UCAR website. I replaced the file 'surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr2000_c190214.nc' with this new file. However, I get the same error as before:

------------------------------------------------------------
Attempting to read annual vegetation data .....
(GETFIL): attempting to find local file
surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc
(GETFIL): using
/BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/surfdata_map/release-clm5.0.1
8/surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc
Attempting to read monthly vegetation data .....
nstep = 0 month = 1 day = 1
(GETFIL): attempting to find local file
surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc
(GETFIL): using
/BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/surfdata_map/release-clm5.0.1
8/surfdata_0.9x1.25_16pfts_Irrig_CMIP6_simyr1850_c170824.nc
Successfully read monthly vegetation data for
month 1

Successfully initialized the land model
begin initial run at:
nstep= 0 year= 1 month= 1 day=
1 seconds= 0

************************************************************

dtime_sync= 1800 dtime_clm= 1800 mod = 0
(shr_dmodel_readstrm) open : /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc
(shr_dmodel_readstrm) file lb: /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc 12
(shr_dmodel_readstrm) file ub: /BIGDATA2/app/CESM_inputdata_2/inputdata/lnd/clm2/prescribed_data/LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc 1
Input soil moisture dataset is not vegetated as expected: gridcell= 1
active = T
ENDRUN:
ERROR:
('PrescribedSoilMoistureInterp') ERROR:: The input soil moisture stream is NOT
vegetated for one of the land points

I noted that the LS3MIP prescribed soil moisture file 'LFMIP-pdLC-SST.H2OSOI.0.9x1.25.20levsoi.natveg.1980-2014.MONS_climo.c190716.nc' ends in 'c190716', while the surface dataset file ends in 'c170824' and 'c190214'. I think this is when the file was created. Do I need to use the same version of the LS3MIP prescribed soil moisture file? But I only find the version 'c190716' in the UCAR website.

Any hints would be helpful! Thanks in advance!

Best,
Kaiwen

 

[oleson]

Right, F2000climo is not a transient simulation.
The dates on the surface dataset and soil moisture files don't have to agree.
But you probably wouldn't want to use an 1850 surface dataset in a year 2000climo simulation anyway.
I think it's just that your code and compset/case are not similar enough to the original code and case that was used in the f.e21.FHIST_SSP126.f09_f09_mg17.CMIP6-lfmip-pdLC.001 simulation.
I've tarred up the original case directory for the historical part of that simulation and put it on our ftp site:
ftp://ftp.cgd.ucar.edu/pub/oleson/f.e21.FHIST_SSP126.f09_f09_mg17.CMIP6-lfmip-pdLC.001.HIST.tar.gz
Maybe you can figure out how to configure your simulation similarly.
Another thing you can try is to set soilm_ignore_data_if_missing = .true. in your user_nl_clm. That would bypass the error but you'd want to check to see your soil moisture against the prescribed to see how different it is.
Otherwise, I think you'll need to create your own soil moisture file and use the prescribed soil moisture capability in CLM.

 

[kwzhang]

Right, F2000climo is not a transient simulation.
The dates on the surface dataset and soil moisture files don't have to agree.
But you probably wouldn't want to use an 1850 surface dataset in a year 2000climo simulation anyway.
I think it's just that your code and compset/case are not similar enough to the original code and case that was used in the f.e21.FHIST_SSP126.f09_f09_mg17.CMIP6-lfmip-pdLC.001 simulation.
I've tarred up the original case directory for the historical part of that simulation and put it on our ftp site:
ftp://ftp.cgd.ucar.edu/pub/oleson/f.e21.FHIST_SSP126.f09_f09_mg17.CMIP6-lfmip-pdLC.001.HIST.tar.gz
Maybe you can figure out how to configure your simulation similarly.
Another thing you can try is to set soilm_ignore_data_if_missing = .true. in your user_nl_clm. That would bypass the error but you'd want to check to see your soil moisture against the prescribed to see how different it is.
Otherwise, I think you'll need to create your own soil moisture file and use the prescribed soil moisture capability in CLM.

Dear Oleson,
Very thanks for your nice suggestions! :)
Best,
Kaiwen