Is PTS_MODE working in CESM 2.1?

[ayhwong]

I tried to follow 1.6.2. Running a single point using global data - PTS_MODE — ctsm release-clm5.0 documentation to set up 1-pt CLM5.0 run at CESM 2.1. Yet create_newcase script cannot read --pts_lon and --pts_lat. Is this option is not working, could anyone point me to a guide of setting up a PTS_MODE run?

 

[erik]

Support for settings PTS_LAT/PTS_LON from create_newcase was dropped at some point in cime. You can still use it and we do test for it in the release code. But, what we do is set

PTS_MODE and PTS_LAT, PTS_LON after the case is created. A sample test case that we ran for the release code is here on cheyenne...


/glade/p/cgd/tss/ctsm_baselines/release-clm5.0.34/SMS_Ld1_Mmpi-serial.f45_f45_mg37.I2000Clm50SpGs.cheyenne_gnu.clm-ptsRLA/

 

[ayhwong]

I have follow https://www.cgd.ucar.edu/events/2019/ctsm/files/practical4-wieder.pdf to set up the 1-pt run. Here are my change in the namelist:

./xmlchange GET_REFCASE=FALSE
./xmlchange RUN_TYPE=startup

# for 1pt
./xmlchange CLM_FORCE_COLDSTART=on
./xmlchange COST_PES=36
./xmlchange TOTALPES=1
./xmlchange NTASKS=1
./xmlchange NTASKS_PER_INST=1
./xmlchange ROOTPE=0
./xmlchange JOB_QUEUE=share
./xmlchange MPILIB=mpi-serial
./xmlchange JOB_WALLCLOCK_TIME=1:00:00

./xmlchange ATM_DOMAIN_FILE="domain.lnd.fv0.9x1.25_gx1v7_145_64.55.151020.nc"
./xmlchange ATM_DOMAIN_PATH="/glade/work/ayhwong/single_point"
./xmlchange LND_DOMAIN_FILE="domain.lnd.fv0.9x1.25_gx1v7_145_64.55.151020.nc"
./xmlchange LND_DOMAIN_PATH="/glade/work/ayhwong/single_point"

./xmlchange DATM_CLMNCEP_YR_START=1990
./xmlchange DATM_CLMNCEP_YR_END=2014

./xmlchange CALENDAR='NO_LEAP'
./xmlchange RUN_STARTDATE=0001-01-01
./xmlchange START_TOD=00000
./xmlchange RUN_REFDATE=0001-01-01
./xmlchange RUN_REFTOD=00000

./xmlchange EXEROOT=${exeroot}
./xmlchange STOP_OPTION=nyear
./xmlchange STOP_N=10

./xmlchange --append CLM_BLDNML_OPTS=" -drydep"
./xmlchange DEBUG="TRUE"

echo "hist_mfilt = 20" >> user_nl_clm
echo "hist_nhtfrq = -8760" >> user_nl_clm
echo "hist_empty_htapes = .true." >> user_nl_clm
echo "hist_fincl1 = 'TOTECOSYSC', 'TOTECOSYSN', 'TOTSOMC', 'TOTSOMN', 'TOTVEGC', 'TOTVEGN', 'TLAI', 'GPP', 'CPOOL', 'NPP', 'TWS','H2OSNO'" >> user_nl_clm

echo " fsurdat = '/glade/work/ayhwong/single_point/surfdata_0.9x1.25_78pfts_CMIP6_simyr2000_145_64.55_c200616.nc' " >> user_nl_clm
echo " use_soil_emis = .true. " >>$fname
echo "mapalgo = 'nn','nn','nn','nn','nn'" >> user_nl_datm

But the following error happened during runtime:



I have attached the log files for the run. Does anyone know what may have happened? Thanks a lot!

 

[oleson]

Did you do this:

./xmlchange MPILIB=mpi-serial

I see a setting of mpt in your case.

 

[ayhwong]

I double checked that I have included ./xmlchange MPILIB=mpi-serial in my setting

 

[oleson]

Is this your case?

/glade/work/ayhwong/cases/arctic_low_N

I see <entry id="MPILIB" value="mpt"> in your env_build.xml

 

[ayhwong]

Yes that's the run directory. The run is created by /glade/u/home/ayhwong/my_work/create_test_pulse_1_pt.csh

And you are right that I forgot to change "MPILIB" to "mpi-serial" in the previous run. I have just created a new run with "mpi-serial", and here is the new error message:

 

[oleson]

The code base you seem to be using doesn't look like cesm2.1:

/glade/work/ayhwong/pulse_code/clean_code_base/no_weathering/

Is this modified code?