How to continue running final spin up

[Li Ming]

Hello,@oleson . I am using release-clm5.0.35 to complete the spin up process for the region, following the instruction document 1.5.5. Spinup of CLM5.0-BGC-Crop — ctsm release-clm5.0 documentation (escomp.github.io)，I first did a 400 years AD-spin up with perfect results, then I planned to do a 600 year final spin up, but the program terminated unexpectedly at 500 years and no r-file was generated in the result. Is there a way for me to continue the final spin up for the last 100 years? If there is a way, how do I need to set it up? Below is the setup I have during the execution of final spin up.

./create_newcase --case TP-final-spin-up --res CLM_USRDAT --compset 2000_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_SROF_SGLC_SWAV --compiler intel --mach myintel --run-unsupported

cp /gpfs/home/limingwang/cesm/scratch/TP-AD-spin-up/run/TP-AD-spin-up.clm2.r* .
cp /gpfs/home/limingwang/cesm/scratch/TP-AD-spin-up/run/rpointer.* .

<entry id="DIN_LOC_ROOT_CLMFORC" value="/gpfs/home/limingwang/cesm/CMFD/Tibet-result">
<entry id="ATM_DOMAIN_FILE" value="domain.lnd.qzgy-rqj_noocean.nc.230604.nc">
<entry id="ATM_DOMAIN_PATH" value="/gpfs/home/limingwang/cesm/inputdata/wodewenjian/domain">
<entry id="LND_DOMAIN_FILE" value="domain.lnd.qzgy-rqj_noocean.nc.230604.nc">
<entry id="LND_DOMAIN_PATH" value="/gpfs/home/limingwang/cesm/inputdata/wodewenjian/domain">
<entry id="DATM_CLMNCEP_YR_START" value="1979">
<entry id="DATM_CLMNCEP_YR_END" value="1998">
<entry id="STOP_OPTION" value="nyears">
<entry id="STOP_N" value="600">

use_init_interp = .true
finidat ='/gpfs/home/limingwang/clm5new/cime/scripts/TP-final-spin-up/TP-AD-spin-up.clm2.r.0401-01-01-00000.nc'
fsurdat ='/gpfs/home/limingwang/cesm/inputdata/wodewenjian/surface/qzgy-rqj/surfdata_qzgy-rqj_hist_16pfts_Irrig_CMIP6_simyr2000_c231031.nc'

I have another question, if I can't continue to complete the final spin up, can I use the result file of the 400-year AD-spin up (e.g. TP-AD-spin-up.clm2.r.0401-01-01-00000.nc) as the initial file of the model? The problem I am researching is the historical changes in NPP & NEP

Thanks for your reply.

 

[oleson]

If the model ran continuously for 600 years (without restarting), and crashed before the end of the 600 years, then no r-file will be generated and there won't be a way to redo the spinup for the last 100 years.
In the second part of the spinup, it looks like you are using the initial file from the end of the AD spinup, which is proper. However, the second part of the spinup should be run for several hundred years to get the carbon pools into equilibrium. Much of that time is spent getting the soil carbon pools to equilibrate.

 

[Li Ming]

Thank you for your reply, I reran the 200 year final spin up, and it produced the result file and r file without any problem, but I would like to continue to run the 200 year final spin up and generate an r file every 5 years, so I have made the following changes to env_run, am I making the correct changes? The model run looks fine.

result files：
TP-final-spin-up.clm2.r.0201-01-01-00000.nc
TP-final-spin-up.clm2.rh0.0201-01-01-00000.nc
TP-final-spin-up.cpl.r.0201-01-01-00000.nc
TP-final-spin-up.datm.rs1.0201-01-01-00000.bin

Env_run setting:
<entry id="CONTINUE_RUN" value="TRUE">
<entry id="REST_OPTION" value="nyears">
<entry id="REST_N" value="5">

Thanks

 

[slevis]

I think your changes seem fine.