SSP370 compset with minimal temperature changes

[Yuan Sun]

Hi, all,

I ran land-only ssp3-7.0 composet based on 15-year spin-up. but I found the outputs changes moderately.

./create_newcase --case $HOME/scratch/Projects/scratch/20231203Case1 --compset SSP370_DATM%GSWP3v1_CLM50%SP_SICE_SOCN_MOSART_CISM2%NOEVOLVE_SWAV --res f09_g17 --machine csf3 --run-unsupported

./xmlchange RUN_STARTDATE=2015-01-01
./xmlchange NTASKS=512
./xmlchange STOP_OPTION=nyears
./xmlchange STOP_N=10
./xmlchange RESUBMIT=8
./xmlchange RUN_REFCASE=20231126SpinUp
./xmlchange RUN_REFDATE=2015-01-01
./xmlchange RUN_TYPE=hybrid
./xmlchange GET_REFCASE=FALSE
./xmlchange JOB_WALLCLOCK_TIME=96:00:00
./xmlchange MAX_MPITASKS_PER_NODE=32
./xmlchange MAX_TASKS_PER_NODE=32

vim user_nl_clm
fsurdat = '/mnt/iusers01/fatpou01/sees01/a16404ys/scratch/Projects/inputdata/project1/surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr1850_c231120_urbanalbedo0.9.nc'
flanduse_timeseries = '/mnt/iusers01/fatpou01/sees01/a16404ys/scratch/Projects/inputdata/lnd/clm2/surfdata_map/release-clm5.0.18/landuse.timeseries_0.9x1.25_SSP3-7.0_16pfts_Irrig_CMIP6_simyr1850-2100_c190214.nc'

annual_TSA_U is annual averaged Urban 2m air temperature; annual_TSA_R is annual averaged Rural 2m air temperature; JJA is June, July, and August averaged; TBOT is atmospheric air temperature.

I am not sure what is wrong and appreciate any comments.

Best,
Yuan

 

[slevis]

@Yuan Sun were you expecting a warming trend? If so, my best guess for the lack of warming trend is in your compset:
SSP370_DATM%GSWP3v1_CLM50%SP_SICE_SOCN_MOSART_CISM2%NOEVOLVE_SWAV
This compset suggests to me that you are driving the land model with GSWP3 datm files, i.e. files that range from 1901 to 2014. You should be able to confirm by looking in the atm.log file.

If you want a warming trend, you either have to run with active atmosphere, or use simulated datm files instead of observed.

 

[oleson]

@slevis is correct, you won't see a temperature trend out of the box with this compset because it is using GSWP3 atmospheric forcing, not SSP atmospheric forcing. Apologies if I didn't make that clear in our previous communications.
You have a couple of choices here. You can either force the model with atmospheric output from a fully coupled simulation in which the coupler history output was saved. Or you can use the "anomaly forcing" capability in the datm described here:

1.5.9. Running with anomaly forcing — ctsm release-clm5.0 documentation

Either way, you can contact me via email for detailed instructions. Alternatively, if you are still in contact with @bowen , you can ask him about these methods since I believe he has run with at least one of these methods.

 

[Yuan Sun]

Thanks for the comments @slevis @oleson @bowen. It is a valuable lesson I learned. I will post my modified running later.

Best,
Yuan