What version of the code are you using?
CTSM5.2.015
Have you made any changes to files in the source tree?
No
Describe every step you took leading up to the problem:
user_nl_clm and user_nl_datm_streams are attached below.
Describe your problem or question:
I am running a high-resolution regional (1/32° over CONUS) land-only simulation using CTSM5.2.015 on derecho, forced by bias-corrected ERA5 data regridded to match the 1/32° mesh. For spin-up, I applied cyclic forcing using the 2010-2014 ERA5 climatology.
When spinning up, the model ran successfully from 2010-01 to 2010-12, generating a restart file for 2011-01. However, after resubmitting, I encounter the error: “Longwave down sent from the atmosphere model is negative or zero.” However, setting RUN_STARTDATE=2011-01-01 and running from that date directly works without any issues.
Here's what I have tried to debug:
1. Forcing data check: I have checked the forcing data and there is no negative or zero FLDS.
2. Configurations check:
1) Change tintalgo for forcing
By default it was
I changed it to
also tried
both failed with same error as above
2) Change use_init_interp=.true.
Initially I set use_init_interp=.false., changing it to .true. (cleaned init_generated_files or created a new case) didn't work.
3) Change STOP_N/RUN_STARTDATE/STOP_OPTION
Different dates for stopping/resubmitting all resulted in the same error. The error arises only upon resubmission, while the model runs fine for continuous three-year spans (e.g., 2010-01 to 2012-12).
Given that the forcing data appears to be correct, there seemed to be something wrong with the restart/initialization process. Any suggestions or insights would be appreciated! Thanks!
CTSM5.2.015
Have you made any changes to files in the source tree?
No
Describe every step you took leading up to the problem:
Code:
# create and setup case
CASEDIR="/glade/work/yifanc17/cases/WFDE5.Regridded.CTSM5.2.015.ExplicitAC.CONUS.1d32deg.CTRL.SPINUP.c241106"
/glade/work/yifanc17/code/ctsm5.2.015/cime/scripts/create_newcase --case "${CASEDIR}" --compset 2000_DATM%GSWP3v1_CLM50%SP_SICE_SOCN_MOSART_SGLC_SWAV --res CLM_USRDAT --run-unsupported
./xmlchange LND_DOMAIN_MESH=/glade/work/yifanc17/02_data/cesmdata/meshdata/CONUS_3p75/CONUS_1d32deg_mesh_WFDE5_masked_c241028.nc # regional mesh
./xmlchange ATM_DOMAIN_MESH=/glade/work/yifanc17/02_data/cesmdata/meshdata/CONUS_3p75/CONUS_1d32deg_mesh_WFDE5_masked_c241028.nc
./xmlchange MASK_MESH=/glade/work/yifanc17/02_data/cesmdata/meshdata/CONUS_3p75/CONUS_1d32deg_mesh_WFDE5_masked_c241028.nc
./xmlchange PIO_TYPENAME=pnetcdf,PIO_NETCDF_FORMAT=64bit_data # for parallel IO
./xmlchange RUN_STARTDATE=2010-01-01,DATM_YR_START=2010,DATM_YR_ALIGN=2010,DATM_YR_END=2014,RESUBMIT=1,STOP_N=1,STOP_OPTION=nyears # unit testuser_nl_clm and user_nl_datm_streams are attached below.
Describe your problem or question:
I am running a high-resolution regional (1/32° over CONUS) land-only simulation using CTSM5.2.015 on derecho, forced by bias-corrected ERA5 data regridded to match the 1/32° mesh. For spin-up, I applied cyclic forcing using the 2010-2014 ERA5 climatology.
When spinning up, the model ran successfully from 2010-01 to 2010-12, generating a restart file for 2011-01. However, after resubmitting, I encounter the error: “Longwave down sent from the atmosphere model is negative or zero.” However, setting RUN_STARTDATE=2011-01-01 and running from that date directly works without any issues.
Here's what I have tried to debug:
1. Forcing data check: I have checked the forcing data and there is no negative or zero FLDS.
2. Configurations check:
1) Change tintalgo for forcing
By default it was
Code:
CLMGSWP3v1.Solar:tintalgo = coszen
CLMGSWP3v1.TPQW:tintalgo = linear
Code:
CLMGSWP3v1.Solar:tintalgo = nearest
CLMGSWP3v1.TPQW:tintalgo = linear
Code:
CLMGSWP3v1.Solar:tintalgo = nearest
CLMGSWP3v1.TPQW:tintalgo = nearest2) Change use_init_interp=.true.
Initially I set use_init_interp=.false., changing it to .true. (cleaned init_generated_files or created a new case) didn't work.
3) Change STOP_N/RUN_STARTDATE/STOP_OPTION
Different dates for stopping/resubmitting all resulted in the same error. The error arises only upon resubmission, while the model runs fine for continuous three-year spans (e.g., 2010-01 to 2012-12).
Given that the forcing data appears to be correct, there seemed to be something wrong with the restart/initialization process. Any suggestions or insights would be appreciated! Thanks!