git describe:
cesm3_0_beta09-0-gfcbfad5
------------------------------------------------------------------------
--res hcru_hcru_mt13
--compset I1850Clm60BgcCrop
--compiler gnu
--mach sofia
I submitted a proposal for using CTSM6 on the HPC Sofia (VSC Belgium) for the ISIMIP4-FT.
I turned on the high frequency output, the HumanIndexMod, and irrigation. The simulation was 1 year to get a good representation of high frequency output.
user_nl_clm
limit_irrigation_if_rof_enabled= .true.
use_groundwater_irrigation= .true.
calc_human_stress_indices = 'ALL' !! GEMS removing for the cold start
! DAILY gridcell
hist_fincl6 = 'SWMP65','SWMP65_R','SWMP65_U','SWMP80','SWMP80_R','SWMP80_U','WBT','WBT_R','WBT_U','EPT','EPT_R','EPT_U','THIP','THIP_R','THIP_U','SWBGT','SWBGT_R','SWBGT_U','HUMIDEX','HUMIDEX_R','HUMIDEX_U','HIA','HIA_R','HIA_U','APPAR_TEMP','APPAR_TEMP_R','APPAR_TEMP_U','WBA','WBA_R','WBA_U','DISCOI','DISCOI_R','DISCOI_U','THIC','THIC_R','THIC_U','TSA', 'TSA_U', 'TSA_R', 'PSurf', 'RH2M', 'RH2M_U', 'RH2M_R','FSDS', 'RAIN', 'WASTEHEAT','FSDS'
hist_nhtfrq(6) = -24
hist_mfilt(6) = 365
hist_dov2xy(6) = .true.
But, I get weak scaling results from these tests:
Does CTSM not scale strongly? When I ran B-Case CESM3 scaling tests back in 2024, the model scaled almost 1:1 on NCAR Derecho and CSCS Eiger. Is this due to the cubic-Earth grids? Or is it that CAM uses so much more resources that the CTSM weak scaling is << than CAM strong scaling? Should I be trying to setup I1850Clm60BgcCrop with a cubic earth grid? I would be looking for an equivalent to 360x720 resolution (hcru_hcru_mt13).
Cheers,
-Jonathan
cesm3_0_beta09-0-gfcbfad5
------------------------------------------------------------------------
--res hcru_hcru_mt13
--compset I1850Clm60BgcCrop
--compiler gnu
--mach sofia
I submitted a proposal for using CTSM6 on the HPC Sofia (VSC Belgium) for the ISIMIP4-FT.
I turned on the high frequency output, the HumanIndexMod, and irrigation. The simulation was 1 year to get a good representation of high frequency output.
user_nl_clm
limit_irrigation_if_rof_enabled= .true.
use_groundwater_irrigation= .true.
calc_human_stress_indices = 'ALL' !! GEMS removing for the cold start
! DAILY gridcell
hist_fincl6 = 'SWMP65','SWMP65_R','SWMP65_U','SWMP80','SWMP80_R','SWMP80_U','WBT','WBT_R','WBT_U','EPT','EPT_R','EPT_U','THIP','THIP_R','THIP_U','SWBGT','SWBGT_R','SWBGT_U','HUMIDEX','HUMIDEX_R','HUMIDEX_U','HIA','HIA_R','HIA_U','APPAR_TEMP','APPAR_TEMP_R','APPAR_TEMP_U','WBA','WBA_R','WBA_U','DISCOI','DISCOI_R','DISCOI_U','THIC','THIC_R','THIC_U','TSA', 'TSA_U', 'TSA_R', 'PSurf', 'RH2M', 'RH2M_U', 'RH2M_R','FSDS', 'RAIN', 'WASTEHEAT','FSDS'
hist_nhtfrq(6) = -24
hist_mfilt(6) = 365
hist_dov2xy(6) = .true.
But, I get weak scaling results from these tests:
Does CTSM not scale strongly? When I ran B-Case CESM3 scaling tests back in 2024, the model scaled almost 1:1 on NCAR Derecho and CSCS Eiger. Is this due to the cubic-Earth grids? Or is it that CAM uses so much more resources that the CTSM weak scaling is << than CAM strong scaling? Should I be trying to setup I1850Clm60BgcCrop with a cubic earth grid? I would be looking for an equivalent to 360x720 resolution (hcru_hcru_mt13).
Cheers,
-Jonathan