Description:
Same as ctsm51_ctsm51d150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD but run on Derecho.
The purpose is to see what differences we get on Derecho compared to Cheyenne, and to establish a new baseline on Derecho for the dead veg discussion.
Differences would not be entirely due to Cheyenne-> Derecho because of answer-changing tag differences between ctsm5.1.dev160 (required for Derecho support) and ctsm5.1.dev150 (used in the Cheyenne version of the simulation), but my plan is to try to revert the answer-changing items for this simulation.
Case directory:
Locally:
/glade/work/oleson/ctsm5.1.dev160/cime/scripts/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD
user_nl_clm:
hist_nhtfrq = 0
hist_mfilt = 1
hist_empty_htapes = .false.
hist_fincl1 = 'FSDSND','FSDSVD','FSDSNI','FSDSVI','RAIN_FROM_ATM','SNOW_FROM_ATM','FLDS_NOT_DOWNSCALED','ZBOT','Tair_from_atm','Thair_from_atm','QBOT_NOT_DOWNSCALED','PBOT_NOT_DOWNSCALED','Rho_from_atm','PCO2','UWIND','VWIND','BCPHIDRY','BCPHODRY','BCPHIWET','OCPHIDRY','OCPHODRY','OCPHIWET','DSTWET1','DSTDRY1','DSTWET2','DSTDRY2','DSTWET3','DSTDRY3','DSTWET4','DSTDRY4','ATM_TOPO'
use_c13 = .true.
use_c14 = .true.
use_c13_timeseries = .true.
use_c14_bombspike = .true.
fsurdat = '/glade/work/slevis/git/mksurfdata_toolchain/tools/mksurfdata_esmf/surfdata_ne30np4.pg3_SSP5-8.5_78pfts_CMIP6_1850-2100_c230227.nc'
!KO This file should not create any differences with ctsm5.1.dev150 (see /glade/campaign/cgd/tss/people/oleson/modify_param/modify_ctsm51_params.c231117_with_ABCD_blk_A5BCD.ncl)
paramfile = '/glade/campaign/cgd/tss/people/oleson/modify_param/ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc'
snicar_snobc_intmix = .true.
!KO No longer in namelist, now in parameter file
!fresh_snw_rds_max = 400.d
snow_thermal_cond_method = 'Sturm1997'
SourceMods:
src.clm/atm2lndType.F90: Add some additional history fields
src.clm/clm_varpar.F90: nlevurb=5 -> nlevurb=10
src.clm/surfrdMod.F90: Remove read of PFTDATA_MASK
src.clm/Wateratm2lndType.F90: Add some additional history fields
src.clm/CNPhenologyMod.F90 decreasing temperature thresholds to -3C, reducing the minimum critical daylength by 25%, doubling the snow threshold (I think to 20 cm)
src.clm/controlMod.F90 remove error associated with snicar_snobc_intmix
src.clm/SnowSnicarMod.F90 set flg_snoage_scl to .true. so that xdrdt change takes effect
src.clm/BareGroundFluxesMod.F90 revert forcing height change
src.clm/CanopyFluxesMod.F90 revert forcing height change
src.clm/LakeFluxesMod.F90 revert forcing height change
Output:
Output (if still available):
/glade/derecho/scratch/oleson/archive/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD/
Contacts:
@olyson
@wwieder
Extra details:
I've designated this combination simulation as ABCDE_blk_A5BCD, where the first ABCD is:
A - SturmSnowtk (implemented via namelist snow_thermal_cond_method = 'Sturm1997')
B - decidphenotrig (implemented via src.clm/CNPhenologyMod.F90)
C - frootleaf (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc)
D - FUNfracfixers (implemented via ctsm51_params.c2311117.nc.ABCD_blk_A5BCD.nc)
E - new SNICAR (on by default)
blk - snicar_snobc_intmix = .true. (implemented via namelist, per Issue #36 )
and the next ABCD is (per Issue #36 )
A5 - xdrdt = 5 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc and takes effect by code change to SnowSnicarMod.F90)
B - scvng_fct_mlt_sf = 0.5 (implemented via ctsm51_params.c2311117.nc.ABCD_blk_A5BCD.nc)
C - snw_rds_refrz = 1500 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc)
D - fresh_snw_rds_max = 400._r8 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc)
See discussion in https://github.com/NCAR/LMWG_dev/discussions/3
AD spunup for 40 years.
I believe I've backed out the answer-changing stuff from ctsm5.1.dev150 to ctsm5.1.dev160. One exception might be a change in dev154 as noted in the change log: "6dc1966: This git merge escomp/master probably does change answers from previous commit". Regardless, it looks like the differences in the diagnostics are just noise, nothing systematic, so likely just due to machine differences (cheyenne->derecho).
Complete for now.
Description: Same as ctsm51_ctsm51d150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD but run on Derecho. The purpose is to see what differences we get on Derecho compared to Cheyenne, and to establish a new baseline on Derecho for the dead veg discussion. Differences would not be entirely due to Cheyenne-> Derecho because of answer-changing tag differences between ctsm5.1.dev160 (required for Derecho support) and ctsm5.1.dev150 (used in the Cheyenne version of the simulation), but my plan is to try to revert the answer-changing items for this simulation.
Case directory: Locally: /glade/work/oleson/ctsm5.1.dev160/cime/scripts/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD
Sandbox: Locally: /glade/work/oleson/ctsm5.1.dev160
usernl changes:
user_nl_clm: hist_nhtfrq = 0 hist_mfilt = 1 hist_empty_htapes = .false. hist_fincl1 = 'FSDSND','FSDSVD','FSDSNI','FSDSVI','RAIN_FROM_ATM','SNOW_FROM_ATM','FLDS_NOT_DOWNSCALED','ZBOT','Tair_from_atm','Thair_from_atm','QBOT_NOT_DOWNSCALED','PBOT_NOT_DOWNSCALED','Rho_from_atm','PCO2','UWIND','VWIND','BCPHIDRY','BCPHODRY','BCPHIWET','OCPHIDRY','OCPHODRY','OCPHIWET','DSTWET1','DSTDRY1','DSTWET2','DSTDRY2','DSTWET3','DSTDRY3','DSTWET4','DSTDRY4','ATM_TOPO' use_c13 = .true. use_c14 = .true. use_c13_timeseries = .true. use_c14_bombspike = .true. fsurdat = '/glade/work/slevis/git/mksurfdata_toolchain/tools/mksurfdata_esmf/surfdata_ne30np4.pg3_SSP5-8.5_78pfts_CMIP6_1850-2100_c230227.nc' !KO This file should not create any differences with ctsm5.1.dev150 (see /glade/campaign/cgd/tss/people/oleson/modify_param/modify_ctsm51_params.c231117_with_ABCD_blk_A5BCD.ncl) paramfile = '/glade/campaign/cgd/tss/people/oleson/modify_param/ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc'
snicar_snobc_intmix = .true.
!KO No longer in namelist, now in parameter file !fresh_snw_rds_max = 400.d
snow_thermal_cond_method = 'Sturm1997'
SourceMods:
src.clm/atm2lndType.F90: Add some additional history fields src.clm/clm_varpar.F90: nlevurb=5 -> nlevurb=10 src.clm/surfrdMod.F90: Remove read of PFTDATA_MASK src.clm/Wateratm2lndType.F90: Add some additional history fields src.clm/CNPhenologyMod.F90 decreasing temperature thresholds to -3C, reducing the minimum critical daylength by 25%, doubling the snow threshold (I think to 20 cm) src.clm/controlMod.F90 remove error associated with snicar_snobc_intmix src.clm/SnowSnicarMod.F90 set flg_snoage_scl to .true. so that xdrdt change takes effect src.clm/BareGroundFluxesMod.F90 revert forcing height change src.clm/CanopyFluxesMod.F90 revert forcing height change src.clm/LakeFluxesMod.F90 revert forcing height change
Diagnostics:
https://webext.cgd.ucar.edu/I1850/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD/lnd/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD_31_40-ctsm51_ctsm51d150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD_31_40/setsIndex.html
Output: Output (if still available): /glade/derecho/scratch/oleson/archive/ctsm51_ctsm51d160r150_ne30pg3ne30pg3mg17_GSWP3V1_ABCDE_blk_A5BCD_1850AD/
Contacts: @olyson @wwieder
Extra details: I've designated this combination simulation as ABCDE_blk_A5BCD, where the first ABCD is: A - SturmSnowtk (implemented via namelist snow_thermal_cond_method = 'Sturm1997') B - decidphenotrig (implemented via src.clm/CNPhenologyMod.F90) C - frootleaf (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc) D - FUNfracfixers (implemented via ctsm51_params.c2311117.nc.ABCD_blk_A5BCD.nc) E - new SNICAR (on by default) blk - snicar_snobc_intmix = .true. (implemented via namelist, per Issue #36 ) and the next ABCD is (per Issue #36 ) A5 - xdrdt = 5 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc and takes effect by code change to SnowSnicarMod.F90) B - scvng_fct_mlt_sf = 0.5 (implemented via ctsm51_params.c2311117.nc.ABCD_blk_A5BCD.nc) C - snw_rds_refrz = 1500 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc) D - fresh_snw_rds_max = 400._r8 (implemented via ctsm51_params.c231117.nc.ABCD_blk_A5BCD.nc)
See discussion in https://github.com/NCAR/LMWG_dev/discussions/3 AD spunup for 40 years. I believe I've backed out the answer-changing stuff from ctsm5.1.dev150 to ctsm5.1.dev160. One exception might be a change in dev154 as noted in the change log: "6dc1966: This git merge escomp/master probably does change answers from previous commit". Regardless, it looks like the differences in the diagnostics are just noise, nothing systematic, so likely just due to machine differences (cheyenne->derecho). Complete for now.