You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Description:
2 degree clone of #100 using using 5.4 surface datasets, Jordan over glacier, Sturm over land and lakes + paramfile='/glade/work/linnia/CLM6-PPE/ctsm6_cal/paramfiles/caltest0030.nc'
@olyson these runs are wrapping up now. Can you compare #101 and 103-107 in ILAMB?
Note, I just reran #101 to get a CO2 time series that extends through 2023 and 103-107 also use this CO2 streams file
We can look at diags or LDF for any pairings here if it's helpful (I'm doing the full historical runs in LDF for 101 vs. 105 now) , but likely don't need to look at them all? @linniahawkins, do you want diagnostics run on any of these?
Thanks Keith. My initial reaction is that we overshot the calibration: lowering LAI, GPP, TOTVEGC led to survivability issues and too weak of a carbon cycle, and much too low of carbon stocks.
water: ET was low to begin with and we made it lower, similarly we exacerbated the other biases in the water cycle.
I'm investigating potential causes: (1) sparse grid:full grid (2) biome mean metrics and (3) observational data used in our calibration (regridding issues?).
Activity
wwieder commentedon May 19, 2025
@olyson can you help with urban tall building restart magic for this series of 2 degree cases so I can interpolate from the pSASU to HIST cases?
olyson commentedon May 19, 2025
Try these at /glade/derecho/scratch/oleson/ANALYSIS:
ctsm53041_54surfdata_PPEcal23_103_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
ctsm53041_54surfdata_PPEcal25_104_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
ctsm53041_54surfdata_PPEcal27_105_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
ctsm53041_54surfdata_PPEcal30_106_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
ctsm53041_54surfdata_PPEcal32_107_pSASU.clm2.r.0161-01-01-00000_64bitoffset.nc
wwieder commentedon May 20, 2025
@olyson these runs are wrapping up now. Can you compare #101 and 103-107 in ILAMB?
Note, I just reran #101 to get a CO2 time series that extends through 2023 and 103-107 also use this CO2 streams file
We can look at diags or LDF for any pairings here if it's helpful (I'm doing the full historical runs in LDF for 101 vs. 105 now) , but likely don't need to look at them all? @linniahawkins, do you want diagnostics run on any of these?
olyson commentedon May 20, 2025
I'm validating an installation of a new version of ILAMB that @slevis-lmwg and I worked on today, so I will likely start a comparison tomorrow.
linniahawkins commentedon May 21, 2025
Thanks! I think ILAMB+LDF should be sufficient for now.
olyson commentedon May 21, 2025
The ILAMB comparison is here
linniahawkins commentedon May 22, 2025
Thanks Keith. My initial reaction is that we overshot the calibration: lowering LAI, GPP, TOTVEGC led to survivability issues and too weak of a carbon cycle, and much too low of carbon stocks.
water: ET was low to begin with and we made it lower, similarly we exacerbated the other biases in the water cycle.
I'm investigating potential causes: (1) sparse grid:full grid (2) biome mean metrics and (3) observational data used in our calibration (regridding issues?).
[-]PPEcal25: ctsm53041_54surfdata_PPEcal30_106_ [AD, SASU, pSASU, HIST][/-][+]PPEcal30: ctsm53041_54surfdata_PPEcal30_106_ [AD, SASU, pSASU, HIST][/+]