Can I use compset which is not scientific and tested when running CESM2.1.3

[Jerry]

Hi, everyone,
I want to run a fully-coupled compset with external foring set as year 2000 level.
However, I can't find a compset named as "B2000".
I tried to run with the compset with longname "2000_CAM60_CLM50%BGC-CROP_CICE_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD".
That is, I replaced the initialization time in the longname of “B1850” from 1850 to 2000.
I found it works, at least without reporting errors. But I don't know if it's legal, and if there's a risk of a potential energy imbalance.
Replies from anyone will be appreciated.

 

[dbailey]

You should be able to do 2000 instead of 1850. You can check the CO2 values to make sure they are consistent with 2000 levels. That should be a supported compset (longname), but we might not have a short name.

 

[Jerry]

You should be able to do 2000 instead of 1850. You can check the CO2 values to make sure they are consistent with 2000 levels. That should be a supported compset (longname), but we might not have a short name.

Thanks for your reply, Dbailey.
I check the co2 value by "./xmlquery CCSM_CO2_PPMV", and the result is 367.0.
Does this mean that the external forcing has been set at the year 2000 level？

 

[dbailey]

Yes. You should also check the cam namelist for the co2vmr value.

 

[Jerry]

Yes. You should also check the cam namelist for the co2vmr value.

That's what confused me, because I didn't find the "co2vmr" parameter in atm_in.
Parameters about co2 are:
&chem_surfvals_nl
flbc_cycle_yr = 2000
flbc_file = '/public/home/amduan/cesm/new_inputdata/atm/waccm/lb/LBC_2000climo_CMIP6_0p5degLat_c180227.nc'
flbc_list = 'CO2','CH4','N2O','CFC11eq','CFC12'
flbc_type = 'CYCLICAL'
scenario_ghg = 'CHEM_LBC_FILE'
I am not sure if it is normal.

 

[Jerry]

That's what confused me, because I didn't find the "co2vmr" parameter in atm_in.
Parameters about co2 are:
&chem_surfvals_nl
flbc_cycle_yr = 2000
flbc_file = '/public/home/amduan/cesm/new_inputdata/atm/waccm/lb/LBC_2000climo_CMIP6_0p5degLat_c180227.nc'
flbc_list = 'CO2','CH4','N2O','CFC11eq','CFC12'
flbc_type = 'CYCLICAL'
scenario_ghg = 'CHEM_LBC_FILE'
I am not sure if it is normal.

Another question is the spinup time. I has run a B2000 compset which hybrid from a CMIP6 year-2000 initial conditions.
And I wonder how long it would take for the model to reach a balanced state.
I have run for 30 years and found the TOM heat budget (FSNT-FLNT) ranges in [-1.5 1.5] W/m2, so now it is a balanced state?

 

[dbailey]

That's what confused me, because I didn't find the "co2vmr" parameter in atm_in.
Parameters about co2 are:
&chem_surfvals_nl
flbc_cycle_yr = 2000
flbc_file = '/public/home/amduan/cesm/new_inputdata/atm/waccm/lb/LBC_2000climo_CMIP6_0p5degLat_c180227.nc'
flbc_list = 'CO2','CH4','N2O','CFC11eq','CFC12'
flbc_type = 'CYCLICAL'
scenario_ghg = 'CHEM_LBC_FILE'
I am not sure if it is normal.

This might be your version of the code.

 

[dbailey]

Another question is the spinup time. I has run a B2000 compset which hybrid from a CMIP6 year-2000 initial conditions.
And I wonder how long it would take for the model to reach a balanced state.
I have run for 30 years and found the TOM heat budget (FSNT-FLNT) ranges in [-1.5 1.5] W/m2, so now it is a balanced state?

This should be balanced after 30 years. What is the mean value?

 

[Jerry]

This should be balanced after 30 years. What is the mean value?

Thanks for your reply.
I have attached the picture of TOM heat budget. I think it is now in a balanced state.
I'd like to ask another question, if you don't mind.
I have run a 50-yr experiment and set the DOUT_S_SAVE_INTERIM_RESTART_FILES=TRUE to get restart filles for every single month.
And I want to know how to hybrid from a month that is not January (Here we want to hybrid from 0050-05-01).
Because I find that the timestamp of the cam initial file doesn't match the timestamps of the other files under the $DOUT_S_ROOT/$CASE/rest/YYYY-MM-01-00000 directory.
And that would cause the hybrid run to report an error with " ERROR: GETFIL: FAILED to get b.hybrid.from.CMIPhist2000.CTRL.branch.cam.i.0050-05-01-00000.nc"
So where should I get the required initial file of "b.hybrid.from.CMIPhist2000.CTRL.branch.cam.i.0050-05-01-00000.nc"?
And I also attach the rest directory files screenshot as well as the reported error of hybrid run.

 

[slevis]

I'm pretty sure that you could start any simulation (continue or hybrid or branch or startup) from any day of the year, as long as you have the necessary files and fulfill the requirements for each type of run as explained in the env_run.xml file.

 

[Jerry]

I'm pretty sure that you could start any simulation (continue or hybrid or branch or startup) from any day of the year, as long as you have the necessary files and fulfill the requirements for each type of run as explained in the env_run.xml file.

Slevis, Thanks for your reply.
I retried with a branch run instead of a hybrid run and succeed.
Does that mean the branch run is more relaxed comparing to hybrid run?

 

[dbailey]

It should be the reverse. A hybrid run is more relaxed. A branch run is like a "continue" run.