Known bug in CESM release F_2000_WACCM (FW) compset

[mmills@ucar_edu]

We have discovered a bug in the current release of CESM for the WACCM present-day compset (F_2000_WACCM or FW). Before configuring this case, users should modify the file env_conf.xml using the command:

xmlchange -file env_conf.xml -id CCSM_CO2_PPMV -val 368.9

This variable had been set to 1e-6, as it was believed it was not used by the model. It is used in this compset, however. If it is not changed, the land model will crash about 80 days into the run.

 

[mmills@ucar_edu]

This bug has since been fixed in a later release (cesm1_0_1).

 

[gchiodo@fis_ucm_es]

Dear all,

I recently noticed that the same bug may be affecting the -compset B_1850_WACCM_CN.
of the cesm1_0_2 distribution. I am running a 20-yr pre-industrial control run (not transient) but I see following settings in my env_conf.xlm:




I think that value of CO2 is quite unrealistic.
Should I change it manually and build the model again?

Thanks a lot

 

[mmills@ucar_edu]

No, you should not change that. The namelist variable CCSM_CO2_PPMV is only used in the F compsets. In the B compsets, the CO2 comes from other input files. We have run the B_1850_WACCM_CN compset for 150 years without running into any problems with this.