Total Solar Irridiance Modification in CAM5

[jgvirgin@uwaterloo_ca]

Hello,

I am trying to run a series of perturbed parameter experiments using CESM1.2.2 with a reduced solar constant. Using the B1850CN compset (with CAM4) I can modify the namelist parameter 'solar_const' in the user_nl_cam file with ease. However, when using the same compset with CAM5 (B1850C5CN), I am running into some problems. The atmosphere namelist for CAM5 with RRTMG for radiation doesn't have a solar_const variable, and when I try to include one and simultaneously remove reference to the solar_data_file I get this warning when running ./preview_namelist:

CAM build-namelist - WARNING: It is not allowed to set both solar_const and solar_data_file. solar_const will be ignored.

When I try to change the radiation model back to CAMRT, I get incompatibility errors with other model components. My question is, what's the best way to modify the TSI when running CESM1.2.2 with RRTMG? I presume the TSI is now calculated as opposed to specified, so does this mean I'd have to make a new solar spectral irradiance file with modified spectral irradiance values?

Any advice at all would be appreciated,

Jack

 

[brianpm]

Hi Jack,

When I have modified the TSI with CESM1 and CESM2, that is exactly how I've done it: I change the spectral irradiance file. There might be combination of runtime parameters that you could use to set a solar constant, but like your experience, I have not had luck doing that. If you are interested, I have a small python script that takes an input solar spectral file and a target TSI to produce a new file that can be used. The downside is that it isn't physical because it just scales the spectrum, so the output doesn't match the spectrum that you'd expect from a blackbody, but it seems to work ok.

 

[jgvirgin@uwaterloo_ca]

Hi Jack,

When I have modified the TSI with CESM1 and CESM2, that is exactly how I've done it: I change the spectral irradiance file. There might be combination of runtime parameters that you could use to set a solar constant, but like your experience, I have not had luck doing that. If you are interested, I have a small python script that takes an input solar spectral file and a target TSI to produce a new file that can be used. The downside is that it isn't physical because it just scales the spectrum, so the output doesn't match the spectrum that you'd expect from a blackbody, but it seems to work ok.

Hi Brian,

Okay, I'll do that. Thanks for the advice! The python script would also be much appreciated.

Jack