Marine low cloud differences between CESM2.1 and CESM2.2?

[cll026]

I'm using cesm2.2.1-rc.01 (CAM tag cam_cesm2_2_rel_07) for a study that depends on the distribution of marine low cloud, particularly subtropical stratocumulus. My analysis regions were originally derived from CESM2.1 output, and I'm trying to work out whether they're still appropriate for CESM2.2. Are there known differences in marine low cloud between the CESM2.1.z and CESM2.2 series? The release notes mention "answer-changing developments," but the CAM6.3 changes appear centered on dynamical cores and I haven't found anything specific to CLUBB or low cloud. Relatedly, does any free-running CESM2.2 output at f09 exist publicly, ideally with COSP enabled? Everything CESM2.2 at f09 I've found on GLADE is nudged and/or built with cosp=0.

Thanks in advance!

 

[brianpm]

Hi Claire,

I'd recommend following the guidance on Community Earth System Model 2 (CESM2) | Community Earth System Model and stick to the 2.1 series if possible. The main difference with the 2.2 series is supposed to be some dynamical core capabilities, as you mentioned, but there could be some other differences too. I don't know of any public runs of 2.2.x with COSP turned on (there could be some, but I don't know where they would be).

There's also a good chance that there are a few changes even within the 2.0-2.1 series of models that could impact low clouds. We noticed a change in cloud feedbacks and climate sensitivity (see Section 7 of https://doi.org/10.1175/JCLI-D-23-0250.1).

If you are working on the NCAR HPC system and want me to take a quick look at your setup, just shoot me an email. You could also have an LLM try to identify the differences between the versions and your 2.1 vs 2.2 cases. For purposes of defining regions of subtropical stratocumulus, the differences should be relatively small across all the CESM2 versions.