CAM7 crashing with error: "negative moist layer thickness. timestep or remap time too large"

[michelle_dvorak]

Hello, I am running CESM3 alpha09 release and am wondering what namelist modifications I need to make to get around the error:

"negative moist layer thickness. timestep or remap time too large"... in the cesm.log file this error is preceded by lines similar to:

dec1278.hsn.de.hpc.ucar.edu 938: k,dp_star_moist,u,v,T 57 4.228320909785 -7.64 4.97 246.53
dec1278.hsn.de.hpc.ucar.edu 938: k,dp_star_moist,u,v,T 58 2.925703691101 -6.59 2.77 245.47

In the atm.log the last few lines read:

READ_NEXT_TRCDATA Q

nstep, te 1 0.26151251179149861E+10 0.26150903599374218E+10 -0.19259775100985295E-02 0.98522738570991787E+05 0.20548077300190920E+03
nstep, te 2 0.26151218962278867E+10 0.26150853398998923E+10 -0.20256205159242292E-02 0.98523008251867577E+05 0.20548077300190920E+03

I don't understand what these massive values are corresponding to.

I have implemented a unique topography file across MOM6, CLM6 and CAM7. My only namelist changes to CAM7 is the bnd_topo file. I do not know if that has to do with this, and realize there is a lot of tuning that is still going into this model development. It's not really clear to me (via GitHub) if there is a standard protocol for namelist changes for this version of CESM3 -- any advice?

 

[slevis]

I do not have answers to your questions, but I have a troubleshooting suggestion that I often recommend:
Have you tried running the same code without your change in topography? This may at least narrow down the source of the problem.