Erik Kluzek

The change has to do with how MPI is binding processors on Derecho. This must have some influence on how the code is running for threading. There were problems with...

Note there are some issues with threading for CN-Matrix. 1. Threading code uses different calls than without it 2. Matrix is using the OMP_NUM_THREADS env variable rather than what's passed...

1. So the problem is that OMP_NUM_THREADS is a proper UNIX environment variable and NOT part of the env_run.xml file. So it has to be treated in a different way...

In PR #2545 I saw a test fail in comparison baseline that is because of this issue see this comment: https://github.com/ESCOMP/CTSM/pull/2545#issuecomment-2249546181 The test is: ERP_P64x2_Lm13.f10_f10_mg37.IHistClm60Bgc.derecho_intel.clm-monthly--clm-matrixcnOn_ignore_warnings so we should expect this...

@wwieder this isn't something we really need now, but I would like to check off the next three boxes in the checklist at the top by ctsm6.0.0. So I put...

time_bounds is an expected part of the CF convention, so I do endorse using it for anything with time. That's likely why some tools might expect it to be there....

Good point @samsrabin. That convinces me we should remove it for I fields then. That would then be in line with the convention. We don't similarly report on the grid...

FYI: @wwieder and @slevis-lmwg I removed the discussion label here. It didn't seem like it should be converted to a discussion.

Thanks @lawrencepj1 @slevis-lmwg and @wwieder for the discussion. I agree with everyone we certainly shouldn't add plantation forests in the model now. I will create an issue for us doing...

OK, I created a science discussion for this here: https://github.com/ESCOMP/CTSM/discussions/3045 and an issue about potential veg in #3046