ctsm5.4.044: Merge b4bdev 20260605#4083
Conversation
…iogeochemistry-BGC-spinup.rst Add double quotes around module load ncl Co-authored-by: Keith Oleson <oleson@ucar.edu>
…gcspinups_sec1.5.7
…of-figs-tables Tech Note: Delete lists of figures and tables
matlab script for appending WIEMIP co2 scenario data to TRENDY2025
Overflow respiration bug fixes PR ESCOMP#4014 Issue 3491 Changes answers in MIMICS cases only.
samsrabin
left a comment
There was a problem hiding this comment.
Pre-approving with a brief glance, assuming it's a straightforward merge. Also assuming that the updated submodules are intentional.
|
I don't get how this happened. But, I assume it's just an extremly rare glitch with git. The changes to CLMBuildNamelist.pm in ctsm5.4.042 somehow got lost. git thinks they are there so I can't just remerge it in. So I have to do it more or less by hand. |
For some reason CLMBuildNamelist.pm got screwed up and the updates in
ctsm5.4.042 backed out. git had the history and thought it was correct
so merely merging it again didn't correct it.
So I fixed it by hand by doing this:
git show {ctsm5.4.042:bld/CLMBuildNamelist.pm} > CLMBuildNamelist.pm
|
The two hillslope tests that showed different answers now pass: SMS_D_Mmpi-serial_Ld5.5x5_amazon.I2000Clm60Bgc.derecho_gnu.clm-HillslopeC so the testing is now as expected |
|
I think I see what happened. There was a merge of master onto @swensosc branch in PR #4057 and then a mistaken undoing of a merge that was OK here: The b4b-dev branch already correctly had that change from ctsm5.4.042, and so didn't need it backed out. But, I also see how it's easy to get that wrong, it was basically the classic "off by one" problem we easily run into. So then I ended up undoing the above commit. So it wasn't a glitch in git, it was a simple human mistake. |
Ah, so this was another example of what I remembered this morning. A way to think about it is that we would have likely encountered significant difficulty merging to b4b-dev if we had kept the merged master in Sean's PR, so we would have reconsidered a merge to b4b-dev. Github would have also shown unexpected diffs and might have given warnings about the branch having conflicts or being inconsistent with base. In any case, I'm glad you cleared it up @ekluzek :-) |
Description of changes
Bring latest b4b-dev to master.
Mostly updates to the documentation, especially the tech note.
Also update of submodules to almost the latest ones. This includes a few updates in cdeps to help with spinup. One specific update is to allow using the CO2 from the CPLHIST files for spinup cases. Also use daily files for CO2 CPLHIST rather than 3-hourly to sync with the change in CMEPS.
NOTE: This brings in @samsrabin update to cime that will fix the glitch we have with baselines on Izumi!
Also update the FATES parameter generation for testing
And update how the ch4 history fields are handled
Specific notes
Contributors other than yourself, if any:
CTSM issues resolved or otherwise addressed, if any:
Any user interface changes (namelist or namelist defaults changes)?
Add cplhist option to DATM_CO2_TSERIES
** PR's Brought in**
Total 22
Changes to code: 4
Changes to documentation: 18
Changes to Tech Note: 14
Other documentation updates: 4
Testing planned or performed, if any:
(After tag is made kick off fates tests for Derecho and Izumi)
Requirements before merge: