Bugfix/atrayano/uw regression fix #875
Conversation
Sync develop into main
Sync develop into main
There was a problem hiding this comment.
The cnvtr dependence was originally added to reduce entrainment in the presence of deep convective cold pools. GF cold pool development was later set aside and we switched to a precip dependence. Removing this is fine, but we should verify it doesn't negatively impact skill.
|
@narnold1 thanks for your quick response! I wanted to hear your opinion, and this was the reason I made this PR "draft". Let see how this impacts the skill. If we need to keep your "cntvr" dependence, we could fix the regression issue differently |
|
I'll run a test this afternoon to verify that the effect is small. |
|
In replay experiments with L137 and GFDL microphysics, this results in a small improvement in 850 mb temperature and small degradation in water vapor. I suspect the overall change is either neutral or slightly beneficial. Expected to be zero diff for default L72 configuration. |
This fixes #874. The choice of the limiter was suggested by @wmputman. We need to coordinate with @narnold1