search

NCAR / ccpp-physics

GFS physics for CCPP
Other
58 stars 145 forks source link

Too little cloud ice aloft when using Thompson MP #966

Closed grantfirl closed 1 year ago

grantfirl commented 2 years ago

Description

Greg Thompson reported issues with ice clouds aloft that can be improved through tuning of the Thompson MP scheme.

Steps to Reproduce

Use the Thompson MP scheme in the UFS and examine results.

lisa-bengtsson commented 1 year ago

@grantfirl, @RuiyuSun is working on a PR for this and has presented the results and optimizations in the microphysics working group, is this a separate issue? @gthompsnWRF is part of that working group and should be aware of the updates discussed. Thanks!

grantfirl commented 1 year ago

@lisa-bengtsson No, I don't think so. I only added an issue here because I didn't already see a GitHub issue associated with @gthompsnWRF 's code changes in https://github.com/ufs-community/ccpp-physics/pull/1 and the UFS code managers are requiring a GitHub issue as part of the UFS review/merge process. @RuiyuSun, if this is incorrect, that you have code changes above and beyond @gthompsnWRF 's changes in https://github.com/ufs-community/ccpp-physics/pull/1, please correct me.

lisa-bengtsson commented 1 year ago

@grantfirl ok, thanks. I recall Ruiyu mentioned the update from Greg (that was tested in the RRFS) to be part of the other modifications found necessary for the GFS. I think Ruiyu will create another PR after some more testing in the global model.

It is a big challenge that updates can be made to one model system (or suite) and impact the forecast performance in another. @yangfanglin it is perhaps something that should be coordinated between the SRW and MRW applications? Or elsewhere?

yangfanglin commented 1 year ago

My understanding is that changes of MP processes in Greg's PR have already been included in Ruiyu's tests. Greg's cloud fraction computation method is optional and is controlled by a namelist. I'd suggest having Greg's PR committed.

Ruiyu will finish his tests and create another PR with more changes on top of Greg's PR.

lisa-bengtsson commented 1 year ago

@yangfanglin it is good that Ruiyu already have Greg's updates as part of his tests, but in general it seems pretty dangerous that schemes that are shared between suites that are used for different applications (RRFS, GFS in this case) can directly enter a control code-base without carefully assessing the impact of the result in the other application.

RuiyuSun commented 1 year ago

@lisa-bengtsson @grantfirl , @yangfanglin is right that the maximum ice number change is included in our GFS coupled and uncoupled tests. We will create a new PR for the changes. One of the main purposes of the changes is to increase ice at high levels.

grantfirl commented 1 year ago

This was addressed in https://github.com/ufs-community/ccpp-physics/pull/1