-
Notifications
You must be signed in to change notification settings - Fork 54
Introduce Thompson MP namelist options to adjust rain and snow max terminal fall speeds; Enable RRTMGP working with nesting; Introduce SASAS namelist options to control convective adjustment time #332
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Changes from 3 commits
24f17a7
462bf71
330755a
01898e4
2aa85ba
5350bf8
d177274
1430d38
fd734a0
c67e810
2c90394
5c430f9
a4793b8
89e7c2c
605d486
d08592d
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -48,7 +48,7 @@ module rrtmgp_lw_cloud_optics | |
| ! ###################################################################################### | ||
| !> | ||
| subroutine rrtmgp_lw_cloud_optics_init(rrtmgp_root_dir, rrtmgp_lw_file_clouds, & | ||
| nrghice, mpicomm, mpirank, mpiroot, errmsg, errflg) | ||
| nrghice, mpicomm, mpirank, mpiroot1, errmsg, errflg) | ||
|
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. @dustinswales @BinLiu-NOAA I am worried about this change. The host model defines mpiroot, not the scheme. You are hardcoding it in the RRTMGP scheme to be zero. This is a dangerous assumption in my opinion. Can you not just pass in the value you need (i.e. zero) from the host model?
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. @climbfuji I got a similar concern on this. @Qingfu-Liu and @dustinswales, not sure if it's easy to come up a more generalized solution for this.
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Yeah, this is not an ideal solution. Sorry for not proposing something cleaner. The problem was related to nested domains and initialization. Setting mpiroot=0 in the scheme avoided the situation where the nested domain(s) would try to reset this global data already being set by the parent domain. @BinLiu-NOAA @climbfuji Hang tight and I'll work on a fix
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Actually, I just realized that the UWM PR does not add any new testing for the HAFS_v2 suites being introduced to NOAA-EMC/ufsatm.
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. @dustinswales We do plan to update some existing HAFS RTs in UWM to be aligned with future HAFS v2.2 configurations/suites. However, these HAFS RTs related updates will most likely come from a later/future PR, since the HAFSv2.2 final configurations are still evolving (hopefully will be finalized in the next few weeks). Meanwhile, definitely can provide you a canned test case for developing/testing alternative solutions before the HAFS RT update PR. Thanks!
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
In terms alternative solutions, I recall @SamuelTrahanNOAA and @JiayiPeng-NOAA worked/implemented an approach for a similar issue related nesting/moving-nesting support for the stochastics physics. Please feel free to comment/suggest if you have some ideas. Thanks!
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
@BinLiu-NOAA That works for me. (No rush as I will be off until 1/5/26.)
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Agreed RE: the need for a more general solution. Do the nest physics always get initialized after the parent domain physics? If so, would a logical
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. @climbfuji @grantfirl We could just set Init_parm%master = 0 in fv3/atmos_model.F90: And revert all these changes to the GP scheme files. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I don't understand this completely, but what @grantfirl suggests seems to make the most sense - keep track of the initialization through a host model (or scheme module) variable? |
||
|
|
||
| ! Inputs | ||
| character(len=128),intent(in) :: & | ||
|
|
@@ -61,7 +61,8 @@ subroutine rrtmgp_lw_cloud_optics_init(rrtmgp_root_dir, rrtmgp_lw_file_clouds, | |
| mpicomm !< MPI communicator | ||
| integer, intent(in) :: & | ||
| mpirank, & !< Current MPI rank | ||
| mpiroot !< Master MPI rank | ||
| mpiroot1 !< Master MPI rank | ||
| integer :: mpiroot | ||
|
|
||
| ! Outputs | ||
| character(len=*), intent(out) :: & | ||
|
|
@@ -77,6 +78,7 @@ subroutine rrtmgp_lw_cloud_optics_init(rrtmgp_root_dir, rrtmgp_lw_file_clouds, | |
| errmsg = '' | ||
| errflg = 0 | ||
|
|
||
| mpiroot = 0 | ||
| ! Filenames are set in the physics_nml | ||
| lw_cloud_props_file = trim(rrtmgp_root_dir)//trim(rrtmgp_lw_file_clouds) | ||
|
|
||
|
|
||
Uh oh!
There was an error while loading. Please reload this page.