Adding udm microphysics as option 27 #2147
Conversation
|
It has failed the tests, so I need to know what happened. |
|
forgot to add new module, done now. |
|
The regression test results: |
|
@dudhia |
| PRESENT( QH_CURR ) .AND. & | ||
| PRESENT( QNN_CURR ) .AND. PRESENT ( QNC_CURR ) .AND. & | ||
| PRESENT( QNR_CURR ).AND. & | ||
| PRESENT( RAINNC ) .AND. PRESENT ( RAINNCV ) ) THEN |
There was a problem hiding this comment.
Do we still check all these variables' presence if we only have one dynamic core?
There was a problem hiding this comment.
I think this is part of a bigger question, but if they are optional in the driver we need these.
|
@dudhia Can you or Songyou add some more description to the scheme, especially for its difference and improvements over WDM7? Is the computation speed improved? |
|
UDM mp largely adopts microphysical processes in WDM7, but with |
|
@dudhia @Songyou184 I took the liberty and added it to the PR itself. It may be useful to add it to the release notes too. |
|
@weiwangncar have we done enough for approval now? |
|
@dudhia Does Intel/ifort or oneAPI or gfortran also require lower optimization to compile? Do you have a runtime number when compared to WDM7? |
|
Hong has found a solution to avoid the optimization problem with the Intel
ftn option and we will add that soon.
…On Wed, Jan 22, 2025 at 7:38 PM weiwangncar ***@***.***> wrote:
@dudhia <https://github.com/dudhia> Does Intel/ifort or oneAPI or
gfortran also require lower optimization to compile? Do you have a runtime
number when compared to WDM7?
—
Reply to this email directly, view it on GitHub
<#2147 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEIZ77B4EUR5455JKO776NL2MBIZZAVCNFSM6AAAAABUI2JPXCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMBYG4ZDINRQG4>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
Hong should finish updates next week. |
|
I am expecting a cmake fix for this in addition to making the small change
for -O3 to work.
…On Wed, Feb 5, 2025 at 3:49 PM weiwangncar ***@***.***> wrote:
Merged #2147 <#2147> into develop.
—
Reply to this email directly, view it on GitHub
<#2147 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEIZ77H7JOX2KUWVTFAEIDD2OKIQZAVCNFSM6AAAAABUI2JPXCVHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJWGIYDCNRZGU3DANY>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
TYPE: bug fix KEYWORDS: cmake, compilation SOURCE: internal DESCRIPTION OF CHANGES: Problem: PR #2147 added a new file that was not added to the CMake build Solution: Add the corresponding file to the respective CMakeLists.txt
TYPE: new feature
KEYWORDS: UFS Double Moment Microphysics scheme (UDM)
SOURCE: Songyou Hong via internal
DESCRIPTION OF CHANGES:
The new microphysics option is based on 7-class WDM7 with revision and improvements and will be mp_physics=27.
LIST OF MODIFIED FILES:
M Registry/Registry.EM_COMMON
M phys/Makefile
M phys/module_microphysics_driver.F
M phys/module_physics_init.F
M phys/module_mp_udm.F
TESTS CONDUCTED:
RELEASE NOTE: An new microphysics, UFS Double Moment (UDM), 7-class microphysics from Songyou Hong is added (mp_physics=27). UDM mp largely adopts microphysical processes in WDM7, but with bug fixes or revisions based on literature and accumulated realism. UDM mp utilizes the in-cloud microphysics concept (Kim and Hong 2018), with the addition of water-friendly aerosols for CCN initialization. Semi-lagrangian sedimentation of Juang and Hong (2010) is also re-configured for computational efficiency and numerical accuracy. All production terms are optimized by introducing a cloud-top definition for hydrometeors.