Update WW3 cap for 2-way ocean-wave coupling #1348
Conversation
model/src/wav_import_export.F90
Outdated
| @@ -357,6 +359,11 @@ subroutine import_fields( gcomp, time0, timen, rc ) | |||
| if (state_fldchk(importState, 'So_u')) then | |||
| call SetGlobalInput(importState, 'So_u', vm, global_data, rc) | |||
| if (ChkErr(rc,__LINE__,u_FILE_u)) return | |||
| do isea = 1,nsea | |||
There was a problem hiding this comment.
This needs to be isolated so as to not to affect other users of the WW3 mesh cap.
|
@binli2337 is there a corresponding ufs-weather-model PR or did I miss it? |
|
@binli2337 We need a ufs-weather-model PR to correspond with this. I did a quick test and my log is here: I did not get the same results as you - all tests were the same, which honestly I was a little surprised about. I agree with @DeniseWorthen though- it would be best if this was isolated with a flag to turn on/off this feature. |
only when the case_name is ufs.hafs.
|
@JessicaMeixner-NOAA @DeniseWorthen I have updated the code to isolate the changes to the HAFS applications. I will submit a corresponding ufs-weather-model PR soon. Thanks. |
model/src/wav_import_export.F90
Outdated
| @@ -357,6 +360,11 @@ subroutine import_fields( gcomp, time0, timen, rc ) | |||
| if (state_fldchk(importState, 'So_u')) then | |||
| call SetGlobalInput(importState, 'So_u', vm, global_data, rc) | |||
| if (ChkErr(rc,__LINE__,u_FILE_u)) return | |||
| do isea = 1,nsea | |||
| if(casename == "ufs.hafs" .and. abs(global_data(isea)-fillv).lt.0.01) then | |||
There was a problem hiding this comment.
What about:
if(casename == "ufs.hafs") then
do isea = 1,nsea
if abs(global_data(isea)-fillv).lt.0.01) then
global_data(isea)=0.0
That way for non-hafs applications you are not going through this loop.
There was a problem hiding this comment.
@JessicaMeixner-NOAA I will revise the wav_import_export.F90 file.
The changes in WW3 alone will not change any baselines in the UFS regression tests since ocean-current fields are not transferred to WW3 using the emc/develop branch of CMEPS. A related CMEPS PR will transfer the fields.
There was a problem hiding this comment.
Thanks for the explanation @binli2337 !
There was a problem hiding this comment.
@binli2337 In the global model, we do export So_u and So_v to WAV.
There was a problem hiding this comment.
Good point @DeniseWorthen
@binli2337 can you share the ufs-waether-model PR ?
There was a problem hiding this comment.
@JessicaMeixner-NOAA The ufs-weather-model branch is at
https://github.com/binli2337/ufs-weather-model/tree/202501b/ocean_wave_update. A PR will be submitted soon.
model/src/wav_import_export.F90
Outdated
| @@ -371,6 +374,14 @@ subroutine import_fields( gcomp, time0, timen, rc ) | |||
| if (state_fldchk(importState, 'So_u')) then | |||
| call SetGlobalInput(importState, 'So_u', vm, global_data, rc) | |||
| if (ChkErr(rc,__LINE__,u_FILE_u)) return | |||
|
|
|||
| if(casename == "ufs.hafs") then | |||
There was a problem hiding this comment.
I think you need trims on your casename comparisons?
There was a problem hiding this comment.
@DeniseWorthen Updates have been made. Thanks!
model/src/wav_import_export.F90
Outdated
| @@ -380,6 +391,13 @@ subroutine import_fields( gcomp, time0, timen, rc ) | |||
| if (state_fldchk(importState, 'So_v')) then | |||
| call SetGlobalInput(importState, 'So_v', vm, global_data, rc) | |||
| if (ChkErr(rc,__LINE__,u_FILE_u)) return | |||
| if(casename == "ufs.hafs") then | |||
There was a problem hiding this comment.
Does this need the same update that was done above?
|
@binli2337 - the dev/ufs-weather-model branch was updated yesterday. Please merge those changes into your branch here when you get a chance. |
|
@JessicaMeixner-NOAA Updates have been made. Thanks! |
Pull Request Summary
The file wav_import_export.F90 located at model/src will be updated to convert a missing value to zero for the HAFS application.
Description
The HAFS application in the ufs-weather-model will be updated to include 2-way ocean-wave coupling capability. The related CMEPS component will be updated to interpolate ocean-current fields from MOM6 grid to WW3 grid. Since regional MOM6 domain and regional WW3 domain do not completely overlap, a missing value is assigned to the interpolated fields in the non-overlapped regions. In the CMEPS component, the missing value is 9.99e20.
To prevent a model crash, a short-term solution is to convert the missing value to zero in the file model/src/wav_import_export.F90. In the future, the ocean-current fields in the non-overlapped regions can be provided from a global MOM6 output.
Issue(s) addressed
-fixes #1346
Commit Message
Converting a missing value to zero for the imported ocean-current fields.
Check list
Testing