Updates to support UPP use with MPAS by jaymes-kenyon · Pull Request #1018 · NOAA-EMC/UPP

jaymes-kenyon · 2024-08-19T21:06:26Z

This PR contains the initial modifications to support UPP functionality with MPAS (specifically, MPASSIT-generated netCDF history files). A newINITPOST_MPAS.F interface is proposed, copied from INITPOST.F. Several static control files are also included; these control files are currently in use by GSL in its real-time HFIP and RRFS-MPAS runs.

This PR addresses issue #1008. Thanks to @SamuelTrahanNOAA for the help in making changes to getVariable.f and getIVariableN.f.

….xml file is adapted from hrrr_postcntrl.xml.

…lencing print statements for closely spaced model levels, and (3) adding comments regarding the "T" array

…h netCDF variable, grid locations are assigned SPVAL wherever "_FillValue" is found.

…ther.

WenMeng-NOAA · 2024-08-22T12:21:28Z

@jaymes-kenyon Do you have the sample MPAS model outputs for my UPP standalone test?

jaymes-kenyon · 2024-08-22T14:10:21Z

@WenMeng-NOAA — Here's a sample netCDF file on Hera; it's a 12-h forecast from the RRFS-MPAS prototype configuration (3-km CONUS):
/scratch2/BMC/wrfruc/kenyon/sample/rrfs_mpas/2024082100/MPAS-A_out.2024-08-21_12.00.00.nc

Here's an example "itag" file for testing (note the MODELNAME, SUBMODELNAME, and fileNameFlat):

&model_inputs
fileName='/mnt/lfs4/BMC/wrfruc/kenyon/mpas_upp/rrfs/2024082100/MPAS-A_out.2024-08-21_12.00.00.nc'
fileNameFlux='/mnt/lfs4/BMC/wrfruc/kenyon/mpas_upp/rrfs/2024082100/MPAS-A_out.2024-08-21_12.00.00.nc'
IOFORM='netcdfpara'
grib='grib2'
DateStr='2024-08-21_12:00:00'
MODELNAME='RAPR'
SUBMODELNAME='MPAS'
fileNameFlat='postxconfig-NT-rrfs_mpas.txt'
/

Thanks for the testing!

WenMeng-NOAA · 2024-08-23T16:15:08Z

@jaymes-kenyon Thanks for sharing with me your itag. My test results are at /home/Wen.Meng/stmp2/mpas_2024082112 on Hera. Please validate and let me if you see issues.
PS. I would suggest to remove the line "fileNameFlat=..." in tag. The 'fileNameFlat' input is not needed for ingesting WRFIO model output, like rap/hrrr/mpas.

jaymes-kenyon · 2024-08-23T16:54:00Z

@WenMeng-NOAA — Thanks! Your test results look good.

Thanks also for the clarification regarding the use of "fileNameFlat" within the itag file (for WRFIO applications). I see now that WRFPOST.F specifies a default name (fileNameFlat='postxconfig-NT.txt'). So, the chosen flat file just needs to be linked within the run directory as this default name.

Another MPAS-related PR will likely be ready soon, but I wanted to get this PR done first. Thanks again!

WenMeng-NOAA · 2024-08-23T17:15:11Z

@jaymes-kenyon Thanks for verifying. In my test, I didn't add the entry "fileNameFlat" in itag. Instead, I copied file 'postxconfig-NT-rrfs_mpas.txt" into the run directory and rename as 'postxconfig-NT.txt'.
BTW we will proceed your PR to the merging process.

WenMeng-NOAA · 2024-08-23T17:39:09Z

The github workflow failed at GCC Linux build. Do you know the root cause? @SamuelTrahanNOAA @AlexanderRichert-NOAA @jaymes-kenyon

WenMeng-NOAA · 2024-08-23T17:53:34Z

@FernandoAndrade-NOAA You may start the UPP RTs on R&D machines. There should be no baseline updates with this PR.

jaymes-kenyon · 2024-08-23T18:00:34Z

@WenMeng-NOAA — It seems to be failing in getVariable.f and getIVariableN.f (both changed in this PR) due to "undefined reference to ext_ncd_get_var_ti_real_ ".

Specifically, getVariable.f and getIVariableN.f now use calls to ext_ncd_get_var_ti_real, which is a subroutine within NCEPlibs-wrf_io (https://github.com/NOAA-EMC/NCEPLIBS-wrf_io). Is this library available in the automatic testing setup?

WenMeng-NOAA · 2024-08-26T13:13:54Z

The UPP RTs have been successfully completed on WCOSS2 without baseline changes.

epic-cicd-jenkins · 2024-08-26T14:36:21Z

Orion UPP Regression Tests FAILED

epic-cicd-jenkins · 2024-08-26T14:36:27Z

Hera UPP Regression Tests FAILED

epic-cicd-jenkins · 2024-08-26T14:36:33Z

Hercules UPP Regression Tests FAILED

FernandoAndrade-NOAA · 2024-08-26T14:37:12Z

Please ignore the Jenkins failures, I will run manually.

FernandoAndrade-NOAA · 2024-08-26T15:38:41Z

RTs showed no changes across Hera, Hercules, and Orion. @WenMeng-NOAA I'm assuming we will be waiting for additional comments on the CI failures before the merge.

WenMeng-NOAA · 2024-08-26T18:30:13Z

@FernandoAndrade-NOAA I successfully built an UPP executable with GNU compiler on Hera (See /home/Wen.Meng/stmp2/jaymes/UPP/tests/compile_upp_guc.log). I would think the CI workflow should be updated with this PR. Since we have completed the UPP RTs on WCOSS2 and R&D machines, we could skip this CI issue and address it later. Please let me know your thoughts.

AlexanderRichert-NOAA · 2024-08-26T18:55:28Z

This appears to be a problem with the CMake configuration, and I would strongly recommend against merging as is. Specifically, the modifications to getIVariableN.f & getVariable.f use subroutines from wrf_io, but wrf_io isn't being found because BUILD_WITH_WRFIO is not enabled. Note that in CMakeLists.txt, lines 156-157, the inclusion of getIVariableN.f and getVariable.f is unconditional (specifically, it doesn't depend on BUILD_WITH_WRFIO being enabled). So I imagine you'll want to either make it conditional, or always use wrf_io.

SamuelTrahanNOAA · 2024-08-26T19:01:41Z

What happens if the CMakeLists change is reverted?

We could explicitly enable the build with wrfio when using MPAS.

WenMeng-NOAA · 2024-08-26T20:53:50Z

@AlexanderRichert-NOAA @SamuelTrahanNOAA Thanks for comments on narrowing down the checklist. In the CMakelist.txt, the default setting is "build without WRF-IO lib. @jaymes-kenyon Could you change "OFF" to "ON" in that line in your branch to see if it resolve the CI issue?

SamuelTrahanNOAA · 2024-08-26T21:04:03Z

I found the problem: io_int_stubs.f lacks ext_ncd_get_var_ti_real

The correct fix is to add the missing stub.

EDIT: I mistyped the subroutine name. It ends with _real, not _integer.

SamuelTrahanNOAA · 2024-08-26T21:16:10Z

@jaymes-kenyon - Please merge this pull request to your branch:

add ext_ncd_get_var_ti_real stub so UPP can compile without wrf-io library jaymes-kenyon/UPP#1

It'll add the missing stub so your branch can compile without the wrf-io library.

add `ext_ncd_get_var_ti_real` stub so UPP can compile without wrf-io library

SamuelTrahanNOAA · 2024-08-26T21:26:26Z

All three github tests passed. It seems the fixed worked. Do you want to retest other platforms? Or are the github tests sufficient for a fix of adding one stub?

WenMeng-NOAA · 2024-08-26T21:32:54Z

All three github tests passed. It seems the fixed worked. Do you want to retest other platforms? Or are the github tests sufficient for a fix of adding one stub?

@SamuelTrahanNOAA Thanks for the fix. The UPP RTs are conducted with WRF_IO enabled. I will rerun the RTs on WCOSS2 to confirm.

WenMeng-NOAA · 2024-08-26T22:07:18Z

Re-ran the UPP RTs on WCOSS2. No baseline update is needed.

WenMeng-NOAA · 2024-08-26T22:08:05Z

@FernandoAndrade-NOAA Please approval this PR for merging process.

SamuelTrahanNOAA

These code changes have the Official Sam Seal of Approval: 🦭👍

Keep up the good job, everyone!

WenMeng-NOAA · 2024-08-27T12:20:39Z

This PR is ready for merging.

* origin/develop: Updates to support MPAS applications (RRFS and HFIP) (NOAA-EMC#1029) Add 6-hrly max winds at 10m to SFS (NOAA-EMC#1019) Update jet.lua (NOAA-EMC#1027) Updates to support UPP use with MPAS (NOAA-EMC#1018) [develop]: Turn on WARN_AS_ERROR for Doxygen Documentation Builds (NOAA-EMC#1023) Sample PR for testing (NOAA-EMC#1007) Enable template 4-49 to obtain aerosol ensemble information (NOAA-EMC#1011) Relocate gfs/gefs/sfs control files (NOAA-EMC#990) update g2tmp on aws to 1.12.0 (NOAA-EMC#996) [develop]: Update requirements files (NOAA-EMC#994) Bump zipp from 3.17.0 to 3.19.1 in /doc/UserGuide (NOAA-EMC#991) remove stack-python load for Gaea build (NOAA-EMC#993) Bump certifi from 2024.2.2 to 2024.7.4 in /doc/UserGuide (NOAA-EMC#988) [develop]: Add Doxygen variable definitions to modules (NOAA-EMC#965) Read bucket either as an integer or float (NOAA-EMC#984) Update Orion module file for Rocky 9 upgrade (NOAA-EMC#987) UPP bugfixes for inline post and g2tmpl 1.12.0 compatibility (and number concentration updates) (NOAA-EMC#974) Sync the RRFS hotfix NOAA-EMC#967 and NOAA-EMC#968. (NOAA-EMC#979) Bump urllib3 from 2.2.1 to 2.2.2 in /doc/UserGuide (NOAA-EMC#981) Fix for hydrostatic runs - write out Omega and HGT from model output (NOAA-EMC#959) Bump requests from 2.31.0 to 2.32.0 in /doc/UserGuide (NOAA-EMC#963) Correct rt.sh paths (NOAA-EMC#957) Updates to use GRIB2 template 4.9 for QPFFFG and QPFARI variables in RRFS (NOAA-EMC#929) Add hourly averaged smoke PM2.5 and dust PM10 (NOAA-EMC#917) Update Namelist documentation for all namelist options (NOAA-EMC#940) update gtg code revision to v.4.16.3 (NOAA-EMC#954) Bump jinja2 from 3.1.3 to 3.1.4 in /doc/UserGuide (NOAA-EMC#953) Clean up labeling for 15-min accumulated fields (NOAA-EMC#948) Updates to the GSL exp1 ceiling diagnostic (NOAA-EMC#947) Additional RT log information (NOAA-EMC#941) add cmake load to modulefile/gaea.lua (NOAA-EMC#945) RRFS updates (NOAA-EMC#938) add module for AWS (NOAA-EMC#934)

jaymes-kenyon and others added 14 commits August 6, 2024 18:16

First draft of static control files for HFIP work. The hfip_postcntrl…

b90c146

….xml file is adapted from hrrr_postcntrl.xml.

Updates to INITPOST.F, including: (1) hard-coding prec_acc_dt, (2) si…

3715895

…lencing print statements for closely spaced model levels, and (3) adding comments regarding the "T" array

Silencing the "ZERO T" print statements in MDLFLD

c292731

Adding read-in of netCDF "_FillValue" when SUBMODELNAME=MPAS. For eac…

a4b7fe7

…h netCDF variable, grid locations are assigned SPVAL wherever "_FillValue" is found.

Adding static control files for RRFS-MPAS runs.

d3062b3

Merge branch 'NOAA-EMC:develop' into mpas

4372f3e

Removing previous changes to INITPOST.F and MDLFLD.f

527c253

Creating "INITPOST_MPAS" from "INITPOST"

3b489eb

Adding call to INITPOST_MPAS

dbc19b8

Adding INITPOST_MPAS to CMakeLists.txt

2105127

Modifying several write statements in INITPOST_MPAS

3f25532

Silencing print statements for model levels that are "too close" toge…

e3cc350

…ther.

Merge branch 'mpas' of https://github.com/jaymes-kenyon/UPP into mpas

bbc93f8

Removing temporary print statement in getVariable.f

d1bf361

jaymes-kenyon requested a review from WenMeng-NOAA as a code owner August 19, 2024 21:06

WenMeng-NOAA requested a review from MatthewPyle-NOAA August 20, 2024 18:23

WenMeng-NOAA added the MPAS label Aug 20, 2024

WenMeng-NOAA linked an issue Aug 20, 2024 that may be closed by this pull request

New UPP branch for MPAS-related development? #1008

Closed

Merge branch 'develop' into mpas

49fc204

WenMeng-NOAA added the Ready for Review This PR is ready for code review. label Aug 23, 2024

Merge branch 'NOAA-EMC:develop' into mpas

be87393

WenMeng-NOAA requested a review from FernandoAndrade-NOAA August 23, 2024 17:52

FernandoAndrade-NOAA added Jenkins-run-RT and removed Jenkins-run-RT labels Aug 26, 2024

FernandoAndrade-NOAA and others added 3 commits August 26, 2024 15:26

add hera rt log

70a06b6

add orion rt log

7f8e3cb

add hercules rt log

2b7fd82

add stub for ext_ncd_get_var_ti_real to compile without wrf-io lib

0f0cf1a

Merge pull request #1 from SamuelTrahanNOAA/mpas

760d529

add `ext_ncd_get_var_ti_real` stub so UPP can compile without wrf-io library

WenMeng-NOAA approved these changes Aug 26, 2024

View reviewed changes

SamuelTrahanNOAA approved these changes Aug 26, 2024

View reviewed changes

FernandoAndrade-NOAA approved these changes Aug 26, 2024

View reviewed changes

WenMeng-NOAA merged commit f5e4a18 into NOAA-EMC:develop Aug 27, 2024

WenMeng-NOAA mentioned this pull request Aug 27, 2024

Updates in develop branch since 07/2024 #985

Closed

Conversation

jaymes-kenyon commented Aug 19, 2024

Uh oh!

WenMeng-NOAA commented Aug 22, 2024

Uh oh!

jaymes-kenyon commented Aug 22, 2024

Uh oh!

WenMeng-NOAA commented Aug 23, 2024

Uh oh!

jaymes-kenyon commented Aug 23, 2024

Uh oh!

WenMeng-NOAA commented Aug 23, 2024

Uh oh!

WenMeng-NOAA commented Aug 23, 2024

Uh oh!

WenMeng-NOAA commented Aug 23, 2024

Uh oh!

jaymes-kenyon commented Aug 23, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

epic-cicd-jenkins commented Aug 26, 2024

Uh oh!

epic-cicd-jenkins commented Aug 26, 2024

Uh oh!

epic-cicd-jenkins commented Aug 26, 2024

Uh oh!

FernandoAndrade-NOAA commented Aug 26, 2024

Uh oh!

FernandoAndrade-NOAA commented Aug 26, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

AlexanderRichert-NOAA commented Aug 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

SamuelTrahanNOAA commented Aug 26, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

SamuelTrahanNOAA commented Aug 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

SamuelTrahanNOAA commented Aug 26, 2024

Uh oh!

SamuelTrahanNOAA commented Aug 26, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

WenMeng-NOAA commented Aug 26, 2024

Uh oh!

SamuelTrahanNOAA left a comment

Choose a reason for hiding this comment

Uh oh!

WenMeng-NOAA commented Aug 27, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

AlexanderRichert-NOAA commented Aug 26, 2024 •

edited

Loading

SamuelTrahanNOAA commented Aug 26, 2024 •

edited

Loading