Having issues running UPP V3.0

[laurenfrederick]

This has been moved over from a question in the UFS community discussion forum: https://github.com/orgs/ufs-community/discussions/21

From @jwraby:

I just compiled and am running UPP V3.0 on an HPC using an input single hour netdf wrfout file produced by WRF-CHEM V3.8. The output GRIB file WRFPRS_d03.27 contains the expected output fields which I specified in the wrf_cntrl.parm file (pressure and hybrid level output), but no diagnostic (SHELTER) level output. I checked the wrf_cntrl.parm file to confirm that the first bit is 1 for the SHELTER level output. The log file (unipost_d03.27.out) has the following error:
At line 134 of file MDL2AGL.f
Fortran runtime error: Index '-1' of dimension 2 of array 'lvls' below lower bound of 1
Error termination. Backtrace:
#0 0x83ea60 in mdl2agl_
at /p/home/jraby/UPPV3.0/src/unipost/MDL2AGL.f:134
https://github.com/orgs/ufs-community/discussions/1 0x718c70 in process_
at /p/home/jraby/UPPV3.0/src/unipost/PROCESS.f:95
https://github.com/orgs/ufs-community/discussions/2 0x7f2c94 in wrfpost
at /p/home/jraby/UPPV3.0/src/unipost/WRFPOST.f:793
https://github.com/orgs/ufs-community/discussions/3 0x7f41f6 in main
at /p/home/jraby/UPPV3.0/src/unipost/WRFPOST.f:134

In another test, I used the same UPP version and input a wrfout file produced by WRF V4.1. This WRF run was configured to use the newer hybrid mass sigma-pressure coordinate rather than the traditional sigma terrain following coordinate. It also uses vertical nesting. There are also model levels which are below 10 meters AGL in this configuration. This wrfout file is subhourly (15-min interval) with a forecast valid time given in the filename and the startdate given by dumping the file to text and looking for START_DATE. The lead time in hours is 11 based on the difference between the startdate and the date/time of the filename The run has an immediate error and the unipost_d01.11.out file has one error as follows:
"error opening /p/home/jraby/UPPV3.0/wrfprd/wrfout_d01_2022-07-05_23:00:00".

Updated information: Today I was able to resolve the error: "error opening /p/home/jraby/UPPV3.0/wrfprd/wrfout_d01_2022-07-05_23:00:00", by converting the wrfout file from netCDF-4 to 64-bit offset. The UPP is still unable to complete the run and has new errors.
From the unipost_d01.11.out file it shows that it can open/read the wrfout file which means that it can read it in 64-bit offset format.
But there are new errors:
Error reading T from /p/home/jraby/UPPV3.0/wrfprd/wrfout_d01_2022-07-05_23:00:00
ndim = 3
end_index(1) 250
end_index(2) 250
end_index(3) 60

There are similar worded errors for U, V, W, QVAPOR, and P_HYD.

There is also another error at the end of the log file as follows:
Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.

Backtrace for this error:
#0 0x2aaaabc6c3ff in ???
https://github.com/orgs/ufs-community/discussions/1 0x2aaaab5100cc in ???
https://github.com/orgs/ufs-community/discussions/2 0x57b644 in initpost_
at /p/home/jraby/UPPV3.0/src/unipost/INITPOST.f:434
https://github.com/orgs/ufs-community/discussions/3 0x7f200c in wrfpost
at /p/home/jraby/UPPV3.0/src/unipost/WRFPOST.f:687
https://github.com/orgs/ufs-community/discussions/4 0x7f41f6 in main
at /p/home/jraby/UPPV3.0/src/unipost/WRFPOST.f:134
./run_unipost_minutes: line 406: 281950: ${RUN_COMMAND} > unipost_${domain}.${fhr}.out 2>&1

Is there an explanation for the inconsistent results obtained with the two wrfout files? Any troubleshooting assistance for the specific errors described would also be appreciated.

[gspetro-NOAA]

Hello @jwraby,

Support for WRF ended a few years ago, and we unfortunately don’t have the resources to dig into WRF-related issues. Additionally, UPP v3.0 is pretty old. Is there a reason you need to use that version particularly? The last version of the UPP to support WRF was v4.1.0, and you can find a little more about that in Discussion #614. You may also find some useful information in #679, which talks about some of the changes that are required for running WRF with UPP v11.0.0 (the latest version of the UPP). In general, you may find support on the WRF/MPAS forum for running the UPP with WRF.

Not sure what your current WRF work entails, but it might be worth switching to the Unified Forecast System (UFS). There is a lot more support available for the UFS, and the UPP is included in certain application workflows, like the Short-Range Weather (SRW) Application workflow.

Best,
Gillian Petro | Advanced User Support (AUS) Team

[jwraby]

Thanks for your reply and the information, Gillian. We are using UPP V3.0 which we have modified to support our research that uses WRF for simulations on sub-kilometer grids in complex terrain. I think my primary issue is not properly/successfully compiling the UPP and I'm working on that aspect hoping that re-compiling it with the required libraries, etc. will result in a successful build. Perhaps then, I'll be able to run the UPP on the WRF output with better results. R/ John Mr. John W. Raby Devcom Army Research Laboratory White Sands Missile Range, NM 88002 (575) 678-2004 DSN 258-2004 FAX (575) 678-1230 DSN 258-1230 Email: ***@***.*** From: Gillian Petro ***@***.***> Sent: Wednesday, August 2, 2023 8:26 AM To: NOAA-EMC/UPP ***@***.***> Cc: Raby, John W CIV USARMY DEVCOM ARL (USA) ***@***.***>; Mention ***@***.***> Subject: [URL Verdict: Neutral][Non-DoD Source] Re: [NOAA-EMC/UPP] Having issues running UPP V3.0 (Discussion #752) Hello @jwraby <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2Fjwraby&data=05%7C01%7Cjohn.w.raby2.civ%40army.mil%7C286d36db330947c046f f08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43%7C0%7C0%7C63826583264504629 2%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haW wiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=gMv7Z9%2FFp9u26AibiKS9jWGL54D1Ofa6zFN oNADPHvw%3D&reserved=0> , Support for WRF ended a few years ago, and we unfortunately don't have the resources to dig into WRF-related issues. Additionally, UPP v3.0 is pretty old. Is there a reason you need to use that version particularly? The last version of the UPP to support WRF was v4.1.0, and you can find a little more about that in this discussion: #614 <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2FNOAA-EMC%2FUPP%2Fdiscussions%2F614&data=05%7C01%7Cjohn.w.raby2.civ%40ar my.mil%7C286d36db330947c046ff08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43 %7C0%7C0%7C638265832645046292%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLC JQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=TYv2leeiu aD6Fg6rNVvoFz5S%2BgDM6Fr2OQAGYsCGTyU%3D&reserved=0> . You may also find some useful information in #679 <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2FNOAA-EMC%2FUPP%2Fdiscussions%2F679&data=05%7C01%7Cjohn.w.raby2.civ%40ar my.mil%7C286d36db330947c046ff08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43 %7C0%7C0%7C638265832645046292%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLC JQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=PCV0X2Dnp RplWd8Wn%2FYpS%2F8laGw7PC7HCy%2Frc3w5Bzc%3D&reserved=0> , which talks about some of the changes that are required for running WRF with UPP v11.0.0 (the latest version of the UPP). In general, you may find support on the WRF/MPAS forum <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fforum.mm m.ucar.edu%2Fforums%2Fwrf-upp.133%2F&data=05%7C01%7Cjohn.w.raby2.civ%40army. mil%7C286d36db330947c046ff08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43%7C 0%7C0%7C638265832645046292%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQI joiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=ElDNtqlRE1Kv E0pVJeR7xAVoGMJH9X7gA3jnpEM%2B5aM%3D&reserved=0> for running the UPP with WRF. Not sure what your current WRF work entails, but it might be worth switching to the Unified Forecast System (UFS). There is a lot more support available for the UFS, and the UPP is included in certain application workflows, like the Short-Range Weather (SRW) Application <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2Fufs-community%2Fufs-srweather-app&data=05%7C01%7Cjohn.w.raby2.civ%40arm y.mil%7C286d36db330947c046ff08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43% 7C0%7C0%7C638265832645046292%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJ QIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=QgdPqw4Jgk Rg3jad0OdZ6ILsovCURH7E3aKN6v7p3iE%3D&reserved=0> workflow. Best, Gillian Petro | Advanced User Support (AUS) Team - Reply to this email directly, view it on GitHub <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2FNOAA-EMC%2FUPP%2Fdiscussions%2F752%23discussioncomment-6617403&data=05% 7C01%7Cjohn.w.raby2.civ%40army.mil%7C286d36db330947c046ff08db9364a0c5%7Cfae6 d70f954b481192b60530d6f84c43%7C0%7C0%7C638265832645046292%7CUnknown%7CTWFpbG Zsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3 000%7C%7C%7C&sdata=7GiGAjTO8PuDbK75P0Smf%2B7mUNuwTfPpbC7GTVB6aeY%3D&reserved =0> , or unsubscribe <https://usg01.safelinks.protection.office365.us/?url=https%3A%2F%2Fgithub.c om%2Fnotifications%2Funsubscribe-auth%2FAUX5FH74IYQAOWQJ7QQD773XTJPQVANCNFSM 6AAAAAA26T2OWU&data=05%7C01%7Cjohn.w.raby2.civ%40army.mil%7C286d36db330947c0 46ff08db9364a0c5%7Cfae6d70f954b481192b60530d6f84c43%7C0%7C0%7C63826583264504 6292%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1 haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=ejCa%2BC3yPYm7z%2B4CMie%2Byv152neT LKwPNaQSBRVn9AU%3D&reserved=0> . You are receiving this because you were mentioned. <https://github.com/notifications/beacon/AUX5FHY3XDDWMHATAMMG7P3XTJPQVA5CNFS M6AAAAAA26T2OWWWGG33NNVSW45C7OR4XAZNRIRUXGY3VONZWS33OINXW23LFNZ2KUY3PNVWWK3T UL5UWJTQAMT4TW.gif> Message ID: ***@***.*** ***@***.***> >