[BenjaminBlake-NOAA]

This PR resolves issue #1629

A user reported that the values for Lifted Index (UPP ID 030) in the RRFS output have an apparent temperature offset, and thus are not consistent with the expected temperature-difference scale. In CLDRAD.f, it was discovered that TFRZ (273.15 K) is being added to the Lifted Index. Here are the values of Lifted Index before and after my code changes:

143:717243998:vt=2026032700:500-1000 mb:36 hour fcst:LFTX Surface Lifted Index [K]:
    ndata=14452641:undef=0:mean=281.216:min=262.3:max=314.9
143:717243998:vt=2026032700:500-1000 mb:36 hour fcst:LFTX Surface Lifted Index [K]:
    ndata=14452641:undef=0:mean=8.05802:min=-10.8:max=41.8

@WenMeng-NOAA @ShellieRowe-NOAA I have RRFS run directories here on WCOSS2 and Ursa if you'd like to take a look:

WCOSS2: /lfs/h2/emc/vpppg/noscrub/Benjamin.Blake/upp-pr1630-lftx/rrfs_2026032512
Ursa: /scratch4/NCEPDEV/stmp/Benjamin.Blake/upp-pr1630-intel/rrfs_2026032512

LFTX is in the 2DFLD and NATLEV GRIB2 files.

Tagging @AndrewBenjamin-NOAA for awareness.

[ShellieRowe-NOAA]

I reviewed and tested PR #1630 to confirm the correction to the RRFS Lifted Index calculation. I checked out the PR on Ursa, compiled UPP, and ran the RRFS regression test, which completed successfully. I reviewed the test log and compared the output files with the original results. The pressure-level file was unchanged, while the other two showed differences consistent with the expected Lifted Index correction. I also examined the difference files and found that the Lifted Index values changed by approximately 273.15, as expected. The comparison flagged a specific humidity field, so I checked that field in both the original and new output files. Both contained undefined values, which explained the warning. Overall, the results support the intended correction, and I did not find any unexpected numerical differences in the comparisons.

…

On Wed, Oct 21, 2026 at 2:19 PM BenjaminBlake-NOAA ***@***.***> wrote: @BenjaminBlake-NOAA <https://github.com/BenjaminBlake-NOAA> requested your review on: NOAA-EMC/UPP#1630 <#1630> Fix Lifted Index calculation in RRFS. — Reply to this email directly, view it on GitHub <#1630?email_source=notifications&email_token=CNF5QAD7A3R2UWLFCTWNLLD5VED3RA5CNFSNUABQM5UWIORPF5TWS5BNNB2WEL2JONZXKZKFOZSW45CON52GSZTJMNQXI2LPNYXTGMZYHAYDKOJTGM2DBJTSMVQXG33OWBZGK5TJMV3V64TFOF2WK43UMVSKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#event-33880593340>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CNF5QAEHVIYZ6VXUHN4SKDL5VED3RAVCNFSNUABFKJSXA33TNF2G64TZHMYTSNRUGUYDAMZXHNEXG43VMU5TKOJWG4YDAMBYG4Z2C5QC> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/CNF5QAH5YTGHCUVXRHZ7YNT5VED3RA5CNFSNUABQM5UWIORPF5TWS5BNNB2WEL2JONZXKZKFOZSW45CON52GSZTJMNQXI2LPNYXTGMZYHAYDKOJTGM2DBJTSMVQXG33OWBZGK5TJMV3V64TFOF2WK43UMVSKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android <https://github.com/notifications/mobile/android/CNF5QAGVVUN4INI745ZB47L5VED3RA5CNFSNUABQM5UWIORPF5TWS5BNNB2WEL2JONZXKZKFOZSW45CON52GSZTJMNQXI2LPNYXTGMZYHAYDKOJTGM2DBJTSMVQXG33OWBZGK5TJMV3V64TFOF2WK43UMVSKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because your review was requested.Message ID: ***@***.***>

[BenjaminBlake-NOAA]

@ShellieRowe-NOAA Thanks for testing and reviewing the changes! I think we can ignore the flagged specific humidity field since all of the values are undefined.

[WenMeng-NOAA]

@BenjaminBlake-NOAA Your changes look good to me and my test results are consistent with yours.

[WenMeng-NOAA]

@gspetro-NOAA This PR is ready for the UPP RTs on RDHPCS. Please refer to @BenjaminBlake-NOAA 's testing:
/scratch4/NCEPDEV/stmp/Benjamin.Blake/upp-pr1630-intel/rrfs_2026032512

The only baseline change is the rrfs test.

[BenjaminBlake-NOAA]

@WenMeng-NOAA @ShellieRowe-NOAA Here are several plots illustrating the differences in the lifted index.

Before changes (all values are greater than 260K):

After changes:

And here is the Best (4 layer) Lifted Index which compares well with the Lifted Index:

[gspetro-NOAA]

Just an update: Baselines have been updated successfully for Ursa (llvm & intel classic) and Hercules, and everything passes on rerun. For some reason, the baseline creation run on Orion keeps timing out. I'm hoping with my latest increase in wallclock time, this next run will pass, and then the test time will be normal once the baselines are updated, but I have to wait for the run to complete.

[BenjaminBlake-NOAA]

@gspetro-NOAA Sounds good, thanks!

[gspetro-NOAA]

I updated the BL files for rrfs and all tests are now passing on all relevant systems. I unfortunately forgot to push the Orion logs from before I updated the baselines, so I pushed the ones from after. The rt_verify.pl printout from before the BL update showed the rrfs test running significantly overtime on Orion, but after the baseline update, RTs ran normally.

Since I overwrote those logs, here is the printout:

#-----------------
There were no errors detected in the work directory
#-----------------

Checking timing results for all regresssion tests...

#-----------------
gefsv13 is 67   seconds UNDER time
nmmb    is 91   seconds UNDER time
hfip    is 85   seconds UNDER time
gfs     is 741  seconds UNDER time
mpas    is 82   seconds UNDER time
rrfs    is 2027 seconds OVER  time
hafs    is 17   seconds UNDER time
sfs     is 53   seconds UNDER time
rap     is 16   seconds UNDER time
gcafs   is 23   seconds UNDER time
gefsv12 is 36   seconds UNDER time
hrrr    is 256  seconds UNDER time
aqm     is 15   seconds UNDER time
3drtma  is 66   seconds UNDER time
aigfs   is 17   seconds UNDER time
#-----------------

Checking for baseline files that have changed...

#-----------------
rrfs:
	1) /work2/noaa/epic/gpetro/orion/RTs/upp-rts/1630/ci/rundir/upp-ORION/rrfs_2026032512/NATLEV36.tm00
[sudo] password for gpetro: 
	   ***New baseline file copied successfully***
	2) /work2/noaa/epic/gpetro/orion/RTs/upp-rts/1630/ci/rundir/upp-ORION/rrfs_2026032512/2DFLD36.tm00
	   ***New baseline file copied successfully***