Changes to regression test files for recent changes to RRTMG and RRTMGP. #262

dustinswales · 2020-11-02T22:18:44Z

This PR contains changes to regression test files relating to recent changes to RRTMG and RRTMGP.

The gfs_physics namelist was modified to accommodate new functionality. Changes to the RT setup for RRTMG and RRTMGP were modified.

Also, GP has input files that need to be updated. These are provided with the rte-rrtmgp code, ccpp/physics/physics/rte-rrtmgp/rrtmgp/data/. Is there a way that we can set up the RT's to use this location when staging input files for the RTs? Just one less place for confusion.

This PR is dependent on:

…r-model into ufs-develop-00a51e1

climbfuji

@dustinswales can you turn on the RRTMGP regression tests again in tests/rt.conf and tests/rt_gnu.conf ? Do they pass?

Also, please update .gitmodules temporarily to point to your fv3atm fork and your branch develop-emc-b955f81? Thanks.

dustinswales · 2020-11-06T18:17:26Z

@climbfuji
The GP tests fail the RT's because the GP input data that gets staged needs updating. (This should really be pulled in directly from the rte-rrtmgp code) I don't know where to look to do this. But if I run the RTs and go into the failed GP RTs run directories and import the updated files, it runs. Of course there is no baseline to compare to for the v15p2 or v16beta w/ GP.

climbfuji · 2020-11-06T19:01:25Z

@climbfuji
The GP tests fail the RT's because the GP input data that gets staged needs updating. (This should really be pulled in directly from the rte-rrtmgp code) I don't know where to look to do this. But if I run the RTs and go into the failed GP RTs run directories and import the updated files, it runs. Of course there is no baseline to compare to for the v15p2 or v16beta w/ GP.

That's good to know! I can help with staging the new data and creating the baselines. Please point me to the files that need to be updated and I can take care of it.

dustinswales · 2020-11-06T20:54:00Z

@climbfuji
The files used for the gas-optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/rrtmgp/data.
For the GP cloud optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/extensions/cloud_optics

climbfuji · 2020-11-06T21:00:40Z

@climbfuji
The files used for the gas-optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/rrtmgp/data.
For the GP cloud optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/extensions/cloud_optics

Where should these files go? Replace the contents of /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/?

[Dom.Heinzeller@hfe09 INTEL]$ ls -l /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/
total 46444
-rw-r--r-- 1 emc.nemspara nems    32212 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-lw.nc
-rw-r--r-- 1 emc.nemspara nems    28888 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-sw.nc
-rw-r--r-- 1 emc.nemspara nems 32530992 Oct 20 00:43 rrtmgp-data-lw-g256-2018-12-04.nc
-rw-r--r-- 1 emc.nemspara nems 14954928 Oct 20 00:43 rrtmgp-data-sw-g224-2018-12-04.nc

(of course not replace , but replace in the next version of the baseline)

dustinswales · 2020-11-06T21:08:39Z

@climbfuji
The files used for the gas-optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/rrtmgp/data.
For the GP cloud optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/extensions/cloud_optics

Where should these files go? Replace the contents of /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/?
[Dom.Heinzeller@hfe09 INTEL]$ ls -l /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/
total 46444
-rw-r--r-- 1 emc.nemspara nems    32212 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-lw.nc
-rw-r--r-- 1 emc.nemspara nems    28888 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-sw.nc
-rw-r--r-- 1 emc.nemspara nems 32530992 Oct 20 00:43 rrtmgp-data-lw-g256-2018-12-04.nc
-rw-r--r-- 1 emc.nemspara nems 14954928 Oct 20 00:43 rrtmgp-data-sw-g224-2018-12-04.nc
(of course not replace , but replace in the next version of the baseline)

This is correct.

climbfuji · 2020-11-06T21:11:40Z

@climbfuji
The files used for the gas-optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/rrtmgp/data.
For the GP cloud optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/extensions/cloud_optics

Where should these files go? Replace the contents of /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/?
[Dom.Heinzeller@hfe09 INTEL]$ ls -l /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/
total 46444
-rw-r--r-- 1 emc.nemspara nems    32212 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-lw.nc
-rw-r--r-- 1 emc.nemspara nems    28888 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-sw.nc
-rw-r--r-- 1 emc.nemspara nems 32530992 Oct 20 00:43 rrtmgp-data-lw-g256-2018-12-04.nc
-rw-r--r-- 1 emc.nemspara nems 14954928 Oct 20 00:43 rrtmgp-data-sw-g224-2018-12-04.nc
(of course not replace , but replace in the next version of the baseline)
This is correct.

Thanks, will do so.

…r-model into ufs-develop-00a51e1

climbfuji · 2020-11-10T17:07:31Z

@climbfuji
The files used for the gas-optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/rrtmgp/data.
For the GP cloud optics: https://github.com/earth-system-radiation/rte-rrtmgp/tree/dtc/ccpp/extensions/cloud_optics

Where should these files go? Replace the contents of /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/?
[Dom.Heinzeller@hfe09 INTEL]$ ls -l /scratch1/NCEPDEV/nems/emc.nemspara/RT/NEMSfv3gfs/develop-20201103/INTEL/FV3_input_data_RRTMGP/
total 46444
-rw-r--r-- 1 emc.nemspara nems    32212 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-lw.nc
-rw-r--r-- 1 emc.nemspara nems    28888 Oct 20 00:43 rrtmgp-cloud-optics-coeffs-sw.nc
-rw-r--r-- 1 emc.nemspara nems 32530992 Oct 20 00:43 rrtmgp-data-lw-g256-2018-12-04.nc
-rw-r--r-- 1 emc.nemspara nems 14954928 Oct 20 00:43 rrtmgp-data-sw-g224-2018-12-04.nc
(of course not replace , but replace in the next version of the baseline)
This is correct.
Thanks, will do so.

@dustinswales FYI, finally doing this on orion because hera remains unusable. I see that the two cloud optics coeffs files are identical between what is currently in the baseline and what is in the rte-rrtmgp repository (at least at the hash currently used inside ccpp-physics), while the two other files are different:

Orion-login-4[9] dheinzel$ diff ./FV3/ccpp/physics/physics/rte-rrtmgp/extensions/cloud_optics/rrtmgp-cloud-optics-coeffs-sw.nc /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-cloud-optics-coeffs-sw.nc

Orion-login-4[10] dheinzel$ diff ./FV3/ccpp/physics/physics/rte-rrtmgp/extensions/cloud_optics/rrtmgp-cloud-optics-coeffs-lw.nc /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-cloud-optics-coeffs-lw.nc

Orion-login-4[12] dheinzel$ diff ./FV3/ccpp/physics/physics/rte-rrtmgp/rrtmgp/data/rrtmgp-data-sw-g224-2018-12-04.nc /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-data-sw-g224-2018-12-04.nc
Binary files ./FV3/ccpp/physics/physics/rte-rrtmgp/rrtmgp/data/rrtmgp-data-sw-g224-2018-12-04.nc and /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-data-sw-g224-2018-12-04.nc differ

Orion-login-4[13] dheinzel$ diff ./FV3/ccpp/physics/physics/rte-rrtmgp/rrtmgp/data/rrtmgp-data-lw-g256-2018-12-04.nc /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-data-lw-g256-2018-12-04.nc
Binary files ./FV3/ccpp/physics/physics/rte-rrtmgp/rrtmgp/data/rrtmgp-data-lw-g256-2018-12-04.nc and /work/noaa/gmtb/dheinzel/TMP_NEW_BASELINE_RRTMGP_20201110/FV3_input_data_RRTMGP/rrtmgp-data-lw-g256-2018-12-04.nc differ

Can you confirm? in this case I only need to update the latter two files.

dustinswales · 2020-11-10T17:13:16Z

@climbfuji
This is correct. The cloud-optics files are unchanged.

climbfuji · 2020-11-10T17:14:14Z

@climbfuji
This is correct. The cloud-optics files are unchanged.

Ok, thanks. Will try to re-enable the RRTMGP regression tests and use those files. Any other changes I need to make for the regression tests, or have you done everything in ufs-weather-model?

dustinswales · 2020-11-10T17:17:02Z

@climbfuji
I think I got everything for the RTs in the UFS.

climbfuji · 2020-11-10T23:45:30Z

@climbfuji
I think I got everything for the RTs in the UFS.

We'll need to make a bunch of changes wrt iovr_{lw,sw} for some of the regression tests. I'll send a PR your way soon!

…s-20201111 ufs-develop-00a51e1 RRTMGP bugfixes

…/ufs-weather-model into ufs-develop-00a51e1

climbfuji · 2020-11-13T00:09:57Z

A note on the regression testing. Compared to the existing baseline, the following tests fail (therefore a new baseline):

These tests fail because of a missing baseline to compare against (new tests):

fv3_ccpp_gfs_v15p2_RRTMGP 041 failed in check_result
fv3_ccpp_gfs_v16beta_RRTMGP 042 failed in check_result
fv3_ccpp_gfs_v15p2_RRTMGP_debug 051 failed in check_result
fv3_ccpp_gfs_v16beta_RRTMGP_debug 052 failed in check_result

These tests fail because of a small change to diagnostic cloud output when HWRF RRTMG (iovr=4) is used. This is ok, the changes are only in the diagnostic output, small and only in one field (tcdc_avebndcl --> IntDiag(nb)%fluxr(:,18))

fv3_ccpp_HAFS_v0_hwrf_thompson 047 failed in check_result
fv3_ccpp_HAFS_v0_hwrf_thompson_debug 058 failed in check_result
fv3_ccpp_esg_HAFS_v0_hwrf_thompson 048 failed in check_result
fv3_ccpp_esg_HAFS_v0_hwrf_thompson_debug 059 failed in check_result

…coss_cray, wcoss_dell_p3

Regression test logs for PR 262

…/ufs-weather-model into ufs-develop-00a51e1

junwang-noaa

need to revert fv3 after fv3 PR is committed.

climbfuji · 2020-11-13T16:39:05Z

Submodule pointer looks ok, ready to merge!

…ommunity#262) Use the GFDL dycore repository dev/emc. The GFDL dev/emc branch of dycore was synchronized with the GFDL master, which included the FV3 202101 upgrade except inline GFDL_MP : https://github.com/NOAA-GFDL/GFDL_atmos_cubed_sphere/releases/tag/FV3-202101-public

* Print usage() message if machine name is not passed. * Bug fix with macos modulefiles.

dustinswales added 3 commits October 30, 2020 17:27

Changes to regression test files for recent changes to RRTMG and RRTMGP.

f978e9d

Merge branch 'develop' of https://github.com/ufs-community/ufs-weathe…

00adc4a

…r-model into ufs-develop-00a51e1

Synced with upstream

656b9c6

This was referenced Nov 6, 2020

Improvements to radiation code NCAR/ccpp-physics#514

Merged

Infrastructure improvements for radiation NOAA-EMC/fv3atm#193

Merged

climbfuji requested review from DusanJovic-NOAA, climbfuji, junwang-noaa and SMoorthi-emc and removed request for DusanJovic-NOAA November 6, 2020 17:38

climbfuji requested changes Nov 6, 2020

View reviewed changes

Updated .gitmodules.

728fb71

Merge branch 'develop' of https://github.com/ufs-community/ufs-weathe…

59ccbba

…r-model into ufs-develop-00a51e1

Removed redundant importing of kind_phys in physics

5f2d742

dustinswales added 2 commits November 10, 2020 21:53

Apply patch from Dom

61eff43

add units to index_for_diagnostic_printout metadata

ec1fee4

climbfuji added 2 commits November 11, 2020 05:31

Fix submodule pointers for CMEPS and NEMS

b2bdc41

Replace iovr_{lw,sw} with iovr for HWRF regression tests

aa24555

climbfuji and others added 4 commits November 12, 2020 10:11

Regression test logs for orion.intel

f733094

Merge pull request #8 from climbfuji/ufs-weather-model-rrtmgp-bugfixe…

c5a11dc

…s-20201111 ufs-develop-00a51e1 RRTMGP bugfixes

Bugfix in GP sampling routines.

f290cc8

Merge branch 'ufs-develop-00a51e1' of https://github.com/dustinswales…

9199ed4

…/ufs-weather-model into ufs-develop-00a51e1

climbfuji approved these changes Nov 13, 2020

View reviewed changes

climbfuji added 2 commits November 12, 2020 17:13

Increase number of nodes for test fv3_ccpp_gocart_clm on Cheyenne

f0d3736

Regression test logs for cheyenne.gnu, cheyenne.intel, orion.intel, w…

ad1af47

…coss_cray, wcoss_dell_p3

DusanJovic-NOAA approved these changes Nov 13, 2020

View reviewed changes

climbfuji and others added 5 commits November 12, 2020 20:26

Regression test log for hera.gnu

d564f22

Regression test log for hera.intel

2d3a9c0

Merge pull request #9 from climbfuji/regtest_logs_rrtmgp

409abc8

Regression test logs for PR 262

Revert submodule pointer for ccpp-physics. Revert FV3 .gitmodules.

db61b51

Merge branch 'ufs-develop-00a51e1' of https://github.com/dustinswales…

0474acb

…/ufs-weather-model into ufs-develop-00a51e1

junwang-noaa approved these changes Nov 13, 2020

View reviewed changes

Revert submodule pointer for FV3. Revert UFS .gitmodules.

c077a7c

DusanJovic-NOAA merged commit 2e25df5 into ufs-community:develop Nov 13, 2020

dustinswales deleted the ufs-develop-00a51e1 branch November 18, 2020 21:33

epic-cicd-jenkins pushed a commit that referenced this pull request Apr 17, 2023

Print usage if machine name is not passed to lmod-setup.sh/csh. (#262)

c118a5f

* Print usage() message if machine name is not passed. * Bug fix with macos modulefiles.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Changes to regression test files for recent changes to RRTMG and RRTMGP. #262

Changes to regression test files for recent changes to RRTMG and RRTMGP. #262

dustinswales commented Nov 2, 2020 •

edited

Loading

climbfuji left a comment

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

climbfuji commented Nov 10, 2020

dustinswales commented Nov 10, 2020

climbfuji commented Nov 10, 2020

dustinswales commented Nov 10, 2020

climbfuji commented Nov 10, 2020

climbfuji commented Nov 13, 2020

junwang-noaa left a comment

climbfuji commented Nov 13, 2020

Changes to regression test files for recent changes to RRTMG and RRTMGP. #262

Changes to regression test files for recent changes to RRTMG and RRTMGP. #262

Conversation

dustinswales commented Nov 2, 2020 • edited Loading

climbfuji left a comment

Choose a reason for hiding this comment

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

dustinswales commented Nov 6, 2020

climbfuji commented Nov 6, 2020

climbfuji commented Nov 10, 2020

dustinswales commented Nov 10, 2020

climbfuji commented Nov 10, 2020

dustinswales commented Nov 10, 2020

climbfuji commented Nov 10, 2020

climbfuji commented Nov 13, 2020

junwang-noaa left a comment

Choose a reason for hiding this comment

climbfuji commented Nov 13, 2020

dustinswales commented Nov 2, 2020 •

edited

Loading