Update Spack Stack Location for NOAA Cloud #294
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
EdwardSnyder-NOAA
requested review from
mkavulich,
gsketefian,
JeffBeck-NOAA,
RatkoVasic-NOAA,
BenjaminBlake-NOAA,
ywangwof,
chan-hoo,
panll,
christinaholtNOAA,
christopherwharrop-noaa,
danielabdi-noaa,
mark-a-potts,
jkbk2004,
willmayfield,
dmwright526,
gspetro-NOAA and
natalie-perlin
as code owners
mark-a-potts
approved these changes
Jun 3, 2022
gspetro-NOAA
approved these changes
Jun 4, 2022
|
The builds failed on Hera and Jet. Do these Jenkins tests run on the cloud platforms? |
@kgerheiser Since this adjustment is only for the noaacloud build modulefile, it shouldn't matter for this PR whether builds pass or fail on Hera and Jet, no? (Of course, we'd want to see why builds are failing there, but it shouldn't be due to this PR, since nothing affecting Hera or Jet was changed...) |
|
@kgerheiser - we have the capability to run tests on the cloud platforms through Jenkins, we just don't have the cloud tests integrated into GitHub yet. I was able to build the SRW App on AWS and run the comprehensive tests on it through Jenkins on Friday. The Miniconda3 used for the AWS build was missing on GCP and Azure, but I've resolved that and ran a single E2E on GCP, which was a successful run. I'm still having spack issues on Azure but you are already aware of that. @gspetro-NOAA - Both Hera and Jet are crashing through our Jenkins pipeline with other PRs. I believe the Hera issue is related to a variable in |
|
Fixed the meta-moduels on Azure. And for Miniconda this should be loaded beforehand as well: And that will bring And then when loading |
|
Thanks @kgerheiser! The SRW App builds and runs on AWS, Azure, and GCP with the new spack stack location. More work is needed to integrate the Jenkins E2E test pipeline with Azure, GCP, Jet, and Hera, which will be addressed in a future PRs. |
|
Thanks for running those tests. This is some good real world testing. |
mkavulich
added a commit
to mkavulich/ufs-srweather-app
that referenced
this pull request
Aug 26, 2022
* Fix to post flat file. * Create MET and METplus config files under ush/templates/parm * Added script to pull and reorg ccpa data. Added a script to run gridstat with METplus. Updated MET and METplus config files. * Added new jjob for running grid-stat vx. Updated setup.sh to include grid-stat vx. Updated run_gridstatvx script. * Fixed typo on script name from ksh to sh * Moved some hard coded items out from the script to the XML * Updates to get METplus to run with fewer hard-coded paths. * Updates to add grid-stat task to XML generation. * Bug fixes for adding grid-stat to XML generation * Updates to remove hard-coded paths in config files * Change log dir to put master_metplus log file with other logs under log/, rather than default logs/. * Updates to generate xml without hard-coded paths for MET * Add hera gridstat module file * Add METplus point-stat task for both sfc and upper air * Small tweaks to remove hard coded paths and add some flexibility * Updates for adding point-stat into auto-generated xml * Add in function to set point-stat task to FALSE * Final tweaks to get it to generate the xml correctly * Minor updates to run ensure 0,6,12,18 * Tweaks to var list for Point-Stat * Add METplus settings to config_defaults * Move quote for end of settings and fix extra comment. * Fix typos to populate templates correctly * Updated to include SCRIPTSDIR and other MET specific settings along with updates to FHR syntax * Update module loads on hera * Fixed comment for BOTH_VARn_THRESH to avoid syntax issues * Added files to run grid_stat for a variety of accumulation intervals, including 3, 6, and 24h * Added module load hpss * Remove module load informatino from these scripts * Updated the method of turning on/off vx tasks using jinja template if statement * Remove commented out lines of code. Fixed typo. Removed gen_wflow.out file. * Updated pull scripts to have file names dependent on date to pull from HPSS. Updated to export a few more local variables that METplus conf needed in scripts. Updated workflow to use service queue (for now) to for 1h grid_stat and point_stat run and default for 3+h accumulation grid_stat runs) * moved common_hera.conf to common.conf - no platform specific information included that needs to be handled. * Remove common_hera.conf * Add scripts to pull and process MRMS data from NOAA HPSS * Updates for REFC vx tasks * updates to obs pull scripts * Update for adding in reflectivity verification using MRMS analyses and updating name of model output to RRFS rather than HRRR * Updates to account for CCPA issues on HPSS - day off for 00-05 UTC directories * Verification mods to feature/add metplus (#1) * Remove unused/outdated code (ufs-community#313) ## DESCRIPTION OF CHANGES: * In setup.sh and generate_FV3LAM_wflow.sh, remove temporary codes that fix bugs in the FV3_GFS_2017_gfdlmp_regional suite definition file because those bugs have been fixed (in the ufs-weather-model repo). * In setup.sh, remove block of code that is no longer necessary because chgres_cube can now initialize from external model data with either 4 or 9 soil levels, and run with LSMs of either 4 or 9 soil levels. * Remove modifications to LD_LIBRARY_PATH in exregional_run_fcst.sh. * For the make_ics and make_lbcs tasks, move the setting of APRUN and other machine-specific actions from the J-job to the ex-script in order to be consistent with the other workflow tasks. * Fix indentation and edit comments. * Remove unused file load_fv3gfs_modules.sh. ## TESTS CONDUCTED: Ran two WE2E tests on hera, new_ESGgrid and new_GFDLgrid: * new_ESGgrid uses the FV3_GFS_2017_gfdlmp_regional suite. The test was successful. * new_GFDLgrid uses the FV3_GFS_2017_gfdlmp suite. The test was successful. ## ISSUE (optional): This resolves issue ufs-community#198. * Add and call a function that checks for use of Thompson microphysics parameterization in the SDF and if so, adjusts certain workflow arrays to contain the names and other associated values of the fixed files needed by this parameterization so that those files are automatically copied and/or linked to. (ufs-community#319) ## DESCRIPTION OF CHANGES: Add and call a function that checks for use of Thompson microphysics parameterization in the suite definition file (SDF). If not, do nothing. If so, add to the appropriate workflow arrays the names and other associated values of the fixed files needed by this parameterization so that they are automatically copied and/or linked to instead of being regenerated from scratch in the run_fcst task. ## TESTS CONDUCTED: On hera, ran two WE2E tests, one in NCO mode (nco_RRFS_CONUS_25km_HRRRX_RAPX) and the other in community mode (suite_FV3_GSD_v0). These use suites FV3_GSD_SAR and FV3_GSD_v0, respectively, and both of these call Thompson microphysics. Both succeeded. ## ISSUE (optional): This PR resolves issue ufs-community#297. * RRFS_v1beta SDF changes after reverting from GSL to GFS GWD suite (ufs-community#322) (ufs-community#327) ## DESCRIPTION OF CHANGES: Removed checks on the RRFS_v1beta SDF implemented for use with the GSL GWD suite (now uses the GFS GWD suite). No longer copies staged orography files necessary for the GSL GWD suite. ## TESTS CONDUCTED: Runs to completion on Hera. End-to-end runs DOT_OR_USCORE and suite_FV3_RRFS_v1beta succeeded on Cheyenne. Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Adjust RRFS_CONUS_... grids (ufs-community#294) ## DESCRIPTION OF CHANGES: * Adjust RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grid parameters so that: * All grids, including their 4-cell-wide halos, lie completely within the HRRRX domain. * All grids have dimensions nx and ny that factor "nicely", i.e. they don't have factors greather than 7. * The write-component grids corresponding to these three native grids cover as much of the native grids as possible without going outside of the native grid boundaries. The updated NCL scripts (see below) were used to generate the write-component grid parameters. * For the RRFS_CONUS_13km grid, reduce the time step (DT_ATMOS) from 180sec to 45sec. This is necessary to get a successful forecast with the GSD_SAR suite, and thus likely also the RRFS_v1beta suite. * Modify WE2E testing system as follows: * Add new tests with the RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grids that use the GFS_v15p2 and RRFS_v1beta suites (which are now the ones officially supported in the first release of the short-range weather app) instead of the GFS_v16beta and GSD_SAR suites, respectively. * For clarity, rename the test configuration files that use the GFS_v16beta and GSD_SAR suites so they include the suite name. * Update list of WE2E tests (baselines_list.txt). * Update the NCL plotting scripts to be able to plot grids with the latest version of the workflow. ## TESTS CONDUCTED: On hera, ran tests with all three grids with the GFS_v15p2 and RRFS_v1beta suites (a total of 6 tests). All were successful. * Remove redundant model_configure.${CCPP_PHYS_SUITE} template files; use Jinja2 to create model_configure (ufs-community#321) ## DESCRIPTION OF CHANGES: * Remove model_configure template files whose names depend on the physics suite, i.e. files with names of the form model_configure.${CCPP_PHYS_SUITE}. Only a single template file is needed because the contents of the model_configure file are not suite dependent. This leaves just one template file (named model_configure). * Change the function create_model_configure_file.sh and the template file model_configure so they use jinja2 instead of sed to replace placeholder values. * Absorb the contents of the write-component template files wrtcmp_lambert_conformal, wrtcmp_regional_latlon, and wrtcmp_rotated_latlon into the new jinja2-compliant model_configure file. We can do this because Jinja2 allows use of if-statements in the template file. * In the new model_configure jinja2 template file, include comments to explain the various write-component parameters. ## TESTS CONDUCTED: On Hera, ran the two WE2E tests new_ESGgrid and new_GFDLgrid. The first uses a "lambert_conformal" type of write-component grid, and the second uses a "rotated_latlon" type of write-component grid. (The write-component also allows "regional_latlon" type grids, which is just the usual earth-relative latlon coordinate system, but we do not have any cases that use that.) Both tests succeeded. ## ISSUE (optional): This PR resolves issue ufs-community#281. * Add Thompson ice- and water-friendly aerosol climo file support (ufs-community#332) * Add if statement in set_thompson_mp_fix_files.sh to source Thompson climo file when using a combination of a Thompson-based SDF and non-RAP/HRRR external model data * Modify if statement based on external models for Thompson climo file * Remove workflow variable EMC_GRID_NAME (ufs-community#333) ## DESCRIPTION OF CHANGES: * Remove the workflow variable EMC_GRID_NAME. Henceforth, PREDEF_GRID_NAME is the only variable that can be used to set the name of the predefined grid to use. * Make appropriate change of variable name (EMC_GRID_NAME --> PREDEF_GRID_NAME) in the WE2E test configuration files. * Change anywhere the "conus" and "conus_c96" grids are specified to "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Rename WE2E test configuration files with names containing the strings "conus" and "conus_c96" by replacing these strings with "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Update the list of WE2E test names (tests/baselines_list.txt). * Bug fixes not directly related to grids: * In config.nco.sh, remove settings of QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST since these are now set automatically (due to another PR). * In the template file FV3LAM_wflow.xml, add the ensemble member name after RUN_FCST_TN in the dependency of the run_post metatask. ## TESTS CONDUCTED: Since this change only affects runs in NCO mode, the following NCO-mode WE2E tests were rerun on hera, all successfully: ``` nco_EMC_CONUS_3km SUCCESS nco_EMC_CONUS_coarse SUCCESS nco_EMC_CONUS_coarse__suite_FV3_GFS_2017_gfdlmp SUCCESS nco_RRFS_CONUS_25km_HRRRX_RAPX SUCCESS nco_RRFS_CONUS_3km_FV3GFS_FV3GFS SUCCESS nco_RRFS_CONUS_3km_HRRRX_RAPX SUCCESS nco_ensemble SUCCESS ``` * Port workflow to Orion (ufs-community#309) ## DESCRIPTION OF CHANGES: * Add stanzas for Orion where necessary. * Add new module files for Orion. * On Orion, both the slurm partition and the slurm QOS need to be specified in the rocoto XML in order to be able to have wall times longer than 30 mins (the partition needs to be specified because it is by default "debug", which has a limit of 30 mins). Thus, introduce modifications to more easily specify slurm partitions: * Remove the workflow variables QUEUE_DEFAULT_TAG, QUEUE_HPSS_TAG, and QUEUE_FCST_TAG that are currently used to determine whether QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST specify the names of queue/QOS's or slurm partitions. * Add the workflow variables PARTITION_DEFAULT_TAG, PARTITION_HPSS_TAG, and PARTITION_FCST_TAG. These will be used to specify slurm partitions only, and the variables QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST will be used to specify queues/QOS's only. IMPORTANT NOTE: On Orion, in order to load the regional_workflow environment needed for generating an experiment, the user must first issue the following commands: ``` module use -a /apps/contrib/miniconda3-noaa-gsl/modulefiles module load miniconda3 conda activate regional_workflow ``` ## TESTS CONDUCTED: Ran 11 WE2E tests on Orion, Hera, and Cheyenne. Results on Orion: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In the run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Hera: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km SUCCESS new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Cheyenne: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` All succeed except GSD_HRRR_AK_50km on Orion and Cheyenne. It is not clear why grid_GSD_HRRR_AK_50km fails on Orion and Cheyenne but not Hera. Seems to point to a bug in the forecast model. These two failures are not so important since this grid will soon be deprecated. Also tested successfully on Jet by @JeffBeck-NOAA and on Odin and Stampede by @ywangwof. ## ISSUE: This resolves Issue ufs-community#152. ## CONTRIBUTORS: @JeffBeck-NOAA @ywangwof @christinaholtNOAA * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: Jamie Wolff <jwolff@ucar.edu> * Change cov_thresh for REFL to be a true max in nbrhood as SPC does. * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Separated Pull Data Scripts from Run Vx Scripts: Feature/add_metplus (ufs-community#2) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Fixed Vx Task Dependencies in Workflow: Feature/add metplus (ufs-community#3) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Manual merge with develop that didn't seem to work before. Trying to get feature branch updated so it will run again! * Add local module files * Add environment variable for SCRIPTSDIR * Remove echo statement * Remove old module files * Update to config_default for walltime for ndas pull. Update to metplus parm for obs file template. Update to FV3LAM xml to not include 00 hour for verification * Update template to remove full path * Verification channges for obs. (ufs-community#4) * Verification channges for obs. * Update config_defaults.sh for vx description * Update config_defaults.sh to remove extraneous MET info. Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> * Initial METplus .confs and MET config files for EnsembleStat APCP * J-Job script for running ensemble stat * Exregional script for ensemble-stat * Added EnsembleStat.conf for A6 and A24. Added PCPCombine to A3, A6, and A24. * Added EnsembleStatConfig files for 6 and 24h * Copy of workflow template with precipitation ensemble tasks added. Will become main template when testing is complete * Added export statement for number of ensemble members * Added necessary task definitions in ush * Updated workflow to included ENTITY definitions for ensstat * Fixed typo * Added ens vx configs * Pull in updates from develop that were not merging properly. Small change to config.community to turn off vx tasks by default. * Added/mod files for point ens vx. * Updated metplus conf files for ens point vx * Did manual merge of these files because it was not handled properly automatically * Adding additional variables to METplus for regional workflow (ufs-community#5) * Changes made based on meeting with Michelle and Jamie * Updating fork * Cleanup after merge * Added additional ens vx * Ensemble point vx mods * Additional updates for ens and det vx * ensgrid_mean and ensgrid_prob .conf files for APCP * Updates for ensemble vx. * Added mean and prob point-stat configs * Updates to ensgrid_vx * Updates for mean/prob vx. * Updates to FV3LAM_wflow.xml * Deterministic and ensembel vx updates. * Ensgrid mean * Update setup.sh * Changed workflow template title * Updates to deterministic and ensemble verification * Created EnsembleStat METplus conf and MET config files for REFC * Added reflectivity mean and prob METplus and MET config files. Updated APCP mean and prob METplus and MET config files. * Added all J-job scripts, exregional scripts, and necessary definitons for workflow generation for all ensgrid_mean and ensgrid_prob tasks * Updates to workflow to add ensgrid_vx * Changes I made to account for runtime errors. * Made changes to directory structures * Made changes to directory structures and variables * Changed log files and stage dir. * Changes for grid- and point-vx. * Updated METplus ensemble precip conf files. * Mods for ensemble and deterministic vx. * Change to GridStatConfig_REFC_mean * Updated EnsembleStat_REFC.conf * Updated to METv10.0.0 * Updated conf files for paths. * Updated FV3LAM_wflow.xml template. * Mods for vx dependencies * Updated for censor thresh in METplus conf files; changes to FV3LAM_wflow.xml after sync with develop. * Updated exregional_run_fcst.sh generate_FV3LAM_wflow.sh to address merge with develop. * Mods for ensemble precip vx, handling padded/non-padded ensemble member names, fixes for python environment for obs pull. * Changes to RETOP (units) and REFC (naming and level) verification. * Fix OUTPUT_BASE for deterministic vx. * Changes to some verification ex-scripts for syntax and path fixes. Included start end dates of incorrect 01-h CCPA data. Removed some extra lines in FV3LAM_wflow.xml template. * Changed comp. ref. variable name in GridStat_REFC_prob.conf * Changed comp. ref. level in GridStat_REFC_prob.conf * Updated logic for number padding in the directory name when running in ensemble mode. * Added MET ensemble vx WE2E test. * Modified location of obs to live outside cycle dir, allowing for obs to be shared across cycles. * Mods to address comments on PR575. * Updated ensemble METPlus conf files for changes to post output name. * Addessed comments in PR and mods for 10-m WIND. * Addressing final comments in PR. Co-authored-by: Jamie Wolff <jwolff@ucar.edu> Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: lindsayrblank <lblank@ucar.edu> Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> Co-authored-by: PerryShafran-NOAA <62255233+PerryShafran-NOAA@users.noreply.github.com>
mkavulich
added a commit
that referenced
this pull request
Sep 8, 2022
* Fix to post flat file. * Create MET and METplus config files under ush/templates/parm * Added script to pull and reorg ccpa data. Added a script to run gridstat with METplus. Updated MET and METplus config files. * Added new jjob for running grid-stat vx. Updated setup.sh to include grid-stat vx. Updated run_gridstatvx script. * Fixed typo on script name from ksh to sh * Moved some hard coded items out from the script to the XML * Updates to get METplus to run with fewer hard-coded paths. * Updates to add grid-stat task to XML generation. * Bug fixes for adding grid-stat to XML generation * Updates to remove hard-coded paths in config files * Change log dir to put master_metplus log file with other logs under log/, rather than default logs/. * Updates to generate xml without hard-coded paths for MET * Add hera gridstat module file * Add METplus point-stat task for both sfc and upper air * Small tweaks to remove hard coded paths and add some flexibility * Updates for adding point-stat into auto-generated xml * Add in function to set point-stat task to FALSE * Final tweaks to get it to generate the xml correctly * Minor updates to run ensure 0,6,12,18 * Tweaks to var list for Point-Stat * Add METplus settings to config_defaults * Move quote for end of settings and fix extra comment. * Fix typos to populate templates correctly * Updated to include SCRIPTSDIR and other MET specific settings along with updates to FHR syntax * Update module loads on hera * Fixed comment for BOTH_VARn_THRESH to avoid syntax issues * Added files to run grid_stat for a variety of accumulation intervals, including 3, 6, and 24h * Added module load hpss * Remove module load informatino from these scripts * Updated the method of turning on/off vx tasks using jinja template if statement * Remove commented out lines of code. Fixed typo. Removed gen_wflow.out file. * Updated pull scripts to have file names dependent on date to pull from HPSS. Updated to export a few more local variables that METplus conf needed in scripts. Updated workflow to use service queue (for now) to for 1h grid_stat and point_stat run and default for 3+h accumulation grid_stat runs) * moved common_hera.conf to common.conf - no platform specific information included that needs to be handled. * Remove common_hera.conf * Add scripts to pull and process MRMS data from NOAA HPSS * Updates for REFC vx tasks * updates to obs pull scripts * Update for adding in reflectivity verification using MRMS analyses and updating name of model output to RRFS rather than HRRR * Updates to account for CCPA issues on HPSS - day off for 00-05 UTC directories * Verification mods to feature/add metplus (#1) * Remove unused/outdated code (#313) ## DESCRIPTION OF CHANGES: * In setup.sh and generate_FV3LAM_wflow.sh, remove temporary codes that fix bugs in the FV3_GFS_2017_gfdlmp_regional suite definition file because those bugs have been fixed (in the ufs-weather-model repo). * In setup.sh, remove block of code that is no longer necessary because chgres_cube can now initialize from external model data with either 4 or 9 soil levels, and run with LSMs of either 4 or 9 soil levels. * Remove modifications to LD_LIBRARY_PATH in exregional_run_fcst.sh. * For the make_ics and make_lbcs tasks, move the setting of APRUN and other machine-specific actions from the J-job to the ex-script in order to be consistent with the other workflow tasks. * Fix indentation and edit comments. * Remove unused file load_fv3gfs_modules.sh. ## TESTS CONDUCTED: Ran two WE2E tests on hera, new_ESGgrid and new_GFDLgrid: * new_ESGgrid uses the FV3_GFS_2017_gfdlmp_regional suite. The test was successful. * new_GFDLgrid uses the FV3_GFS_2017_gfdlmp suite. The test was successful. ## ISSUE (optional): This resolves issue #198. * Add and call a function that checks for use of Thompson microphysics parameterization in the SDF and if so, adjusts certain workflow arrays to contain the names and other associated values of the fixed files needed by this parameterization so that those files are automatically copied and/or linked to. (#319) ## DESCRIPTION OF CHANGES: Add and call a function that checks for use of Thompson microphysics parameterization in the suite definition file (SDF). If not, do nothing. If so, add to the appropriate workflow arrays the names and other associated values of the fixed files needed by this parameterization so that they are automatically copied and/or linked to instead of being regenerated from scratch in the run_fcst task. ## TESTS CONDUCTED: On hera, ran two WE2E tests, one in NCO mode (nco_RRFS_CONUS_25km_HRRRX_RAPX) and the other in community mode (suite_FV3_GSD_v0). These use suites FV3_GSD_SAR and FV3_GSD_v0, respectively, and both of these call Thompson microphysics. Both succeeded. ## ISSUE (optional): This PR resolves issue #297. * RRFS_v1beta SDF changes after reverting from GSL to GFS GWD suite (#322) (#327) ## DESCRIPTION OF CHANGES: Removed checks on the RRFS_v1beta SDF implemented for use with the GSL GWD suite (now uses the GFS GWD suite). No longer copies staged orography files necessary for the GSL GWD suite. ## TESTS CONDUCTED: Runs to completion on Hera. End-to-end runs DOT_OR_USCORE and suite_FV3_RRFS_v1beta succeeded on Cheyenne. Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Adjust RRFS_CONUS_... grids (#294) ## DESCRIPTION OF CHANGES: * Adjust RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grid parameters so that: * All grids, including their 4-cell-wide halos, lie completely within the HRRRX domain. * All grids have dimensions nx and ny that factor "nicely", i.e. they don't have factors greather than 7. * The write-component grids corresponding to these three native grids cover as much of the native grids as possible without going outside of the native grid boundaries. The updated NCL scripts (see below) were used to generate the write-component grid parameters. * For the RRFS_CONUS_13km grid, reduce the time step (DT_ATMOS) from 180sec to 45sec. This is necessary to get a successful forecast with the GSD_SAR suite, and thus likely also the RRFS_v1beta suite. * Modify WE2E testing system as follows: * Add new tests with the RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grids that use the GFS_v15p2 and RRFS_v1beta suites (which are now the ones officially supported in the first release of the short-range weather app) instead of the GFS_v16beta and GSD_SAR suites, respectively. * For clarity, rename the test configuration files that use the GFS_v16beta and GSD_SAR suites so they include the suite name. * Update list of WE2E tests (baselines_list.txt). * Update the NCL plotting scripts to be able to plot grids with the latest version of the workflow. ## TESTS CONDUCTED: On hera, ran tests with all three grids with the GFS_v15p2 and RRFS_v1beta suites (a total of 6 tests). All were successful. * Remove redundant model_configure.${CCPP_PHYS_SUITE} template files; use Jinja2 to create model_configure (#321) ## DESCRIPTION OF CHANGES: * Remove model_configure template files whose names depend on the physics suite, i.e. files with names of the form model_configure.${CCPP_PHYS_SUITE}. Only a single template file is needed because the contents of the model_configure file are not suite dependent. This leaves just one template file (named model_configure). * Change the function create_model_configure_file.sh and the template file model_configure so they use jinja2 instead of sed to replace placeholder values. * Absorb the contents of the write-component template files wrtcmp_lambert_conformal, wrtcmp_regional_latlon, and wrtcmp_rotated_latlon into the new jinja2-compliant model_configure file. We can do this because Jinja2 allows use of if-statements in the template file. * In the new model_configure jinja2 template file, include comments to explain the various write-component parameters. ## TESTS CONDUCTED: On Hera, ran the two WE2E tests new_ESGgrid and new_GFDLgrid. The first uses a "lambert_conformal" type of write-component grid, and the second uses a "rotated_latlon" type of write-component grid. (The write-component also allows "regional_latlon" type grids, which is just the usual earth-relative latlon coordinate system, but we do not have any cases that use that.) Both tests succeeded. ## ISSUE (optional): This PR resolves issue #281. * Add Thompson ice- and water-friendly aerosol climo file support (#332) * Add if statement in set_thompson_mp_fix_files.sh to source Thompson climo file when using a combination of a Thompson-based SDF and non-RAP/HRRR external model data * Modify if statement based on external models for Thompson climo file * Remove workflow variable EMC_GRID_NAME (#333) ## DESCRIPTION OF CHANGES: * Remove the workflow variable EMC_GRID_NAME. Henceforth, PREDEF_GRID_NAME is the only variable that can be used to set the name of the predefined grid to use. * Make appropriate change of variable name (EMC_GRID_NAME --> PREDEF_GRID_NAME) in the WE2E test configuration files. * Change anywhere the "conus" and "conus_c96" grids are specified to "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Rename WE2E test configuration files with names containing the strings "conus" and "conus_c96" by replacing these strings with "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Update the list of WE2E test names (tests/baselines_list.txt). * Bug fixes not directly related to grids: * In config.nco.sh, remove settings of QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST since these are now set automatically (due to another PR). * In the template file FV3LAM_wflow.xml, add the ensemble member name after RUN_FCST_TN in the dependency of the run_post metatask. ## TESTS CONDUCTED: Since this change only affects runs in NCO mode, the following NCO-mode WE2E tests were rerun on hera, all successfully: ``` nco_EMC_CONUS_3km SUCCESS nco_EMC_CONUS_coarse SUCCESS nco_EMC_CONUS_coarse__suite_FV3_GFS_2017_gfdlmp SUCCESS nco_RRFS_CONUS_25km_HRRRX_RAPX SUCCESS nco_RRFS_CONUS_3km_FV3GFS_FV3GFS SUCCESS nco_RRFS_CONUS_3km_HRRRX_RAPX SUCCESS nco_ensemble SUCCESS ``` * Port workflow to Orion (#309) ## DESCRIPTION OF CHANGES: * Add stanzas for Orion where necessary. * Add new module files for Orion. * On Orion, both the slurm partition and the slurm QOS need to be specified in the rocoto XML in order to be able to have wall times longer than 30 mins (the partition needs to be specified because it is by default "debug", which has a limit of 30 mins). Thus, introduce modifications to more easily specify slurm partitions: * Remove the workflow variables QUEUE_DEFAULT_TAG, QUEUE_HPSS_TAG, and QUEUE_FCST_TAG that are currently used to determine whether QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST specify the names of queue/QOS's or slurm partitions. * Add the workflow variables PARTITION_DEFAULT_TAG, PARTITION_HPSS_TAG, and PARTITION_FCST_TAG. These will be used to specify slurm partitions only, and the variables QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST will be used to specify queues/QOS's only. IMPORTANT NOTE: On Orion, in order to load the regional_workflow environment needed for generating an experiment, the user must first issue the following commands: ``` module use -a /apps/contrib/miniconda3-noaa-gsl/modulefiles module load miniconda3 conda activate regional_workflow ``` ## TESTS CONDUCTED: Ran 11 WE2E tests on Orion, Hera, and Cheyenne. Results on Orion: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In the run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Hera: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km SUCCESS new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Cheyenne: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` All succeed except GSD_HRRR_AK_50km on Orion and Cheyenne. It is not clear why grid_GSD_HRRR_AK_50km fails on Orion and Cheyenne but not Hera. Seems to point to a bug in the forecast model. These two failures are not so important since this grid will soon be deprecated. Also tested successfully on Jet by @JeffBeck-NOAA and on Odin and Stampede by @ywangwof. ## ISSUE: This resolves Issue #152. ## CONTRIBUTORS: @JeffBeck-NOAA @ywangwof @christinaholtNOAA * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: Jamie Wolff <jwolff@ucar.edu> * Change cov_thresh for REFL to be a true max in nbrhood as SPC does. * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Separated Pull Data Scripts from Run Vx Scripts: Feature/add_metplus (#2) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Fixed Vx Task Dependencies in Workflow: Feature/add metplus (#3) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Manual merge with develop that didn't seem to work before. Trying to get feature branch updated so it will run again! * Add local module files * Add environment variable for SCRIPTSDIR * Remove echo statement * Remove old module files * Update to config_default for walltime for ndas pull. Update to metplus parm for obs file template. Update to FV3LAM xml to not include 00 hour for verification * Update template to remove full path * Verification channges for obs. (#4) * Verification channges for obs. * Update config_defaults.sh for vx description * Update config_defaults.sh to remove extraneous MET info. Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> * Initial METplus .confs and MET config files for EnsembleStat APCP * J-Job script for running ensemble stat * Exregional script for ensemble-stat * Added EnsembleStat.conf for A6 and A24. Added PCPCombine to A3, A6, and A24. * Added EnsembleStatConfig files for 6 and 24h * Copy of workflow template with precipitation ensemble tasks added. Will become main template when testing is complete * Added export statement for number of ensemble members * Added necessary task definitions in ush * Updated workflow to included ENTITY definitions for ensstat * Fixed typo * Added ens vx configs * Pull in updates from develop that were not merging properly. Small change to config.community to turn off vx tasks by default. * Added/mod files for point ens vx. * Updated metplus conf files for ens point vx * Did manual merge of these files because it was not handled properly automatically * Adding additional variables to METplus for regional workflow (#5) * Changes made based on meeting with Michelle and Jamie * Updating fork * Cleanup after merge * Added additional ens vx * Ensemble point vx mods * Additional updates for ens and det vx * ensgrid_mean and ensgrid_prob .conf files for APCP * Updates for ensemble vx. * Added mean and prob point-stat configs * Updates to ensgrid_vx * Updates for mean/prob vx. * Updates to FV3LAM_wflow.xml * Deterministic and ensembel vx updates. * Ensgrid mean * Update setup.sh * Changed workflow template title * Updates to deterministic and ensemble verification * Created EnsembleStat METplus conf and MET config files for REFC * Added reflectivity mean and prob METplus and MET config files. Updated APCP mean and prob METplus and MET config files. * Added all J-job scripts, exregional scripts, and necessary definitons for workflow generation for all ensgrid_mean and ensgrid_prob tasks * Updates to workflow to add ensgrid_vx * Changes I made to account for runtime errors. * Made changes to directory structures * Made changes to directory structures and variables * Changed log files and stage dir. * Changes for grid- and point-vx. * Updated METplus ensemble precip conf files. * Mods for ensemble and deterministic vx. * Change to GridStatConfig_REFC_mean * Updated EnsembleStat_REFC.conf * Updated to METv10.0.0 * Updated conf files for paths. * Updated FV3LAM_wflow.xml template. * Mods for vx dependencies * Updated for censor thresh in METplus conf files; changes to FV3LAM_wflow.xml after sync with develop. * Updated exregional_run_fcst.sh generate_FV3LAM_wflow.sh to address merge with develop. * Mods for ensemble precip vx, handling padded/non-padded ensemble member names, fixes for python environment for obs pull. * Changes to RETOP (units) and REFC (naming and level) verification. * Fix OUTPUT_BASE for deterministic vx. * Changes to some verification ex-scripts for syntax and path fixes. Included start end dates of incorrect 01-h CCPA data. Removed some extra lines in FV3LAM_wflow.xml template. * Changed comp. ref. variable name in GridStat_REFC_prob.conf * Changed comp. ref. level in GridStat_REFC_prob.conf * Updated logic for number padding in the directory name when running in ensemble mode. * Added MET ensemble vx WE2E test. * Modified location of obs to live outside cycle dir, allowing for obs to be shared across cycles. * Mods to address comments on PR575. * Updated ensemble METPlus conf files for changes to post output name. * Addessed comments in PR and mods for 10-m WIND. * Addressing final comments in PR. Co-authored-by: Jamie Wolff <jwolff@ucar.edu> Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: lindsayrblank <lblank@ucar.edu> Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> Co-authored-by: PerryShafran-NOAA <62255233+PerryShafran-NOAA@users.noreply.github.com>
danielabdi-noaa
pushed a commit
to danielabdi-noaa/ufs-srweather-app
that referenced
this pull request
Sep 8, 2022
* Fix to post flat file. * Create MET and METplus config files under ush/templates/parm * Added script to pull and reorg ccpa data. Added a script to run gridstat with METplus. Updated MET and METplus config files. * Added new jjob for running grid-stat vx. Updated setup.sh to include grid-stat vx. Updated run_gridstatvx script. * Fixed typo on script name from ksh to sh * Moved some hard coded items out from the script to the XML * Updates to get METplus to run with fewer hard-coded paths. * Updates to add grid-stat task to XML generation. * Bug fixes for adding grid-stat to XML generation * Updates to remove hard-coded paths in config files * Change log dir to put master_metplus log file with other logs under log/, rather than default logs/. * Updates to generate xml without hard-coded paths for MET * Add hera gridstat module file * Add METplus point-stat task for both sfc and upper air * Small tweaks to remove hard coded paths and add some flexibility * Updates for adding point-stat into auto-generated xml * Add in function to set point-stat task to FALSE * Final tweaks to get it to generate the xml correctly * Minor updates to run ensure 0,6,12,18 * Tweaks to var list for Point-Stat * Add METplus settings to config_defaults * Move quote for end of settings and fix extra comment. * Fix typos to populate templates correctly * Updated to include SCRIPTSDIR and other MET specific settings along with updates to FHR syntax * Update module loads on hera * Fixed comment for BOTH_VARn_THRESH to avoid syntax issues * Added files to run grid_stat for a variety of accumulation intervals, including 3, 6, and 24h * Added module load hpss * Remove module load informatino from these scripts * Updated the method of turning on/off vx tasks using jinja template if statement * Remove commented out lines of code. Fixed typo. Removed gen_wflow.out file. * Updated pull scripts to have file names dependent on date to pull from HPSS. Updated to export a few more local variables that METplus conf needed in scripts. Updated workflow to use service queue (for now) to for 1h grid_stat and point_stat run and default for 3+h accumulation grid_stat runs) * moved common_hera.conf to common.conf - no platform specific information included that needs to be handled. * Remove common_hera.conf * Add scripts to pull and process MRMS data from NOAA HPSS * Updates for REFC vx tasks * updates to obs pull scripts * Update for adding in reflectivity verification using MRMS analyses and updating name of model output to RRFS rather than HRRR * Updates to account for CCPA issues on HPSS - day off for 00-05 UTC directories * Verification mods to feature/add metplus (#1) * Remove unused/outdated code (ufs-community#313) ## DESCRIPTION OF CHANGES: * In setup.sh and generate_FV3LAM_wflow.sh, remove temporary codes that fix bugs in the FV3_GFS_2017_gfdlmp_regional suite definition file because those bugs have been fixed (in the ufs-weather-model repo). * In setup.sh, remove block of code that is no longer necessary because chgres_cube can now initialize from external model data with either 4 or 9 soil levels, and run with LSMs of either 4 or 9 soil levels. * Remove modifications to LD_LIBRARY_PATH in exregional_run_fcst.sh. * For the make_ics and make_lbcs tasks, move the setting of APRUN and other machine-specific actions from the J-job to the ex-script in order to be consistent with the other workflow tasks. * Fix indentation and edit comments. * Remove unused file load_fv3gfs_modules.sh. ## TESTS CONDUCTED: Ran two WE2E tests on hera, new_ESGgrid and new_GFDLgrid: * new_ESGgrid uses the FV3_GFS_2017_gfdlmp_regional suite. The test was successful. * new_GFDLgrid uses the FV3_GFS_2017_gfdlmp suite. The test was successful. ## ISSUE (optional): This resolves issue NOAA-GSL#198. * Add and call a function that checks for use of Thompson microphysics parameterization in the SDF and if so, adjusts certain workflow arrays to contain the names and other associated values of the fixed files needed by this parameterization so that those files are automatically copied and/or linked to. (ufs-community#319) ## DESCRIPTION OF CHANGES: Add and call a function that checks for use of Thompson microphysics parameterization in the suite definition file (SDF). If not, do nothing. If so, add to the appropriate workflow arrays the names and other associated values of the fixed files needed by this parameterization so that they are automatically copied and/or linked to instead of being regenerated from scratch in the run_fcst task. ## TESTS CONDUCTED: On hera, ran two WE2E tests, one in NCO mode (nco_RRFS_CONUS_25km_HRRRX_RAPX) and the other in community mode (suite_FV3_GSD_v0). These use suites FV3_GSD_SAR and FV3_GSD_v0, respectively, and both of these call Thompson microphysics. Both succeeded. ## ISSUE (optional): This PR resolves issue ufs-community#297. * RRFS_v1beta SDF changes after reverting from GSL to GFS GWD suite (ufs-community#322) (ufs-community#327) ## DESCRIPTION OF CHANGES: Removed checks on the RRFS_v1beta SDF implemented for use with the GSL GWD suite (now uses the GFS GWD suite). No longer copies staged orography files necessary for the GSL GWD suite. ## TESTS CONDUCTED: Runs to completion on Hera. End-to-end runs DOT_OR_USCORE and suite_FV3_RRFS_v1beta succeeded on Cheyenne. Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Adjust RRFS_CONUS_... grids (ufs-community#294) ## DESCRIPTION OF CHANGES: * Adjust RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grid parameters so that: * All grids, including their 4-cell-wide halos, lie completely within the HRRRX domain. * All grids have dimensions nx and ny that factor "nicely", i.e. they don't have factors greather than 7. * The write-component grids corresponding to these three native grids cover as much of the native grids as possible without going outside of the native grid boundaries. The updated NCL scripts (see below) were used to generate the write-component grid parameters. * For the RRFS_CONUS_13km grid, reduce the time step (DT_ATMOS) from 180sec to 45sec. This is necessary to get a successful forecast with the GSD_SAR suite, and thus likely also the RRFS_v1beta suite. * Modify WE2E testing system as follows: * Add new tests with the RRFS_CONUS_25km, RRFS_CONUS_13km, and RRFS_CONUS_3km grids that use the GFS_v15p2 and RRFS_v1beta suites (which are now the ones officially supported in the first release of the short-range weather app) instead of the GFS_v16beta and GSD_SAR suites, respectively. * For clarity, rename the test configuration files that use the GFS_v16beta and GSD_SAR suites so they include the suite name. * Update list of WE2E tests (baselines_list.txt). * Update the NCL plotting scripts to be able to plot grids with the latest version of the workflow. ## TESTS CONDUCTED: On hera, ran tests with all three grids with the GFS_v15p2 and RRFS_v1beta suites (a total of 6 tests). All were successful. * Remove redundant model_configure.${CCPP_PHYS_SUITE} template files; use Jinja2 to create model_configure (ufs-community#321) ## DESCRIPTION OF CHANGES: * Remove model_configure template files whose names depend on the physics suite, i.e. files with names of the form model_configure.${CCPP_PHYS_SUITE}. Only a single template file is needed because the contents of the model_configure file are not suite dependent. This leaves just one template file (named model_configure). * Change the function create_model_configure_file.sh and the template file model_configure so they use jinja2 instead of sed to replace placeholder values. * Absorb the contents of the write-component template files wrtcmp_lambert_conformal, wrtcmp_regional_latlon, and wrtcmp_rotated_latlon into the new jinja2-compliant model_configure file. We can do this because Jinja2 allows use of if-statements in the template file. * In the new model_configure jinja2 template file, include comments to explain the various write-component parameters. ## TESTS CONDUCTED: On Hera, ran the two WE2E tests new_ESGgrid and new_GFDLgrid. The first uses a "lambert_conformal" type of write-component grid, and the second uses a "rotated_latlon" type of write-component grid. (The write-component also allows "regional_latlon" type grids, which is just the usual earth-relative latlon coordinate system, but we do not have any cases that use that.) Both tests succeeded. ## ISSUE (optional): This PR resolves issue ufs-community#281. * Add Thompson ice- and water-friendly aerosol climo file support (ufs-community#332) * Add if statement in set_thompson_mp_fix_files.sh to source Thompson climo file when using a combination of a Thompson-based SDF and non-RAP/HRRR external model data * Modify if statement based on external models for Thompson climo file * Remove workflow variable EMC_GRID_NAME (ufs-community#333) ## DESCRIPTION OF CHANGES: * Remove the workflow variable EMC_GRID_NAME. Henceforth, PREDEF_GRID_NAME is the only variable that can be used to set the name of the predefined grid to use. * Make appropriate change of variable name (EMC_GRID_NAME --> PREDEF_GRID_NAME) in the WE2E test configuration files. * Change anywhere the "conus" and "conus_c96" grids are specified to "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Rename WE2E test configuration files with names containing the strings "conus" and "conus_c96" by replacing these strings with "EMC_CONUS_3km" and "EMC_CONUS_coarse", respectively. * Update the list of WE2E test names (tests/baselines_list.txt). * Bug fixes not directly related to grids: * In config.nco.sh, remove settings of QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST since these are now set automatically (due to another PR). * In the template file FV3LAM_wflow.xml, add the ensemble member name after RUN_FCST_TN in the dependency of the run_post metatask. ## TESTS CONDUCTED: Since this change only affects runs in NCO mode, the following NCO-mode WE2E tests were rerun on hera, all successfully: ``` nco_EMC_CONUS_3km SUCCESS nco_EMC_CONUS_coarse SUCCESS nco_EMC_CONUS_coarse__suite_FV3_GFS_2017_gfdlmp SUCCESS nco_RRFS_CONUS_25km_HRRRX_RAPX SUCCESS nco_RRFS_CONUS_3km_FV3GFS_FV3GFS SUCCESS nco_RRFS_CONUS_3km_HRRRX_RAPX SUCCESS nco_ensemble SUCCESS ``` * Port workflow to Orion (ufs-community#309) ## DESCRIPTION OF CHANGES: * Add stanzas for Orion where necessary. * Add new module files for Orion. * On Orion, both the slurm partition and the slurm QOS need to be specified in the rocoto XML in order to be able to have wall times longer than 30 mins (the partition needs to be specified because it is by default "debug", which has a limit of 30 mins). Thus, introduce modifications to more easily specify slurm partitions: * Remove the workflow variables QUEUE_DEFAULT_TAG, QUEUE_HPSS_TAG, and QUEUE_FCST_TAG that are currently used to determine whether QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST specify the names of queue/QOS's or slurm partitions. * Add the workflow variables PARTITION_DEFAULT_TAG, PARTITION_HPSS_TAG, and PARTITION_FCST_TAG. These will be used to specify slurm partitions only, and the variables QUEUE_DEFAULT, QUEUE_HPSS, and QUEUE_FCST will be used to specify queues/QOS's only. IMPORTANT NOTE: On Orion, in order to load the regional_workflow environment needed for generating an experiment, the user must first issue the following commands: ``` module use -a /apps/contrib/miniconda3-noaa-gsl/modulefiles module load miniconda3 conda activate regional_workflow ``` ## TESTS CONDUCTED: Ran 11 WE2E tests on Orion, Hera, and Cheyenne. Results on Orion: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In the run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Hera: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km SUCCESS new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` Results on Cheyenne: ``` community_ensemble_2mems SUCCESS DOT_OR_USCORE SUCCESS grid_GSD_HRRR_AK_50km FAILURE - In run_fcst task. * Error message: !!! (1) Error in subr radiation_aerosols: unrealistic surface pressure = 1 NaN new_ESGgrid SUCCESS new_GFDLgrid SUCCESS regional_001 SUCCESS regional_002 SUCCESS suite_FV3_GFS_v15p2 SUCCESS suite_FV3_GFS_v16beta SUCCESS suite_FV3_GSD_SAR SUCCESS suite_FV3_GSD_v0 SUCCESS ``` All succeed except GSD_HRRR_AK_50km on Orion and Cheyenne. It is not clear why grid_GSD_HRRR_AK_50km fails on Orion and Cheyenne but not Hera. Seems to point to a bug in the forecast model. These two failures are not so important since this grid will soon be deprecated. Also tested successfully on Jet by @JeffBeck-NOAA and on Odin and Stampede by @ywangwof. ## ISSUE: This resolves Issue NOAA-GSL#152. ## CONTRIBUTORS: @JeffBeck-NOAA @ywangwof @christinaholtNOAA * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml * Update FV3.input.nml for fhzero = 1.0 * Updated conf files for file name conventions. * Updated MET scripts and MRMS pull scripts. * Removed comments from exregional_get_mrms_files.sh and removed fhzero from FV3.input.yml Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: Jamie Wolff <jwolff@ucar.edu> * Change cov_thresh for REFL to be a true max in nbrhood as SPC does. * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Separated Pull Data Scripts from Run Vx Scripts: Feature/add_metplus (#2) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Fixed Vx Task Dependencies in Workflow: Feature/add metplus (NOAA-GSL#3) * Job script for get_obs_ccpa * Jobs script for get_obs_mrms * Jobs script for get_obs_ndas * Added external variables necessary to get_ccpa script * Updated workflow template with separate get obs tasks * Separated pull scripts from run scripts * Added necessary defaults/values for defining pull tasks * Added module files, default config.sh options, and changed dependencies for vx tasks * Changed name of new workflow to FV3LAM_wflow.xml * Added task get_obs_tn, removed config.sh, updated config_defaults and config.community.sh * Adjusted the community and default config files based on comments * Updated FV3LAM workflow * Fixed discrepancies in config.community.sh * Fixed discrepancies in config_defaults.sh * Fixed discrepancies in config_defaults.sh round 2 * Fixed discrepancies in config_defaults.sh round 3 * Fixed discrepancies in config_defaults.sh round 4 * Fixed discrepancies in config.community.sh round 2 * Fixed discrepancies in config.community.sh round 3 * Fixed discrepancies in generate_FV3LAM_wflow.sh * Fixed discrepancies in generate_FV3LAM_wflow.sh round 2 * Fixed discrepancies in generate_FV3LAM_wflow.sh round 3 * Updated FV3LAM_wflow template * Fixed the dependencies of the vx tasks * Manual merge with develop that didn't seem to work before. Trying to get feature branch updated so it will run again! * Add local module files * Add environment variable for SCRIPTSDIR * Remove echo statement * Remove old module files * Update to config_default for walltime for ndas pull. Update to metplus parm for obs file template. Update to FV3LAM xml to not include 00 hour for verification * Update template to remove full path * Verification channges for obs. (NOAA-GSL#4) * Verification channges for obs. * Update config_defaults.sh for vx description * Update config_defaults.sh to remove extraneous MET info. Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> * Initial METplus .confs and MET config files for EnsembleStat APCP * J-Job script for running ensemble stat * Exregional script for ensemble-stat * Added EnsembleStat.conf for A6 and A24. Added PCPCombine to A3, A6, and A24. * Added EnsembleStatConfig files for 6 and 24h * Copy of workflow template with precipitation ensemble tasks added. Will become main template when testing is complete * Added export statement for number of ensemble members * Added necessary task definitions in ush * Updated workflow to included ENTITY definitions for ensstat * Fixed typo * Added ens vx configs * Pull in updates from develop that were not merging properly. Small change to config.community to turn off vx tasks by default. * Added/mod files for point ens vx. * Updated metplus conf files for ens point vx * Did manual merge of these files because it was not handled properly automatically * Adding additional variables to METplus for regional workflow (NOAA-GSL#5) * Changes made based on meeting with Michelle and Jamie * Updating fork * Cleanup after merge * Added additional ens vx * Ensemble point vx mods * Additional updates for ens and det vx * ensgrid_mean and ensgrid_prob .conf files for APCP * Updates for ensemble vx. * Added mean and prob point-stat configs * Updates to ensgrid_vx * Updates for mean/prob vx. * Updates to FV3LAM_wflow.xml * Deterministic and ensembel vx updates. * Ensgrid mean * Update setup.sh * Changed workflow template title * Updates to deterministic and ensemble verification * Created EnsembleStat METplus conf and MET config files for REFC * Added reflectivity mean and prob METplus and MET config files. Updated APCP mean and prob METplus and MET config files. * Added all J-job scripts, exregional scripts, and necessary definitons for workflow generation for all ensgrid_mean and ensgrid_prob tasks * Updates to workflow to add ensgrid_vx * Changes I made to account for runtime errors. * Made changes to directory structures * Made changes to directory structures and variables * Changed log files and stage dir. * Changes for grid- and point-vx. * Updated METplus ensemble precip conf files. * Mods for ensemble and deterministic vx. * Change to GridStatConfig_REFC_mean * Updated EnsembleStat_REFC.conf * Updated to METv10.0.0 * Updated conf files for paths. * Updated FV3LAM_wflow.xml template. * Mods for vx dependencies * Updated for censor thresh in METplus conf files; changes to FV3LAM_wflow.xml after sync with develop. * Updated exregional_run_fcst.sh generate_FV3LAM_wflow.sh to address merge with develop. * Mods for ensemble precip vx, handling padded/non-padded ensemble member names, fixes for python environment for obs pull. * Changes to RETOP (units) and REFC (naming and level) verification. * Fix OUTPUT_BASE for deterministic vx. * Changes to some verification ex-scripts for syntax and path fixes. Included start end dates of incorrect 01-h CCPA data. Removed some extra lines in FV3LAM_wflow.xml template. * Changed comp. ref. variable name in GridStat_REFC_prob.conf * Changed comp. ref. level in GridStat_REFC_prob.conf * Updated logic for number padding in the directory name when running in ensemble mode. * Added MET ensemble vx WE2E test. * Modified location of obs to live outside cycle dir, allowing for obs to be shared across cycles. * Mods to address comments on PR575. * Updated ensemble METPlus conf files for changes to post output name. * Addessed comments in PR and mods for 10-m WIND. * Addressing final comments in PR. Co-authored-by: Jamie Wolff <jwolff@ucar.edu> Co-authored-by: gsketefian <31046882+gsketefian@users.noreply.github.com> Co-authored-by: Michael Kavulich <kavulich@ucar.edu> Co-authored-by: JeffBeck-NOAA <55201531+JeffBeck-NOAA@users.noreply.github.com> Co-authored-by: lindsayrblank <lblank@ucar.edu> Co-authored-by: Michelle Harrold <Michelle.Harrold@noaa.gov> Co-authored-by: PerryShafran-NOAA <62255233+PerryShafran-NOAA@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
DESCRIPTION OF CHANGES:
Update Spack Stack location in the NOAA Cloud build file.
TESTS CONDUCTED:
Ran the comprehensive suite of tests on AWS through Jenkins successfully. More work is needed to setup Azure and GCP with Jenkins. As a result, I ran a single E2E test on Azure and GCP manually. Both of these tests were successful.
DEPENDENCIES:
None.
DOCUMENTATION:
No updates.
ISSUE (optional):
CONTRIBUTORS (optional):
@kgerheiser who updated the Spack HPC-Stack on NOAA Cloud.