Add S2SW forecast capability to coupled_sprint branch #416

WalterKolczynski-NOAA · 2021-08-17T18:57:10Z

Add the ability to run the S2SW (non-aerosol coupled) free forecast in the coupled_sprint branch.

Part of Epic #366

The specification of CCPP_SUITE in config.base was erroneously using the CCPP_SUITES from the build system instead of specifying a single CCPP_SUITE for the execution. This wasn't preventing the atm-only from running, since the different variable name meant the setting was just ignored. Refs: NOAA-EMC#416

The confignamevarfornems was not being set correctly, so the default of 'atm' was always used. In coupled-crow, this variable was set to nems_temp in the j-job because there is a separate j-job for cold- starting the mediator. Here I just cut out the middle-man and changed the nems_temp assignments to confignamevarfornems in config.base. Refs: NOAA-EMC#416

Resource settings for the forecast job were previously scattered between multiple files. These settings are now consolidated in config.resources. The total number of PEs is now calculated properly if different coupled models are included. The config.fcst script is updated to properly source component scripts if that component is on. New scripts for those components were created where needed, and settings were moved into the appropriate scripts (though it may be appropriate to relocate additional settings in the future). The special load of the wave config in the j-job is removed as it is no longer needed. CDUMP assignment had to be moved up in the j-job to occur before the config files are sourced so the correct settings are used. Undefined DO switches for stochastic components are added to prevent syntax errors downstream. Settings for do_ugwp, do_tofd, and launch_level are removed from the forecast config, as they are now set in the forecast script based on the CCPP suite. Refs: NOAA-EMC#416

Removing the bit that turns off the wave model prior to a certain date, as the coupled model prototypes run far earlier than that date. The code remains in commented form, as we may need to revisit the cutoff and how to handle it going forward. Refs: NOAA-EMC#416

Updates the forecast launcher string on Orion so it uses the correct number of tasks. Refs: NOAA-EMC#416

Two settings specifying the type of external input to WW3 is updated to use CPL when coupling with the associated model. Refs: NOAA-EMC#416

Replaced the bash-specific indirect variable reference (${!<var>}) with one that uses eval (unfortunately). This was necessary because the script is currently a korne-shell script at runtime, which does not recognize the ! notation. Hopefully, this can be restored in the future. Refs: NOAA-EMC#416

The ocn/ice mesh filename was not correctly being set. Refs: NOAA-EMC#416

Changed the slow coupling interval (ocean coupling) from 60 min to 30 min to match value used in old coupled branch. Refs: NOAA-EMC#416

Turns on coupling from waves to atmosphere when running the full coupled (S2SW) app. This includes moving the setting to config.base so it is set alongside other app-specific settings. Refs: NOAA-EMC#416

Changes the default setting for lheatstrg and lseaspray to match those used in the latest prototype. Refs: NOAA-EMC#416

The wrong path to the mosaic fix file was being used for coupled forecasts and the mesh file was not being determined correctly based on the ice resolution. Both are now corrected. Since it is determined by the ice resolution, the ocn/ice meshfile name is now determined in the CICE post-det. Refs: NOAA-EMC#416

Plan changed and _suffix was never used in the resources config, so it is now removed. Refs: NOAA-EMC#416

Refs: NOAA-EMC#416

A version of the atm post script is copied from coupled-crow branch. It contains small differences from the normal script, and in the future we might be able to consolidate back to one script. Refs: NOAA-EMC#416

There was a line that was overriding the previous selection for the number of threads when running at C384. Also changed the settings to use the correct number of threads when running the free forecast vs gdas. Refs: NOAA-EMC#416

The existing of some fix files was being checked with fallback values used if the files are missing. However, this approach does not work with the tiled fix files, as they are passed as 'tileX', while the files are tile1, tile2, etc. This was causing the files checked to always fall back to their non-tiled versions. In the future, we should check for the existence of all fix files, not just the five that were checked here, appropriately handling the tiled versions. Refs: NOAA-EMC#416

The lheatstrg setting is added to the rest of the variables set depending on the lsm type. Now overridden settings in config.fcst are removed. Refs: NOAA-EMC#416

The numebr of PETs for the ocean model is changed to match the value used by the prototypes. Refs: NOAA-EMC#416

Adds the runoff variable for the ocean model to data_table. (The table is otherwise empty.) Refs: NOAA-EMC#416

Exports are added to the DO switches for the stochastic physics components so they actually make it into the forecast scripts. Refs: NOAA-EMC#416

Changed the default number of PEs in nems.configure to be zero for optional components. The old non-zero default causes problems when waves are run without a ocean/ice model because the PE counts for those components is still used for the PET bounds. Refs: NOAA-EMC#416

The nems.configure template for atm-wave was still using the older coupling_interval_sec, which was not being subsituted at runtime. The parameter is updated to coupling_interval_slow_sec to match what is used for the full coupled model, and the substitution is performed whenever any coupling is on, instead of just cplflux. Refs: NOAA-EMC#416

do_tofd was being added to the namelist twice when coupling was off. Refs: NOAA-EMC#416

OCNRES is still needed in some instances even if the ocean model is off, so the specification is moved back to config.base for now. Refs: NOAA-EMC#416

The following tar files for ocean and ice output are added to the archive job to be stored on HPSS: - ocn_2D - ocn_3D - ocn_xsect - ocn_daily - ocn_ice_grib2_0p25 - ocn_ice_grib2_0p5 - wavocn - ice - gfs_flux_1p00 Handling of the archival of gaussian grid NetCDF files is changed. The former, unused, and now misnamed setting of SAVEFCSTNEMSIO has been changed to ARCH_GAUSSIAN and is now also used consistently between arch.sh and hpssarch_gen.sh. gfs_pgrb2b remains controlled by this setting as well, though it is not a gaussian grib. Additionally, there are new settings, ARCH_GAUSSIAN_FHMAX and ARCH_GAUSSIAN_FHINC, to control which atm and sfc files are saved. The default values are 36 and 6, maintaining the previous hardcoded values The archive job is also streamlined by replacing a bunch of repeating code. Instead of creating tar files in small loops, a list of all the tarfiles to archive is created, then a single loop executes the archival and error checking. Only the MOS tarfile remains separate (for now). The forecast input.nml has been added to the gfsa tarball. The WW3 grid definitions (ww3_multi*) have been added to the gfswave tarball. Refs: NOAA-EMC#416

The variable names used in config.defaults.s2sw for changing the archive settings were from an earlier iteration of the changes. Refs: NOAA-EMC#416

…effective radii computation from Thompson MP" and "Fix/improve logic for convective transportable tracers, add GFS_checktracer debugging routine, correct spelling "janic" --> "janjic", fix inconsistent types related with noahmp and log functions" (NOAA-EMC#416) * Add missing active attribute for one variable in ccpp/data/GFS_typedefs.meta * Correct spelling of janic in CCPP standard names, must be janjic * Remove invalid optional and intent attributes from DDT metadata in ccpp/data/GFS_typedefs.meta * Remove support for optional arguments from CCPP framework/metadata

WalterKolczynski-NOAA self-assigned this Aug 17, 2021

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 19, 2021

Fix forecast launcher for Orion

6000368

Updates the forecast launcher string on Orion so it uses the correct number of tasks. Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 23, 2021

Properly define ocn/ice mesh filename

87d37da

The ocn/ice mesh filename was not correctly being set. Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 24, 2021

Turn on lheatstrg and lseaspray

5da523e

Changes the default setting for lheatstrg and lseaspray to match those used in the latest prototype. Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 24, 2021

Remove unused variable in resources

eb7c657

Plan changed and _suffix was never used in the resources config, so it is now removed. Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 24, 2021

Add trailing new line to ocean config

d7c8362

Refs: NOAA-EMC#416

WalterKolczynski-NOAA mentioned this issue Aug 24, 2021

Enable coupled forecast in sprint branch #430

Merged

WalterKolczynski-NOAA linked a pull request Aug 24, 2021 that will close this issue

Enable coupled forecast in sprint branch #430

Merged

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 30, 2021

Modify ocean PETs to match prototypes

3dc7169

The numebr of PETs for the ocean model is changed to match the value used by the prototypes. Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Aug 30, 2021

Add runoff to data_table

a6030fd

Adds the runoff variable for the ocean model to data_table. (The table is otherwise empty.) Refs: NOAA-EMC#416

WalterKolczynski-NOAA added a commit to WalterKolczynski-NOAA/global-workflow that referenced this issue Sep 9, 2021

Remove duplicate do_tofd from namelist

e6cd4fb

do_tofd was being added to the namelist twice when coupling was off. Refs: NOAA-EMC#416

WalterKolczynski-NOAA closed this as completed Sep 15, 2021

WalterKolczynski-NOAA mentioned this issue Dec 2, 2021

Add coupled files to archive and streamline #517

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add S2SW forecast capability to coupled_sprint branch #416

Add S2SW forecast capability to coupled_sprint branch #416

WalterKolczynski-NOAA commented Aug 17, 2021

Add S2SW forecast capability to coupled_sprint branch #416

Add S2SW forecast capability to coupled_sprint branch #416

Comments

WalterKolczynski-NOAA commented Aug 17, 2021