dtc/develop: update from develop 2020/01/27#20
Conversation
* add fh00 post control file, add restart output at specified forecast hours, ugwd bug fixes * update module files to use g2tmpl v1.6.0 and post v8.0.1 * add fh00 post control file in fv3.input for inline post * update in runduration in atmos * update RT baseline directory on hera and wcoss * comment out fv3_wrtGauss_nemsio_c768 on dell * update grib2 table for g2tmpl 1.6.0 * update post * comment out print line in FV3
* update FMS to point to tag 2019.01 of NOAA-GFDL * update stochastic_physics to point to hash 1745422af76d830757cd6035b6ea101e92b4cac1 @pjpegion * update of path to CCPP physics library for CCPP regression tests * add -Wall to compiler flags for GNU compilers * remove warnings for non-existent include directories for GNU compiler * update of regression testing scripts to detect errors in tests/run_test.sh * new regression tests fv3_ccpp_gfs_v15p2, fv3_ccpp_gfs_v15p2_debug, fv3_ccpp_gfs_v16beta, fv3_ccpp_gfs_v16beta_debug * update of rt.conf: remove Cheyenne.intel entries (so that rt.conf is an EMC-maintained regression test configuration; move Cheyenne Intel tests into a separate file that resides in the NCAR dtc/develop branch) * make compile_cmake.sh work with Cheyenne Intel/GNU, and bugfixes for compile_cmake.sh * allow environment variable NEMS_MACHINE to overwrite (or set) MACHINE_ID (see ufs-community#20 for the corresponding PR for the ufs_public_release branch) * new regression test fv3_ccpp_gsd_sar_25km_debug (but not exercised in default rt.conf) * updates for jet and gaea (note: only supporting rt.sh, not NEMSCompsetRun)
* Updated fv3atm (see fv3atm PR ufs-community#48 NOAA-EMC/ufsatm#48) * Updated WW3
…gression test config tests/rt_ipd_prod_only.conf
|
Regression testing on hera.intel: run IPD PROD tests in rt.conf against EMC baseline - all pass. Next, create full baseline for rt.conf (IPD PROD, IPD REPRO, CCPP PROD) and verify against it (IPD PROD, IPD REPRO, CCPPREPRO, CCPP PROD) - all pass. rt_full_create.log Rsync the newly created baseline to the DTC regression test baseline directory (use instead of EMC baseline directory). Then, verify again against the DTC baseline using full rt.conf - all pass. |
|
Regression testing on hera.intel for rt_ccpp_dtc.conf, rt_ccpp_hafs.conf, rt_ccpp_gsd.conf, rt_ccpp_caps.conf (first create, then verify): all pass. rt_ccpp_caps_create.log |
PBL tendencies were missing in two schemes; now fixed. Squashed commit of the following: * fix bugs found in pbl and ozone 3d diagnostic tendencies * remove debugging prints * implied shape arrays for five variables * more block labels in CCPP * yet more bug fixes in ccpp * make the logging less wordy * point to hash 238c84c of ccpp/physics * point to merged version of FV3, with latest ldiag3d/qdiag3d bug fixes
Update dtc/develop from develop:
For regression testing information, see below.