Update from emc/develop 2020/03/12 and other changes

[climbfuji]

Reconfigure the regression test configurations

• stochastic physics baselines now generated by CCPP REPRO and not IPD REPRO
• turn of IPD REPRO Thompson test
• add Thompson aerosol and no-aerosol test in CCPP PROD mode
• add Thompson aerosol and no-aerosol test in CCPP DEBUB mode, add GSD-suite test in DEBUG mode
• updates for fv3atm, GFDL_atmos_cubed_sphere, ccpp-physics as described in the associated PRs below

Associated PRs:

NCAR/GFDL_atmos_cubed_sphere#10
NCAR/ccpp-physics#407
NCAR/fv3atm#30
#27

For regression testing information, see below.

[climbfuji]

Regression testing on hera.intel using full rt.conf against existing/old baseline. The following tests fail as expected (because these tests were newly added, i.e. missing baseline; manually checked that all those tests ran to completion with an exit code zero):

log_hera.intel/rt_132_fv3_ccpp_thompson_prod
log_hera.intel/rt_133_fv3_ccpp_thompson_no_aero_prod
log_hera.intel/rt_138_fv3_ccpp_gsd_debug_prod
log_hera.intel/rt_139_fv3_ccpp_thompson_debug_prod
log_hera.intel/rt_140_fv3_ccpp_thompson_no_aero_debug_prod

All other tests pass.

Compile_hera.intel.log
RegressionTests_hera.intel.log
rt_full_against_existing_baseline.log

[climbfuji]

Regression testing on hera.intel using rt_ccpp_dtc.conf: first create baseline, then verify against (i.e. compare CCPP against IPD) - all tests pass

Create:

Compile_hera.intel.log
RegressionTests_hera.intel.log
rt_ccpp_dtc_create_rocoto.log

Verify:

Compile_hera.intel.log
RegressionTests_hera.intel.log
rt_ccpp_dtc_verify_rocoto.log

[climbfuji]

Regression testing on hera.intel using full rt.conf, creating new baseline, rsync to final location, then verify against it (includes comparing IPD REPRO versus CCPP REPRO)

Create - all pass:

Compile_hera.intel.log
RegressionTests_hera.intel.log
rt_full_create_ecflow.log

Verify - all pass:

Compile_hera.intel.log
RegressionTests_hera.intel.log
rt_full_verify_ecflow.log

[llpcarson]

approved

[climbfuji]

@junwang-noaa @DusanJovic-NOAA for your information, next round of updates going into NCAR dtc/develop in preparation for merging back to EMC develop / NCAR master. More to follow (e.g. Shan's changes on top of Moorthi's, Moorthi's levr bugfix).