ncar/ufs-weather-model

UFS Weather Model

ufs-weather-model's Issues

fv3_ccpp_gfdlmp threading issue for dynamic CCPP build

The fv3_ccpp_gfdlmp regression test setup hangs on macOS for the dynamic CCPP build after the initialization if more than one thread is used. The fv3_gfdlmp (i.e. the non-CCPP test) works fine, and so does the static CCPP build. This problem does not occur for the fv3_ccpp_control test.

Revert submodule for WW3 to use authoritative repository

Change back from NCAR gmtb-ww3 to NOAA-EMC WW3.

dt/develop: new regression test fv3_ccpp_gsd_sar_v1 failing

This test was added in PR #45, but failed because of a missing file parm/ccpp_gsd_sar_v1.nml.IN.

This does not effect any of the other tests.

No regression test for newly added Ferrier-Aligo MP

When #1, the regression test configuration files were added to the repo, but no entry was made in any of the rt*.conf files. It was agreed that the test should go in rt_ccpp_dtc.conf, for example like this:

# Compile with CCPP - dynamic mode
COMPILE | CCPP=Y REPRO=Y | standard | hera.intel | |
# Run tests, input data currently only available on Theia
RUN | fv3_ccpp_regional_c768_FA_update_moist | standard | |

Regression tests using csawmg crashing with GNU compilers

All regression tests involving csawmg (with or without SHOC) are crashing with the GNU compilers. This used to be the case on Theia and Cheyenne with versions 6.x and macOS with versions 6.x/7.x. Needs to be retested on Cheyenne and macOS with newer versions of the compiler.

SDF basic documentation

(Transferred from https://github.com/NCAR/NEMSfv3gfs/issues/97)

Now that SDFs reside in host model repositories and given their propensity to multiply during testing, I think that it would be a good idea to provide some kind of high level documentation on what the suites represent more than just relying on the name of the file. Are comments allowable in the SDFs? Failing that (or in addition), could there be one text file in the /ccpp/suites directory that lists the suites in the repo and what they are and how they differ? Right now, for folks outside of GMTB, I imagine that it is pretty difficult to get a feel for what the suites are and how they differ -- basically they have to read through the SDFs line by line and compare.

From @climbfuji:
Comments are allowed:

Sounds like a very good idea for the UFS release.