search

NOAA-SWPC / GSMWAM-IPE

Coupled NEMS app for WAM-IPE
5 stars 11 forks source link

Workflow refresh #23

Closed akubaryk closed 4 years ago

akubaryk commented 4 years ago

This PR addresses three parts of #22: the extension of med.rc, the writing of SWIO output at the last model timestep, and a for-now default configuration of one-hour output for WAM-IPE and standaloneWAM SWIO output for each model component. It does not link SWIO output into the RUNDIR because of an apparent PNetCDF bug (or inconsistency with the standard NetCDF Fortran library) -- this has been confirmed with PNetCDF 1.10.0 and 1.11.2 on Hera, and 1.8.1 on WCOSS Phase 3. We have the options of cp'ing those files out manually, planning future work to pass a path in the SWIO resource file (which will require further changes to SWIO and COMIO to support), or diagnosing the issue in the PNetCDF library.

ZhuxiaoLi66 commented 4 years ago

I redownloaded the workflow_refresh branch and did an 1hr run, it looks that the gsm.20130316_010000.nc and ipe.20130316_010000.nc both have good file size at stmp4.

As for another way, pulled the updates to my original local workflow_refresh branch and did 'git submodule update --recursive' under GSMWAM-IPE directory, seems still can't get the right files. what else I missed to do?

rmontuoro commented 4 years ago

Commit 7e47447 allows to build the SWPC mediator intermediate 3D mesh using vertical levels imported from IPE if they are not provided in the mediator's configuration file.

rmontuoro commented 4 years ago

The workflow_refresh branch was updated to include the following additional requirements:

ZhuxiaoLi66 commented 4 years ago

num_density_io_refresh_2013031601 num_density_netcdf_output_20150317

ZhuxiaoLi66 commented 4 years ago

mms_io_refresh_2013031601

rmontuoro commented 4 years ago

Updated thermosphere mean molecular mass units to amu (commit abd5678b).

twfang commented 4 years ago

Everything works as expected. I did runs on both Hera and WCOSS and checked the Ne (TEC and fixed altitude) and drifts. I have also modified the output height level to 109 levels but all the way to 2000km.

ZhuxiaoLi66 commented 4 years ago

I did a 24hr workflow_refresh run, checked the number density and mean mass in the SWIO output and the original WAM output, the results are very close.