control_c768 RT test

[RuiyuSun]

Have anyone used control_c768 on Hera?

[gspetro-NOAA]

Hi @RuiyuSun ,

The regression tests are run regularly on Hera whenever new code is ready to be merged. The control_c768 test is run as part of that process. Do you have a specific question about the test?

Best,
Gillian

[BrianCurtis-NOAA]

@gspetro-NOAA Thanks for the clarification, I should have read up on the prior discussion a bit more. The weekly tests are not run, as far as I know. So there's a high chance that this run failure is due to a c768 configuration not getting the testing it needs. I also do not know if the c768 inputs have been updated as needed.

As stated, If there's any output that you do not understand, then feel free to post it here and hopefully you can get some direction on where to go.

[BrianCurtis-NOAA]

@RuiyuSun It does seem odd we have a C768 test without having the tiled fix files for it. My guess is we stopped supporting that resolution in regression testing a while back.

There are workflows available that can generate a set of tiled fix files for you at that resolution if you need them, though.

[RuiyuSun]

@BrianCurtis-NOAA I see. Thanks.

[gspetro-NOAA]

@RuiyuSun I got confirmation from @jkbk2004 that we no longer support this test, so the ICs won't be in the default location. However, you can generate the ICs yourself using GFS data (it takes a moment to load when you click Browse Files) and the UFS_UTILS chgres_cube utility to change the resolution. The WM wiki has some information on how to run a sample case, and it notes that you may need to modify file-related variables in rt.sh (e.g., dprefix, DISKNM) if you are staging the data in a non-default location.

[RuiyuSun]

Thank you @gspetro-NOAA for the instruction!