Errors in wave model when doing BHIST with f09_g17 resolution (CESM2.2.0)

[Yi Yao]

Dear, I want to run BHIST simulations with CESM2.2.0 and I encountered errors (both with and without restart files). More details can be found in followings:

1. model version: CESM2.2.0 (check the file model_version.txt)
2. case create: ./create_newcase --machine hortense --case /dodrio/scratch/users/vsc10154/cases/BHIST_IRR_01 --res f19_g17 --compset BHIST --project '2022_200' --run-unsupported
3. configure files regarding the machine Hortense are attached
4. I started firstly directly, then I got an issue in the wave model (*.log.4491942.230518-002420).

More detailed log files are in the attachment (*.log.4491942.230518-002420).
The cesm log file is too big so I got a screenshot here (don't think it gives any useful information):


5. If i want to restart from an existing CMIP6 cases, then I encoutered an error even ealier:


Btw, when we do another test with the resolution f19_g17, it works perfectly.

I appreciate it if you can give me some tips to solve this issue.

Thanks in advance,
Yi

 

[mlevy]

From the error

 *** WAVEWATCH III ERROR IN W3INIT : 
     NUMBER OF SPECTRAL POINTS LESS THAN NUMBER OF PROC.
     NSEA, NAPROC =     600    1024

 ERROR: WW3 EXTCDE abort

It looks like your PE layout isn't supported by the wave model -- at the current resolution, you need to run the wave model on 600 tasks or fewer. Can you please run ./pelayout from your case directory and report the the results? You probably want to run something like ./xmlchange NTASKS_WAV=600, but it would be helpful to see the general layout of your run.

 

[Yi Yao]

From the error

 *** WAVEWATCH III ERROR IN W3INIT :
     NUMBER OF SPECTRAL POINTS LESS THAN NUMBER OF PROC.
     NSEA, NAPROC =     600    1024

 ERROR: WW3 EXTCDE abort

It looks like your PE layout isn't supported by the wave model -- at the current resolution, you need to run the wave model on 600 tasks or fewer. Can you please run ./pelayout from your case directory and report the the results? You probably want to run something like ./xmlchange NTASKS_WAV=600, but it would be helpful to see the general layout of your run.

Dear Michael, thank you so much for your help! Finally I solved the issue without restartfiles. But for the simulations with restart files, I still encountered issues with restart files as I showed before. My question is: is there difference between the restart files in CESM2.1 and CESM2.2?
Thank in advance!

Cheers,
Yi