Error: Process killed

[nickluos]

Hi,

I am experiencing a persistent WRF crash at a specific simulation time. WRF 4.7.1 consistently crashes at 2019-06-19 05:19:32 UTC across multiple attempts
with different configurations. The previous day (June 18) completes successfully with identical domain and physics settings, but June 19 always fails at the same moment.

Before this try, it crashed at 06-19 02:22 when I used a longer time_step=15.
I would like to ask if this error is related to the time step size or is it due to insufficient memory? Or there are still other boundary conditions that were not considered, which led to this situation？
My file location is /glade/derecho/scratch/tianzel/wrf_OKC_201906_revised.

Many thanks!

 

[Ming Chen]

Hi,
In your namelist.input, you set

cu_physics = 6, 6,

but your grid intervals are 3km and 1km respectively.

In addition, your radt is way too large for 3km run.

In this case, I would suggest that you turn off cumulus scheme, and set radt =3, then try again.

Please let me know if your case still fails with the changes of the above options. Thanks.

 

[nickluos]

Thank you very much. After re-submission, there were no errors and the result looks good.

Hi,
In your namelist.input, you set

cu_physics = 6, 6,

but your grid intervals are 3km and 1km respectively.

In addition, your radt is way too large for 3km run.

In this case, I would suggest that you turn off cumulus scheme, and set radt =3, then try again.

Please let me know if your case still fails with the changes of the above options. Thanks.

 

[Ming Chen]

Thanks for the upadate. I am glad it works!