Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

Incomplete Output for Domain 4 Despite Successful Run

Parvin

New member
Over the past month, I haven't been able to obtain a complete WRF output for domain 4. The model runs successfully without any errors, but the number of records in the wrfout_d04 file is randomly between 90 and 400, instead of the expected 1000. This issue did not occur in my previous runs with the same setup. I’ve attached my namelist.input file for reference. I would appreciate any suggestions or insights.
 

Attachments

  • namelist.input
    4.4 KB · Views: 1
In your namelist.input, I found the option below:

frames_per_outfile = 3000, 3000, 3000, 000

The number for D04 apparently is wrong. Can you change this option and try again?
 
Thank you for your guidance. I have attached the new namelist.input. Unfortunately, the issue still persists.
 

Attachments

  • namelist.input
    4.4 KB · Views: 1
Thank you for uploading your namelist.input. This file looks fine except that the time step is way too small. Can you modify the options as follows:
Code:
 time_step                           = 54
 dx                                  = 9000, 3000, 1000, 333.3
 dy                                  = 9000, 3000, 1000, 333.3
Then try again?

If the issue persists, it is worth trying
Code:
 time_step                           = 27
 dx                                  = 9000, 3000, 1000, 333.3
 dy                                  = 9000, 3000, 1000, 333.3

Please keep me posted about your results. Thanks.
 
Thank you very much for your feedback and suggestions.

Regarding the time step — the reason I am using a very small time step (9 seconds) is that the region I’m working on has steep terrain. Any larger time step results in a vertical CFL error during the simulation.

What’s puzzling to me is that previously, with these same settings, each output file for domain 4 used to include around 1000 records. However, for some time now, only about 90 to 400 records are being saved per run, despite having sufficient disk space available.

I’d really appreciate any insight you might have into why this change has occurred.
 
Hi Pavin,
Please see my comments below:
Thank you very much for your feedback and suggestions.

Regarding the time step — the reason I am using a very small time step (9 seconds) is that the region I’m working on has steep terrain. Any larger time step results in a vertical CFL error during the simulation.
If you increase epssm to a larger value, for example 0.9, can you overcome the CFL issue?
What’s puzzling to me is that previously, with these same settings, each output file for domain 4 used to include around 1000 records. However, for some time now, only about 90 to 400 records are being saved per run, despite having sufficient disk space available.
Which version of WRF did you run before? Note that namelist.input requires dx and dy over all domains. In your original namelist.input. you only set dx and dy over D01 but run over 4-nested domains. The case was done probably due to wrong reasons.
I’d really appreciate any insight you might have into why this change has occurred.
The history_interval generally should be Integer multiples of the time step. This is why I suggest that you try
Code:
 time_step                           = 54
 dx                                  = 9000, 3000, 1000, 333.3
 dy                                  = 9000, 3000, 1000, 333.3
 
Dear Ming,
Thank you very much for your support. I kindly ask for your help in resolving this issue.
My responses are sometimes delayed because, given the model’s time step, it takes a few days for me to see the results. The model version has not changed. The WRF model version is 4.5.2, which I have been using with these settings for about 10 months without any problems. This issue started in early June.
I increased epssm to 0.9, but whenever I select a time step greater than 9 seconds, the model becomes unstable and stops after about 15 time steps.
I have resent the updated namelist.input for your reference.
Thank you again for your help.
Best regards,
Parvin
 

Attachments

  • namelist.input
    4.4 KB · Views: 0
Parvin,

Please confirm that you run the same version of WRF ( i.e. WRFV4.5.2) before with the same namelist.input, and your case can be done successfully.
But since June, you have trouble running the same WRF and same namelist options.

If this is the case, can you recompile WRF, then try again? I am not sure whether your computer system was updated and possibly introduce some issues for WRF simulation.
 
Yes, I ran the same version with the same settings, and none of the configurations were changed on my side. I will try recompiling the model and running the simulations again. Thank you very much for your support.
 
Top