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

Regarding max_dom = 4, Kindly advise on namelists, works well till max_dom =3, then comes out without any information in log etc.

shaildph

New member
I am running WRF 4.3 on AWS cloud, using GFS .25 degree inputs, it works well and produce output for max_dom =3, but when I try to generate the same for max_dom =4, as I wish to simulate at 1 km resolution, it fails, stops in between without any info, usually in max_dom =3, I get tail rsl.out.0000 "successful wrf", I would appreciate your help and guidance. I have attached namelist.wps and namelist.input for your kind perusal.
Regards
Shailendra
 

Attachments

  • namelist.input
    4.5 KB · Views: 8
  • namelist.wps
    949 bytes · Views: 2
cu_physics : It is advised in WRF Solar,
and ratio, I have tried both, what is your advise, is that the only reason for not getting completed in the fourth domain case, Please suggest !!
 
Dear Sir,
I have attached the rsl files as suggested by you, kindly help. Thank you. Regards. Shailendra
@shaildph
Please package your rsl* files together into a single *.tar file and attach that so that we can take a look. Thanks!
Dear Sir,
May I request any updates on this, also, is there a correlation between no. of cores, RAM etc. with the number of domains/size of the domain. Please help. Regards Shailendra
 
I have a few concerns about your namelist options., Please modify the following options:

(1) input_from_file = .true.,.true.,.true.,.true.
(2) parent_time_step_ratio = 1, 3, 3, 3,
(3) radt = 15, 15, 15, 15,
(4) cu_physics = 1, 1, 0, 0,
(5) time_step = 90

Then rerun this case.
 
I have a few concerns about your namelist options., Please modify the following options:

(1) input_from_file = .true.,.true.,.true.,.true.
(2) parent_time_step_ratio = 1, 3, 3, 3,
(3) radt = 15, 15, 15, 15,
(4) cu_physics = 1, 1, 0, 0,
(5) time_step = 90

Then rerun this case.
Dear Sir,
This fails again, though it works well if I make max_dom = 3.
May I request you help me get a sample namelist.wps and namelist.input, which are already tested (working) for max_dom =4.
Kindly help.
Regards
Shailendra
 
Hi Shailendra,

Did you get different error messages with the modified namelist? What are the error messages? How long did the model run before it crashed?

Please see the attached namelist.input for 4-domain nested case. Note that this is just a smaple pone and you need to modify it based on your specific case.
 

Attachments

  • namelist.input.txt
    3.4 KB · Views: 1
Hi Shailendra,

Did you get different error messages with the modified namelist? What are the error messages? How long did the model run before it crashed?

Please see the attached namelist.input for 4-domain nested case. Note that this is just a smaple pone and you need to modify it based on your specific case.
Thank you Sir! I am working with it, will get back to you on this. Regards Shailendra
 
For interruptions in the integration process due to unknown reasons,try using a shorter time_step may be a good idea
 
For interruptions in the integration process due to unknown reasons,try using a shorter time_step may be a good idea
In fact, it was still failing (possible due to large time_step), may I request you to kindly elaborate it, I am sorry, I am new to this field. I would really appreciate it. Regards Shailendra
 
In fact, it was still failing (possible due to large time_step), may I request you to kindly elaborate it, I am sorry, I am new to this field. I would really appreciate it. Regards Shailendra
This is a way used to address a type of problem, such as "too large data difference leads to the integration process unable to converge",
shorter time steps may make the data smoother
 
Can you post your namelist.input (the one for 4-domain nesting) ? Also, what is the error message when the model crashed?
 
Top