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

WRF fails without providing error message


New member

I've been having some issues with getting WRF simulations to complete (no issues with completing real.exe). Initially, I was receiving cfl and TBOUND exceeds table limit error messages, but those have both now been resolved. Unfortunately, I'm still unable to get a simulation to run more than a day of simulation time. It is surprisingly running for less time now than it was when I was receiving the error messages. The rsl* files currently don't contain cfl or TBOUND messages, but I do get the following message:
*WARNING** Time in input file not equal to time on domain **WARNING**
 **WARNING** Trying next time in file  ...
I've had this message during other simulations, but, like with my most recent simulation, it always continues to run for some time even after it appears. From what I can tell, all of the wrfinput, wrfbdy, and wrflowinp files and namelist.input have the same dates, so I'm not sure why it gives me the warning. Is there anything else I can check regarding this error or any others that I haven't checked for yet? Thank you
Can you attach all of your rsl* files (packaged into a single *.tar file), as well as your namelist.input file? Thanks!
Thanks for sending those. The *WARNING* messages you are seeing aren't causing the model to stop. Those are simply stating that the model is looking for the next time available in your wrflowinp* files (because your input files are available every 3 hours, and your wrflowinp boundary times are available every 6 hours). This is not anything to worry about. Unfortunately the rsl files don't provide a lot of useful information. Can you try a few of things:

1) I notice you have p_top_requested set to 10000. Unless you need to have it set to that level, can you try the default value of 5000, instead?
2) Try running with more processors. Can you try maybe 72?
3) If that doesn't work, will you try running with just a single domain, and if that works, try two domains. This will help to know which domain is causing the issue, if one is.
4) Finally, if none of the above help with anything, can you try to just use a very basic (default) namelist - only changing the times and domain settings to match your case, but keeping all other options simple (i.e., don't modify physics, don't even use sst_update). If this works, you will know one of the settings you've introduced is causing the problem, and from there, you can try to slowly add settings back in until you can make it fail, to determine the culprit.