Segmentation Fault when running high resolution over Bogota region

[gfarache]

Hello. I'm trying to execute WRF model over Bogota region with the 3 nested meshes with respectively 9km, 3km and 1km resolution.
The model starts to integrate, but after a few timesteps it crashes indicating "Program received signal SIGSEGV: Segmentation fault - invalid memory reference."
I know that the problem is occurring over the inner domain (1km resolution) because when it is executed only for the 2 coarser nested grids, it runs well. But, I don't know how to solve it or what is the real problem. Anyone could help me?
I'm attaching the namelists and the rsl.out/error of the processor that crashes.

WRF Version: 4.5.2
WPS Version: 4.3.1

 

[William.Hatheway]

Hello. I'm trying to execute WRF model over Bogota region with the 3 nested meshes with respectively 9km, 3km and 1km resolution.
The model starts to integrate, but after a few timesteps it crashes indicating "Program received signal SIGSEGV: Segmentation fault - invalid memory reference."
I know that the problem is occurring over the inner domain (1km resolution) because when it is executed only for the 2 coarser nested grids, it runs well. But, I don't know how to solve it or what is the real problem. Anyone could help me?
I'm attaching the namelists and the rsl.out/error of the processor that crashes.

WRF Version: 4.5.2
WPS Version: 4.3.1
View attachment 16242

First of all happy to see the domain wizard tool being used.


Second did you attach all the rsl.out and rsl.error files? Because if you are using multiple cores then there are different files that might lead to the issue.

you can also use these commands in the /run folder to find which file may contain the issue.

grep -i FATAL rsl.*

grep -i error rsl.*

grep -i SIGSEGV rsl.*

grep -i cfl rsl.*

 

[gfarache]

Hey William, thank you so much for the answer!
I believed that the process with the error was the same as one indicated by the error during the MPI execution.
Using these commands that you recommended, I found out that it was occurring a CFL error, in a process other than the indicated one.
The error is happening in the finer domain, so can I just increase the time step ratio of this domain, or it is recommended to decrease the general time step execution?

 

[William.Hatheway]

Hey William, thank you so much for the answer!
I believed that the process with the error was the same as one indicated by the error during the MPI execution.
Using these commands that you recommended, I found out that it was occurring a CFL error, in a process other than the indicated one.
The error is happening in the finer domain, so can I just increase the time step ratio of this domain, or it is recommended to decrease the general time step execution?

w_damping = 0, -> w_damping = 1,

try that first

 

[gfarache]

Sorry for the delay.
After applying your suggestion, it's working well. Thank you so much William, w_damping = 1 solved the CFL problem.

 

[William.Hatheway]

Sorry for the delay.
After applying your suggestion, it's working well. Thank you so much William, w_damping = 1 solved the CFL problem.

glad to be of help