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

CFL Error after several hours

This post was from a previous version of the WRF&MPAS-A Support Forum. New replies have been disabled and if you have follow up questions related to this post, then please start a new thread from the forum home page.

jahokson

New member
Hi!
I'm getting a CFL error as follows (rsl.out.0012, see link):
d01 2015-10-15_22:13:30 1 points exceeded cfl=2 in domain d01 at time 2015-10-15_22:13:30 hours
d01 2015-10-15_22:13:30 MAX AT i,j,k: 18 202 33 vert_cfl,w,d(eta)= 2.208760 6.219230 7.6782177E-03
d01 2015-10-15_22:14:51 13 points exceeded cfl=2 in domain d01 at time 2015-10-15_22:14:51 hours
d01 2015-10-15_22:14:51 MAX AT i,j,k: 18 202 33 vert_cfl,w,d(eta)= 11.44821 -3.609830 7.6782177E-03

After such error, a "forrtl: severe (174): SIGSEGV, segmentation fault occurred" is shown (rsl.error.0012, see link).

Here are the rsl.error, rsl.out, and namelist.input files: https://drive.google.com/drive/folders/1WUOBP7rGL2SvS9qruHm9bDS4YC7SJlqx?usp=sharing

I'm running a preset moving nest simulation, using a researh system that uses WRF (WRF 3.6.1) as part of the computations.

Update: I have tried running with w_damping=1, but I am still geting a cfl error.

Any help is appreciated. Thank you.
 
Hi,
I looked at your namelist.input. This seems like a moving nest case. Thee are a few issues I am concerned:
(1) Can you run with the automatic moving nest? We don't recommend the option of specified moving nest because it is hard to correctly set the options like move_cd_x, move_ccd_y, move_inteval, etc.
(2) Please turn off adaptive time step and sf_ocean_physics. This will make the case less complicated, which will be helpful for us to figure put what is going.
(3) Please set radt = 27, 27, 27
(4) What is the option of isftcflx = 99?
 
Hi Ming Chen! Thank you for your reply.
To answer your questions:
1. Unfortunately, I need the moving nests. I have a model that computes for the positions.
2. I have tried turning off adaptive time step, but I still get CFL errors. I will try to turn of sf_ocean_physics with adaptive time step off. I will update you soon.
3. Got it.
4. Oh. I think it's a mistake. Should I set this to 0 ?

Thanks
 
Hi, Ming Chen!
Sadly, even with changes (# 2-3), I'm still getting cfl errors. Here are the files this time:
https://drive.google.com/drive/folders/1caIqACkp3ujsZdR0_gJ6qCCaQAqGwb4K?usp=sharing

Thank you
 
Hi, jahokson,
Your namelist.input looks fine except that the time step can be set to 150. The current setting of 60 is small for dx=27km. Also, please turn off feedback and rerun this case. let me know whether it can run at least longer.
I am not that experienced with the preset moving nest and cannot determine whether the errors are caused by the preset options. If this case keeps failing, you may need to print the variables related to CFL violation, and identify possible reasons.
Please keep me updated about the progress. Thanks.
Ming
 
Top