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-Chem Model Error: SIGABRT and Segmentation Fault Issues

S

Sanatan

New member
Hello everyone,

I am running the WRF-Chem (4.4) model and have encountered several issues that I hope to get some assistance with. Below are the details of the errors I am experiencing. The relevant files are also attached for reference.


munmap_chunk(): invalid pointer

Program received signal SIGABRT: Process abort signal.

Backtrace for this error:

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:

And in rsl.out, the message is

in above message, i1 =-391202822
*** module_cmuaq_bulk, aubr aqintegr1 -- svode failed, istate = -3
*** mosaic_cloudchem_driver, subr interface_to_aqoperator1
id,it,jt,kt, istat_fatal, warn = 1 28 21 4 -10 0
dvode-- maxord (=i1) .lt. 0
in above message, i1 =-391202822
*** module_cmuaq_bulk, aubr aqintegr1 -- svode failed, istate = -3
*** mosaic_cloudchem_driver, subr interface_to_aqoperator1
id,it,jt,kt, istat_fatal, warn = 1 28 21 5 -10 0
leaving mosaic_cloudchem_driver - ktau = 1 3


As this seems to be a processor issue, I have tried several options such as 4, 8, 16, or 32 processors, but ended up with the same results. Also, I have enough memory in my system (more than 4TB). As I am a beginner, I guess I am committing some mistakes while setting up the namelist, or it can be something else that I might not be able to figure out.

Any guidance or suggestions on how to resolve these issues would be greatly appreciated.

Thank you in advance for your help!
Sanatan.
 

Attachments

  • rsl.error.0000
    43.4 KB · Views: 1
  • namelist.input
    8.2 KB · Views: 5
I would double check your timesteps and start end date.

Right now it looks like you have a 500sec time step which is enormous and a outer domain of 100km grid space.

also could you please upload your namelist.wps file?
 
Thank you very much William for your time.
Here I have attached namelist.wps file to this message.

Additionally, I should also tell you that when I tried chem_opt=301 and dust_opt=1, the model ran successfully. The namelist with a successful run is also attached to this message named "chem-301-namelist.input".
 

Attachments

  • namelist.wps
    835 bytes · Views: 4
  • chem-301-namelist.input
    8 KB · Views: 5
@William.Hatheway, For chemical boundary conditions, I am using WACCM data (i.e., waccm-20240628124350715760.nc), and for meteorology I am using fnl data. Attached is the namelist used in MOZBC.
 

Attachments

  • mozbc.input
    1.3 KB · Views: 3
This is interesting, so I got the same error "munmap_chunk(): invalid pointer" today while trying a new WRF (v4.4) configuration.

I have been using this same 4.4 version of WRF for roughly two years, but today, I tried implementing some new namelist options and static data (the new MODIS 100m LCZ global land cover). Note I have implemented tons of custom static data previously with no issues. In this new configuration, I turned on NOAH-MP (typically use NOAH) and turned on BEP-BEM (typically use BEP alone) with the namelist "use_wudapt_lcz" to enable the new LCZ map. I switched back to NOAH, and the same error occurred. I believe it concerns me turning on the "use_wudapt_lcz" or BEM options with the new data. Still, I have not done the leg work to figure out what was causing the problem because I simply built WRF v4.6.1, and it works ... not entirely helpful, but maybe some context as to what might be playing into the issue!
 
You must log in or register to reply here.
Top