Case.submit is not successful

[arti-14]

Hi,
I am trying to run the CAM simulation in single column mode.
I am using the large scale forcings from the observations as the IOP data. The initial conditions file, is generated by the CAM (following the instructions given in : 4. Atmospheric configurations (compsets) — camdoc documentation )
But the model is not running.
I have attached the log files generated during the run fro refernce.

I really appreciate help regarding it.
Thank you

 

[jet]

Hi Arti:

Whenever I run into an error (segfault in this case) I always recompile the model with DEBUG set to true in the env_build.xml file and rerun the model. The cesm.log doesn't have any information other than the model died with a segfault. Hopefully debug will give us more information.

cd CASE_DIRECTORY
./xmlchange -id DEBUG -val "TRUE"
./*.build --clean-all
./*.build
./*.submit

Could you also send me your initial file. I see you got by your last error creating it. I think I will need to try it myself.
Just FYI SCAM will work with a 3d initial condition and choose the column closest to the point that is asked for in the namelist. Most of the regridding utilities are made to work with 2d/3d datasets and I think the error you ran into before was due to the single point that you were trying to remap.

jt

 

[arti-14]

Hi,
Thank you so much for your help and reply.
I created the intial file form the CAM simulation for a single point, and then regridded it.
Do I have to run the CAM simulation for a few points, and then regrid it ?
I have copied the initial file to " Arti jadab sent you 1 item "

I followed the steps mentioned above, I have attached the build.log for reference.

Best Regads,
Arti

 

[John Truesdale]

Hi Arti.

There are two files that need to be created to run SCAM. The first is an initial condition file that sets the initial state of the model. For SCAM this initial condition file also has the starting emission fields overwritten by their monthly average. During the SCAM run the emission state Your initial condition file looks OK after a quick check of the fields. Next there is the high frequency IOP file that provides forcing every timestep or every few hours for the IOP run. Can you send me the IOP file that you created as well as your atm_in namelist file.

Thanks
jt

 

[John Truesdale]

ignore the "During the SCAM run the emission state" sentence fragment above. It also looks like you rebuilt the model with debug, could you run it and send me the new atm.log and cesm.log files as well as the IOP file and atm_in namelist. It will still fail but hopefully the atm and cesm logs will have more information about where it is dying.

jt