Closed qianyuzhao closed 2 years ago
Thanks for writing. I think the issue is that INIT_UCX is being called for the aerosol-only simulation but it should only be called for the fullchem simualtion. See the fix in commit https://github.com/geoschem/geos-chem/commit/0d2b8eaea9ae26c159b99083fb2382d017c48d2c, which will go into GEOS-Chem 13.4.0.
Thanks for writing. I think the issue is that INIT_UCX is being called for the aerosol-only simulation but it should only be called for the fullchem simualtion. See the fix in commit 0d2b8ea, which will go into GEOS-Chem 13.4.0.
Thanks for the reply! Finally, the 2x2.5 simulation works by setting "export OMP_STACKSIZE = 14g" and using GEOSFP on our end. The 0.25x0.3125 simulation works by setting "export OMP_STACKSIZE = 256g" on our end. And the version we used is GEOS-Chem 13.3.2.
Great. I'll close out this issue. Thanks for letting us know @nbderenzqy66.
What institution are you from?
Baylor University
Description of the problem
I tried default settings to run GC v13.3.2 simulation of aerosol with MERRA2 2x2.5 and 0.5x0.625 (boundary conditions generated from MERRA2 4x5). The error message from the cluster is: "/var/spool/pbs/mom_priv/jobs/632778.batch.SC: line 36: 20349 Segmentation fault (core dumped) ./gcclassic > GC_13.log" I checked GC_13.log and HEMCO.log, there is no error in those files. I also followed Inter Fortran Compiler (http://wiki.seas.harvard.edu/geos-chem/index.php/Intel_Fortran_Compiler#Resetting_stacksize_for_Linux) and added ulimit commands both before compiling and in my bash file. The same error occured. I tried GEOSFP, GC v13.2.0, GC v12.9.3, all of them occurred segmentation faults.
GEOS-Chem version
13.3.2.
Log files
Build log (if applicable):
Run logs (if applicable): GC_13.log HEMCO.log error.txt
Run script (if applicable): aerosol_v13.txt
Software versions