mark@atmos_colostate_edu
Member
Hello.
I'm attempting to run an extension of a BWSSP585 case with a special 83-level version of WACCM (CESM2.1.4) that Isla Simpson provided for us. Both branch and hybrid runs fail. I wrote interpolation scripts that convert the CAM initialization file and/or restart file to the 83-levels, and thought perhaps this could be the culprit, but it runs okay if I use a 2015 init/restart file. And it also runs fine if I continue the run with the 70 levels it was originally run with.
If I crank up these namelist settings to incredibly high values:
fv_nsplit = 256 ! number of dyn timesteps per physics timestep (default=0)
fv_nspltrac = 256 ! number of tracer advection timesteps per physics timestep (default=0)
fv_nspltvrm = 256 ! number of vertical remapping timesteps per physics timestep
it runs for about 10 days then crashes with this error message:
1147: ERROR: te_map: Lagrangian levels are crossing
Apparently the key log files are too large to upload, but I'm running this on Cheyenne, and the case directory is:
/glade/work/branson/BWSSP585
and the run directory is:
/glade/scratch/branson/BWSSP585/run.
Thanks for any help you can provide.
Mark Branson
I'm attempting to run an extension of a BWSSP585 case with a special 83-level version of WACCM (CESM2.1.4) that Isla Simpson provided for us. Both branch and hybrid runs fail. I wrote interpolation scripts that convert the CAM initialization file and/or restart file to the 83-levels, and thought perhaps this could be the culprit, but it runs okay if I use a 2015 init/restart file. And it also runs fine if I continue the run with the 70 levels it was originally run with.
If I crank up these namelist settings to incredibly high values:
fv_nsplit = 256 ! number of dyn timesteps per physics timestep (default=0)
fv_nspltrac = 256 ! number of tracer advection timesteps per physics timestep (default=0)
fv_nspltvrm = 256 ! number of vertical remapping timesteps per physics timestep
it runs for about 10 days then crashes with this error message:
1147: ERROR: te_map: Lagrangian levels are crossing
Apparently the key log files are too large to upload, but I'm running this on Cheyenne, and the case directory is:
/glade/work/branson/BWSSP585
and the run directory is:
/glade/scratch/branson/BWSSP585/run.
Thanks for any help you can provide.
Mark Branson