CICE | ERROR: ice_read_global_nc: Cannot find variable ulat

[YCCircle]

Hi,

I'm trying to run a SOM case with sea ice:

Compset: 2000_CAM60_SLND_CICE_DOCN%SOM_SROF_SGLC_SWAV
Grid: f09_g16

SOM forcing file: pop_frc.b.e11.B1850C5CN.f09_g16.005.150217.nc

user_nl_cice:
kdyn = 1
kitd = 1
ktherm = 2
prescribed_ice = .false.
grid_file = '.../my_cesm_sandbox/cesm_input_datasets/share/domains/domain.ocn.gx1v6.090206.nc'
kmt_file = '.../my_cesm_sandbox/cesm_input_datasets/share/domains/domain.ocn.gx1v6.090206.nc'
grid_format = 'nc'
grid_type = 'displaced_pole'
ice_ic = '.../my_cesm_sandbox/cesm_input_datasets/ice/cice/b.e15.B1850G.f09_g16.pi_control.25.cice.r.0041-01-01-00000.nc'

After submitting, I met an error written in cesm.log file:

00: ERROR: ice_read_global_nc: Cannot find variable ulat
00: Image PC Routine Line Source
00: cesm.exe 000000000260C28A Unknown Unknown Unknown
00: cesm.exe 000000000236C7FD shr_abort_mod_mp_ 114 shr_abort_mod.F90
00: cesm.exe 000000000236C8AA shr_abort_mod_mp_ 61 shr_abort_mod.F90
00: cesm.exe 000000000202EDCF ice_exit_mp_abort 46 ice_exit.F90
00: cesm.exe 000000000212D818 ice_read_write_mp 1868 ice_read_write.F90
00: cesm.exe 0000000002049830 ice_grid_mp_init_ 171 ice_grid.F90
00: cesm.exe 00000000021A47E4 cice_initmod_mp_c 96 CICE_InitMod.F90
00: cesm.exe 0000000002027A72 ice_comp_mct_mp_i 262 ice_comp_mct.F90
00: cesm.exe 00000000004384EF component_mod_mp_ 257 component_mod.F90
00: cesm.exe 000000000042653F cime_comp_mod_mp_ 1371 cime_comp_mod.F90
00: cesm.exe 0000000000435487 MAIN__ 122 cime_driver.F90
00: cesm.exe 0000000000416D1E Unknown Unknown Unknown
00: libc-2.17.so 00002B34FE635B35 __libc_start_main Unknown Unknown
00: cesm.exe 0000000000416C29 Unknown Unknown Unknown
00: application called MPI_Abort(MPI_COMM_WORLD, 1001) - process 0
00: slurmd[cn7581]: *** STEP 6359023.0 KILLED AT 2023-02-06T19:35:03 WITH SIGNAL 9 ***
yhrun: Job step aborted: Waiting up to 2 seconds for job step to finish.
00: slurmd[cn7581]: *** STEP 6359023.0 KILLED AT 2023-02-06T19:35:03 WITH SIGNAL 9 ***
yhrun: error: cn7582: tasks 16-31: Killed

Is this because the cice namelist is set incorrectly? Any suggestions are appreciated. Thank you.

Best,
Rose

 

[dbailey]

These grid and kmt files are not right for CICE. It needs the POP style grid and kmt files as found in /glade/p/cesm/cseg/inputdata/ocn/pop/gx1v6/grid. These are binary files. I don't believe we have netCDF versions of these.

 

[YCCircle]

Thank you, Dave! You help me a lot. :D

Compset: 2000_CAM60_SLND_CICE_DOCN%SOM_SROF_SGLC_SWAV
Grid: f09_g16

SOM forcing file: pop_frc.b.e11.B1850C5CN.f09_g16.005.150217.nc

user_nl_cice:
kdyn = 1
kitd = 1
ktherm = 2
ice_ic = '/BIGDATA1/yt_ust_shixm_2/huangyy/my_cesm_sandbox/cesm_input_datasets/ice/cice/b.e15.B1850G.f09_g16.pi_control.25.cice.r.0041-01-01-00000.nc'
grid_file = '/BIGDATA1/yt_ust_shixm_2/huangyy/my_cesm_sandbox/cesm_input_datasets/ocn/pop/gx1v6/grid/horiz_grid_20010402.ieeer8'
kmt_file = '/BIGDATA1/yt_ust_shixm_2/huangyy/my_cesm_sandbox/cesm_input_datasets/ocn/pop/gx1v6/grid/topography_20090204.ieeei4'
grid_format = 'bin'
grid_type = 'displaced_pole'

I changed the grid_file and kmt_file as shown above, but still met errors after running for several minutes or several hours. I've tried to recreate and rebuild the case, however, the errors are different every time. I did nothing to the compset, grid files, namelists, but changed the NTASKS, NTHRDS, and DEBUG these times.

first time: forrtl: severe (174): SIGSEGV, segmentation fault occurred

second time: forrtl: error (73): floating divide by zero

third time: Error calling advance_xm_wpxp
Fatal error in CLUBB: at timestep 347

I browse the Forum but cannot find a clear and effective solution. Are there any suggestions? Thanks.

 

[dbailey]

This is more of a CAM question. I am going to move it to the CAM forum.