Picard Convergence Failure

[zwolff@uci_edu]

Hello, I  am running CICE v5 at the default global 3 degree resolution (gx3 grid provided) and the only changes being I am reading in NCEP/NCAR Reanalysis NetCDF files as data (starting with 1980) and almost immediately the model crashes with a picard convergence error. The exact error message from the ice.log file is below.   picard convergence failed!           0  -1.9045826499242648                            NaN           1   0.0000000000000000        0.0000000000000000       -110220000.00000000                1  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                2  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                3  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                4  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                5  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                6  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                7  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649      ------------------------------------- istep1, my_task, iblk =           3           0           3 Global block:           5 Global i and j:          12          38 Lat, Lon:  -15.345122324731349        1.8000000500896045      aice:  0.84439121202229650      ice: Vertical thermo error: temperature_changes_salinity: Picard solver non-convapplication called MPI_Abort(MPI_COMM_WORLD, 128) - process 0 ------------------------------------- picard convergence failed!           0  -1.9045826499242648                            NaN           1   0.0000000000000000        0.0000000000000000       -110220000.00000000                1  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                2  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                3  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                4  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649                5  -2.3330864777318179                            NaN   31.000000000000000        1.0000000000001563E-002  0.75000000000000022       -85268533.086176649      Is this an error with the data, the way it is being read in, or the model configuration? Thank You, Zach Wolff 

 

[dbailey]

Sorry to take so long on this. I was away on vacation. The first thing that concerns me is the NaN appearing in your output. Also, this is a Southern Hemisphere equatorial point where there should not be ice, unless you are running a snowball Earth simulation. I would first check that your forcing is correctly interpolated onto the gx3 grid and make sure it is what you expect. Then I would double check the initial condition (ice_ic) file you are using.Dave