Main menu

Navigation

POP "SOLVER NOT CONVERGED" problem

1 post / 0 new
11438023@...
POP "SOLVER NOT CONVERGED" problem

Dear all,

 

I am running with B_1850_CN configuration and the resolution is f19_g16.    What I did is to remove the seasonal and diurnal cycle of solar insolation in CAM. Of course I changed the processes computing solar zenith angle in other components as well. No change was made to model physics. 

 

The model crashed at year 126 (run_type is startup and coldstart is on) with information “SOLVER NOT CONVERGED”.    I followed the suggested method and decreased POP2 time step from dt_count = 23 to dt_count = 27. However, the model still crashed after 1 year (year 127).    Again I decreased dt_count = 29;  this time the model successfully run for 5 years.      

 

I wrote restart file every year.    I continued my default run with default de_count using the restart file (from year 127, 128, 129, 130, and 131). However, every time the model crashed with the same problem “Error in step…    Solver not converged”.

 

What should I do to overcome this problem?   

 

 

Best,

 

Lei

Who's new

  • rlove@...
  • katsumi@...
  • afox
  • shanru@...
  • yongxiao@...