how to change the ice part timestep

[jyangdas@gmail_com]

Hi:

Recently, I try to run full-coupled CCSM3.0. I change nothing, except seting the solar radiation to be 80% of present level. The model blows up easily, due to "flux conservation error". Thus, I try to change the time step of ice part:

dt = 3600 ===>> dt = 1800
nydn_dt = 120 ===>> nydn_dt = 60
ndte = 1 ===>> ndte = 1

Surprisely, the model can not run again, and I don't get any error message from the log files. Can any body give me some suggestion？

Thanks very much.

Jun Yang
Jul, 06, 2010

 

[dbailey]

Hi,

The ndyn_dt is the number of dynamic subcycles per thermodynamic timestep. So, for example to decrease the timestep in the dynamics, you would try ndyn_dt = 2 to get a dynamic timestep that is half the thermodynamic timestep. However, you have a flux conservation error, which is a thermodynamic issue. This usually indicates an issue with your forcing. Look very carefully at the fluxes coming from the atmosphere and ocean. Make sure these look reasonable. Likely with the solar forcing at 80% present day level, something in your atmosphere is blowing up. If the forcing looks reasonable, and since the timestep reduction is not working, you might try lowering the ice albedos (albsnwi, albsnwv, albicev, albicei).

Dave