Making perturbations in model data

[dbarrie@atmos_umd_edu]

Hi CAM bulletin board,

I am attempting to run CAM in an ensemble mode by manually introducing perturbations into some atmospheric fields. However, I'm having difficulty formulating a sensible approach to this. The history files seem to be impossible to edit, and the initial conditions files do not carry all model fields. The huge dynamical adjustment I imagine accompanies an initial run would probably obscure any perturbations I would manually introduce in such a run. Any help would be greatly appreciated. Thanks.

-Dan

 

[rneale]

What kind of perturbations are you looking for? Do you intend to create just a simple ensemble of climate integrations where you don't really care how the initial conditions are perturbed? If so then run the model for 5 days and output an initial file on every day then use those 5 initial files to initialize 5 startup runs. This is pretty much the way it is done with amip ensemble runs.

If you want something more targeted so you can see the errors grow like in bred modes or singular vector analysis that is used in forecast model ensembles then that is more difficult.