Internal Variability and stochasticity

[jzweifel]

Hello CESM people!

(Useful background information)
My advisor and I are working on a project about sulfate aerosol predictability and its impact on climate variability. In our base-case, we've implemented a large sulfate aerosol plume on the east coast of the United States by changing an input sulfate emissions file.

My advisor then thought it would be a good idea to create cloned cases branching off from our aerosol blob case and run each of them for 3 months to compare and average to account for internal variability.

We did that, but it looks to me that our cloned cases are all identical to each other and that this feature is intended to do exactly that. To make an identical number of copies of a reference case that can then be altered.

(Questions)
Is this an accurate understanding of the create_clone feature? Is there any sort of inherent stochasticity within our model (F2000 compset) or will internal variability arise from us manually inputting a variety of initial conditiions and/or parameters? Are there any scripts or common ways to do something like this?

If you are able to answer any of these questions it would be greatly appreciated!

Thanks for your time and effort helping new users like me!

Jack

 

[mlevy]

CESM is configured so that running the same case twice (on the same machine, with the same compiler and PE layout) provides identical results. Did you make any changes between the different branch cases?

If you want to create an ensemble, I suggest reading through Kay et al. (2015), which describes the first CESM Large Ensemble project. They note

Ensemble members 2–30 were all started on 1 January 1920 using slightly different initial condi- tions (Table 1). Spread in ensemble members 3–30 was generated by round-off level differences in their initial air temperature fields. Specifically, we applied random round-off level (order of 10^–14 K) differences to the air temperature field of ensemble member 1 to generate atmospheric initial conditions for ensemble members 3–30. With the exception of their initial air temperature field, ensemble members 3–30 all had the same initial conditions. For technical reasons, ensem- ble member 2 was started using a 1-day lagged ocean initial condition.

To apply random random-off level noise to the air temperature initial condition, use the pertlim variable in the CAM namelist; you change that variable by adding a line pertlim = {value} to user_nl_cam. I believe the CESM Large Ensemble used values like

pertlim = 1e-14
pertlim = -1e-14
pertlim = 2e-14
pertlim = -2e-14

Also, I'm going to move this to the CAM forum since it's a question about setting up a CAM experiment.

 

[jzweifel]

Thank you, I read through your response and the linked paper - very helpful!