Question about the spin-up process for CESM2?

[mengqi]

Hi,

I’m currently working with CESM2 at the f19_g17 resolution, using the FHIST_BGC compset. I’m planning to start with the spin-up process and was wondering how to access an existing spin-up file.

Any insights would be greatly appreciated!

Thanks!

 

[peverley]

Hi,

I'm not an expert on this process. Does anything in this post help you? How to spin up a transient climate simulation

If not, can you elaborate on what existing spin-up file(s) you are wanting to access?

 

[slevis]

Regarding existing initial condition files for the land model:
- For many model configurations the model is set up to find default initial condition files, so you may not have to worry about looking for those files.
- In general we recommend a new spin-up because we do not guarantee that our default initial conditions are appropriate for your purposes.
- For a new BGC spin-up, you may save computer time by starting from an existing clm initial condition file, but not necessarily and your spin-up process may be cleaner if you start the clm with the cold start option.

 

[mengqi]

Regarding existing initial condition files for the land model:
- For many model configurations the model is set up to find default initial condition files, so you may not have to worry about looking for those files.
- In general we recommend a new spin-up because we do not guarantee that our default initial conditions are appropriate for your purposes.
- For a new BGC spin-up, you may save computer time by starting from an existing clm initial condition file, but not necessarily and your spin-up process may be cleaner if you start the clm with the cold start option.

Thanks, Levis! It helps!

A few additional notes: I'm familiar with running spin-up simulations in CLM5, but I have some concerns about the process in CESM2, where the atmosphere and land models are coupled. Specifically, I was wondering if the spin-up procedure in CESM2 is similar to that in CLM5. I've heard that achieving a quasi-equilibrium state in CESM2 can be challenging and may require substantial computational resources, which is why I'm looking for an existing spin-up file to work from.

Thanks!

 

[slevis]

You would certainly save computer time if you could find an initial file from the same configuration as yours (CESM2, FHIST_BGC, f19_g17). If this configuration automatically selects an initial file for you, then you could try it in case it seems reasonable. If the model does not select one automatically or the outcome is not reasonable, then you can keep asking around and/or you can start your own spin-up.