some questions on the spin-up

[jinmuluo]

What version of the code are you using?
ctsm5.3.065
cesm3_0_beta06 (ctsm5.3.029)

Have you made any changes to files in the source tree?
Some changes in the nitrogen state pools and fluxes.

Describe your problem or question:

Hi,

I have some questions on the model spin-up for different mode versions. Hope someone can give me some advice.
I am using the CESM3-beta6 (ç) and CTSM5.3.065. I want to SPIN-UP the CTSM5.3.065 for a long time and use the restart file to initialize the CESM3-beta06; The CTSM in CESM3-beta6 is slightly earlier than my CTSM tag; can I initialize my CESM3-beta06 (CTSM5.3.029) with the restart from the CTSM5.3.065?

The second question is in my CTSM5.3.029; it needs some more different pools, which my CTSM5.3.065 restart file doesn't have it. Do you think it will cause the model to crash?

Thanks

 

[slevis]

@jinmuluo if I understand the question, you want to spin up ctsm5.3.065 for a long time and use that to initialize cesm3_beta06 (ctsm5.3.029), and you wonder whether the missing pools from 065 will cause 029 to crash.

I think it depends:
You will start the cesm simulation in "startup" or "hybrid" mode and you will set use_init_interp = .true.; the CLM finidat file will not contain all the necessary pools, but where you added the code in 029 for the new pools to appear in restart, you could set interpinic_flag='skip' to avoid a crash. Look for existing examples in the code where we use interpinic_flag='skip' to see how you may use it.

I'm not seeing a second question.

 

[jinmuluo]

@jinmuluo if I understand the question, you want to spin up ctsm5.3.065 for a long time and use that to initialize cesm3_beta06 (ctsm5.3.029), and you wonder whether the missing pools from 065 will cause 029 to crash.

I think it depends:
You will start the cesm simulation in "startup" or "hybrid" mode and you will set use_init_interp = .true.; the CLM finidat file will not contain all the necessary pools, but where you added the code in 029 for the new pools to appear in restart, you could set interpinic_flag='skip' to avoid a crash. Look for existing examples in the code where we use interpinic_flag='skip' to see how you may use it.

I'm not seeing a second question.

Thank you, Sam, I will try this flag later.

best,

 

[jinmuluo]

@jinmuluo if I understand the question, you want to spin up ctsm5.3.065 for a long time and use that to initialize cesm3_beta06 (ctsm5.3.029), and you wonder whether the missing pools from 065 will cause 029 to crash.

I think it depends:
You will start the cesm simulation in "startup" or "hybrid" mode and you will set use_init_interp = .true.; the CLM finidat file will not contain all the necessary pools, but where you added the code in 029 for the new pools to appear in restart, you could set interpinic_flag='skip' to avoid a crash. Look for existing examples in the code where we use interpinic_flag='skip' to see how you may use it.

I'm not seeing a second question.

Hi Sam,

Just one more question, can I use 2-degree results after spin-up to initialize a 1-degree CESM simulation?

 

[slevis]

Yes, the use_init_interp = .true. setting allows that. If you go from f19 to f09 it should work correctly with nearest-neighbor interpolation. Just keep in mind that recently we found that the interpolation may not work correctly from a regular grid (e.g. f09) to an unstructured grid (e.g. ne30).