The CESM2.1.4 B2000 spin-up time

[tyh-iap]

Hello everyone, I'm using the CESM2.1.4 to run B2000 compset with the grid of f19_g17. I modified 1850_CAM60_CLM50%BGC-CROP_CICE_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD to 2000_CAM60_CLM50%BGC-CROP_CICE_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD. The CO2 of this compset is correct in 2000. Now it runs about 120years, and the energy budget is still unbalance. It's so strange. I only have a few HPC resource, Could anyone tell me the spin-up time of B2000 in CESM2.1.4?

 

[tyh-iap]

It seems to run hundreds of years to balance.

 

[slevis]

I don't have an answer to the "spin-up time" question, but I have a thought:
If you already had spun up initial conditions for 1850, you may be better off running with BHist from 1850 to 2000 to generate the B2000 initial condition files.

 

[tyh-iap]

I don't have an answer to the "spin-up time" question, but I have a thought:
If you already had spun up initial conditions for 1850, you may be better off running with BHist from 1850 to 2000 to generate the B2000 initial condition files.

Thank you very much Slevis. I have the B2000 restart files run by CESM2.1.3 in another HPC. However when I use the CESM2.1.4 to start a branch run, there alway has an error like this.. Thus I have to run a cold start in CESM2.1.4.

 

[slevis]

So you're spinning up a BgcCrop case from a cold start. I have two questions:
1) Are you using the AD spin-up method? If not, you should read this section of the CLM User's Guide:

1.5.5. Spinup of CLM5.0-BGC-Crop — ctsm release-clm5.0 documentation

2) You will likely save computer resources if you first spin up the standalone land model. Then you will likely still need to spin up the coupled model, but for less time.