CLM1PT run bgc -cn compset spinup problem

[renqiujie]

Dear all,
I use compsets 2000_DATM%GSWP3v1_CLM50%CN_SICE_SOCN_MOSART_SGLC_SWAV run a single point simulation，atm forcing is GSWP3v1 (1901-2014).
My Settings are as follows:
1 spinup_state=on 2000 years
2 spinup_state=off 400 years
use_fun= .true.
use_nitrif_denitrif=.true.
My resoult of NPP (gc/m^2/s) , I doubt whether the result is correct at present. Could someone please help me? Why does NPP have a jump when the spinup_state changes?

Thank you all

 

[oleson]

Were there any changes made to your case between the spinup and the non-spinup segments of your run? Or did you just change the "spinup_state" namelist item? If you used two separate cases, perhaps compare the two cases. In addition to looking at lnd_in, you might look at datm_in to make sure you are looping over the same datm forcing years in the two instances.

 

[oleson]

I should say that in standard CLM5 BGC cases, we do see a brief increase in LAI/GPP/NPP when transitioning between spinup and normal mode in single point runs, but that should go away after a few years and settle back down to values that are close to those obtained during the spinup. Yours is reaching a different/new equilibrium. You are using CLM5 CN, with a couple of namelist modifications. Not sure if somehow that is causing the equilibrium difference.

 

[renqiujie]

Dear Oleson，
Thank you for your reply.
1.
I want to study the state of carbon and nitrogen cycle at a certain point, so I use CLM50%CN . From the description file of CLM5, I know that CLM5 has a new FUN model, which is not opened by default, so I use "use_fun= .true." and "use_nitrif_denitrif=.true." . Dose this setup correct?

2.
I make sure that the datm in file is consistent. And I have done two kinds of tests, one is the process of spinup and non-spinup in the same case, and the other is the stage of spinup and non-spinup in the two cases respectively, both of which appear to reach a new balance. Do you have any Suggestions?

 

[oleson]

The new carbon/nitrogen cycle in CLM5 is called "BGC", not "CN". CN is the old one. In BGC, FUN is on by default. So you might want to change your setup and repeat your spinups and see how they look. The equivalent compset you would want would be:

2000_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV

or if you want to include crops:

2000_DATM%GSWP3v1_CLM50%BGC-CROP_SICE_SOCN_MOSART_SGLC_SWAV (I2000Clm50BgcCrop)

 

[renqiujie]

OK， Thanks, I'll try your suggestion .
One more question. I want to simulate the NPP changes from 2000 to 2014. Can I use 2000_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV component to start the simulation in 2000 years ?
Or I need to use HIST_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV to start the simulation in 1850 years？

 

[oleson]

If you want to include the transient behavior of the model (e.g., land cover change, changes in nitrogen deposition, etc.) during 2000 to 2014, then you would want to use the historical compset (HIST).

 

[renqiujie]

Dear Oleson,Thanks again for your advice .
I have done the spin and non-spinup processes use 2000_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV wiht atm forcing (1901-2014)，
I built a case with HIST_DATM%GSWP3v1_CLM50%BGC_SICE_SOCN_MOSART_SGLC_SWAV, and the base is set to
do_harvest = .false.
do_transient_crops = .false.
do_transient_pfts = .false.
flanduse_timeseries='landuse.timeseries_1x1_BCO_hist_16pfts_Irrig_CMIP6_simyr1850-2015_c201029.nc'
fsurdat='surfdata_1x1_BCO_hist_16pfts_Irrig_CMIP6_simyr2000_c200918.nc'
I run the case from 2000 - 2014 years. Is this single point operation correct?

 

[oleson]

If you don't want transient landcover, but do want transient CO2, ndep, etc., then that should work. But I would check the lnd and atm log files to make sure things are being implemented as you intended.
Technically, you don't need to set flanduse_timeseries since it won't be used, you could set it to ' ' (blank).

 

[demibing]

may i ask whis is spin and non-spinup? which place the spin or non-spinup be used? why used?

 

[oleson]

Spinup refers to the process of getting the model to an equilibrium state (energy, water, carbon, nitrogen). The purpose is to eliminate drifts in model state that could influence or confound the model state in transient mode.
See slide 4 of the following tutorial:

https://www.cgd.ucar.edu/events/2019/ctsm/files/practical4-wieder.pdf

See also the User's Guide:

1.5.6. Spinning up the Satellite Phenology Model — ctsm CTSM master documentation

1.5.7. Spinup of CTSM1-BGC-Crop — ctsm CTSM master documentation

 

[demibing]

Spinup 是指使模型达到平衡状态（能量、水、碳、氮）的过程。目的是消除模型状态中可能影响或混淆瞬态模式下的模型状态的漂移。
请参阅以下教程的幻灯片 4：

https://www.cgd.ucar.edu/events/2019/ctsm/files/practical4-wieder.pdf

另请参阅用户指南：

https://escomp.github.io/ctsm-docs/...-Satellite-Phenology-Model-CLMSP- spinup.html

https://escomp.github.io/ctsm-docs/...-up-the-biogeochemistry-BGC-spinup.html[ /网址]

Thank you very much.