Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Questions on CLM Spinup

Zhu Zhu

Zhu Zhu
New Member
Hi everyone,

I’m trying to run CESM2.1.5 with the following component set:
2000_CAM60_CLM50%BGCDV-CROP_CICE_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD
However, there is no default land initial file for this configuration. To generate one, I first ran the land model using the I2000Clm50BgcDvCrop component set, driven by a 10-year atmospheric forcing from fully coupled run. Following the user guide’s instructions for accelerated spinup (1.5.5. Spinup of CLM5.0-BGC-Crop — ctsm release-clm5.0 documentation), I turned on the accelerated spinup mode. I ran it for 300 years and it got equilibrium:
1749997320574.png
Then I saved the last restart file from this run and use it as the “finidat” file for another spinup run in normal mode, with same component set and atmospheric forcing files. The outputs are as follows:
1749997994799.png
In the second spinup, I see a pronounced ~10-year oscillation in the time series—something that did not appear during the accelerated run. I’m wondering whether this is an artifact of my atmospheric forcing, or if I’ve made an error in the spinup procedure.
The guide mentions that the second spinup should run for 400+ years. Should I therefore extend this second run to 400 years or more to damp out these oscillations? Any advice or suggestions would be greatly appreciated. Thank you!


Best regards,
Zhu Zhu
 

slevis

Moderator
Staff member
Since you are running the land model with datm (i.e. I-case simulations), then I suspect that you have been cycling 10 years of the datm meteorological data. If so, I would also guess that you could zoom in on the 300-year output and find the same oscillation. And if so, then the oscillation is an artifact of the x and y scales of your plots. But if I'm wrong about that, then I do not know why you didn't get the oscillation in the 300-yr simulation.

On a different note, I am interested to see that you have successfully run with Clm50BgcDv, suggesting that the original CLM-DGVM works in this version of the model. This is encouraging, but keep in mind that we do not support this capability.
 

Zhu Zhu

Zhu Zhu
New Member
Since you are running the land model with datm (i.e. I-case simulations), then I suspect that you have been cycling 10 years of the datm meteorological data. If so, I would also guess that you could zoom in on the 300-year output and find the same oscillation. And if so, then the (1:59) is an artifact of the x and y scales of your plots. But if I'm wrong about that, then I do not know why you didn't get the oscillation in the 300-yr simulation.

On a different note, I am interested to see that you have successfully run with Clm50BgcDv, suggesting that the original CLM-DGVM works in this version of the model. This is encouraging, but keep in mind that we do not support this capability.
slevis, thanks for your response! Both simulations have10-year oscillation. The post-processing script treat these two simulations output differently. For the first simulation, the script reads data every 20 years, so I didn’t see the 10-year oscillation before. Should I read data every 20 years for second simulation to assess whether these varibles have reached equilibrium?

In addition, Is it normal to see such strong interannual variability? The CLM-DGVM is not supported in CESM2.1.5. Does this mean that using Clm50BgcDv could produce incorrect results? I want to run fully coupled CESM2.1.5 with dynamic vegetation turned on. Should I instead use CLM50FATES? I am not sure whether CLM50FATES can be used in fully coupled run.
 

oleson

Keith Oleson
CSEG and Liaisons
Staff member
Generally, assess spinup according the number of years that you are looping over for your atmospheric forcing.
CLM50BgcDv could produce incorrect results since it's not scientifically vetted/supported. So you will have to judge whether it's results are reasonable for your project.
We suggest reading through the FATES documentation to see if it would be an acceptable alternative. I don't think CLM50FATES is scientifically supported in coupled mode.
There is a chapter on FATES in the CLM5 technical note:


The FATES github repository is:


There is a FATES Discussions Board there where you could ask your question about coupled runs.
 

Zhu Zhu

Zhu Zhu
New Member
Generally, assess spinup according the number of years that you are looping over for your atmospheric forcing.
CLM50BgcDv could produce incorrect results since it's not scientifically vetted/supported. So you will have to judge whether it's results are reasonable for your project.
We suggest reading through the FATES documentation to see if it would be an acceptable alternative. I don't think CLM50FATES is scientifically supported in coupled mode.
There is a chapter on FATES in the CLM5 technical note:


The FATES github repository is:


There is a FATES Discussions Board there where you could ask your question about coupled runs.
Thanks for the response! I'll read the FATES documentation.
 

oleson

Keith Oleson
CSEG and Liaisons
Staff member
This post addresses FATES in coupled mode:

 

Zhu Zhu

Zhu Zhu
New Member
This post addresses FATES in coupled mode:

Thanks for the reminder. This is very helpful! It seems that I can only run CLM with FATES offline using the outputs from the coupled simulation. I'll try this way.
 
Top