How to start a spin-up run in CAM6 and correct some outputs?

[nuvolet]

What version of the code are you using?
CESM2.2.2

Have you made any changes to files in the source tree?
Some modifications in the nudging module

Describe every step you took leading up to the problem:
I ran some tests using new re-analisys data and modifying aerosols and after running a month, there are some unusual values for sea salt mass concentrations over areas (North of Africa) that supposedly won't be present (figure enclosed). The first idea is to do a spin-up run to correct those outputs and then run again the model over that first run and correct those anomalies.


Describe your problem or question:
How can I run a spin-up run using the FCnudged compset?
Is there any documentation or example to do that?

Thanks in advance

 

[rrbuchholz]

I'm just wondering what you find unusual about the sea salt values north of Africa? Do you have a difference plot to observations to show how the values should be different?

You can run a spin up in a couple of ways:
1. You can run the model for a year before your desired start date. For example if you are interested in dates onwards from Jan 2024, you would start the simulation in Jan 2023.
2. You can spin up by running for a year and then re-initialize with the spun-up initialization file.

 

[nuvolet]

I do not have plots of difference to observations.
Regarding the spin-up, it is possible to run a 2024 and re-initialize again the same run using the previous 2024 run as a initialization file?
Thanks

 

[rrbuchholz]

OK, then I'm wondering how are you diagnosing that the sea salt patterns are unusual north of Africa? To me it looks perhaps like the winds are transporting sea salt over the continent?

Yes, your spin-up question is the second option if you can't do the first option.

 

[nuvolet]

I forgot to mention that what it is plotted is the sea salt mass concentration for the mode 3 (coarse) at the lowest level in the hybrid level point.
The plot is the monthly average in January 2020, and initially it look suspicious that over the whole month the sea salt in that mode could be that large in that area.

Definitely, I will give a try to run a spin-up case to see if those values for seas salt can be softened.

 

[nuvolet]

Hi Rebecca,

Do you think that this source on how to run spin-up model:

https://www2.cesm.ucar.edu/models/cesm1.2/clm/models/lnd/clm/doc/UsersGuide/x12680.html

could it be use same way for the current CESM2.2.2 and CAM6?.

Thanks in advance

Toni

 

[rrbuchholz]

Hi Toni,

There is a difference between spinning up for climate-scale-runs (i.e. using the fully-coupled model system) and spinning up for nudged atmospheric chemistry runs. For climate runs you notice in that link they say "spin up for a short time of 20 to 50 years". This is in order to get a stable model climate. A 20 year run with full chemistry is quite a long simulation. For atmospheric chemistry purposes, we can assume the model has a stable climate, and we only need to spin up the chemistry for about a year. Sea Surface Temperatures, for example, are data driven - i.e. in env_run.xml: SSTICE_DATA_FILENAME is defined.

I hope that helps clarify spin-up.

Best,
Rebecca