How to run MPAS init_atmosphere with constant SST

J

JiaWang

New member
This is the namelist:
config_sfc_prefix = 'SST'
config_fg_interval = 0

This is the streams:
<immutable_stream name="surface"
type="output"
filename_template="x4.535554.sfc_update.nc"
filename_interval="none"
packages="sfc_update"
output_interval="none" />

But when I run init_atmosphere to generate the sfc_update file, it prints out this endlessly:
Processing file SST:2025-12-09_00
Processing file SST:2025-12-09_00
Processing file SST:2025-12-09_00
Processing file SST:2025-12-09_00

"2025-12-09_00" is the date that SST is provided.

Did I set something wrong in the namelist or streams?
 
Hi JiaWang,

I am not sure that I understand your question completely. Are you trying to create initial condition or sfc_update for MPAS run? Can you upload your namelist.init_atmosphere and streams.init_atmopshere for me to take a look? These files may help me better understand what you intend to do.

You need to create sfc_update file only when you run MPAS with sst_update. Otherwise, you don't need to create this file.

By the way, which version of MPAS you are running? Thanks.
 
Hi Ming,

I am trying to create sfc_update for MPAS run.
I thought MPAS run always needs a surface file (including skin T and sea ice), no matter it is a fixed SST run or an updated SST run.
In a fixed SST run, where is the SST from, if sfc_update file is not provided?
Since the system is down today, I could not download the namelist and streams files today. I will upload them tomorrow.
I am using the latest version 8.4.1.

Best,
Jia
 
Hi Jia,
Please see my answers below:
JiaWang said:
Hi Ming,

I am trying to create sfc_update for MPAS run.
I thought MPAS run always needs a surface file (including skin T and sea ice), no matter it is a fixed SST run or an updated SST run.
In a fixed SST run, where is the SST from, if sfc_update file is not provided?
Click to expand...
You are right that MPAS run needs surface information (skintemp, sst, sea ice, etc.). However, these information are included in the initial condition. You only needs sfc_update when you run with sst_update = true.
JiaWang said:
Since the system is down today, I could not download the namelist and streams files today. I will upload them tomorrow.
I am using the latest version 8.4.1.
Click to expand...
That is fine. We can take a look once the system is back.
JiaWang said:
Best,
Jia
Click to expand...
 
Hi Ming,

When sst_update=true, at the initial time, where is the surface information from, the initial file or the surface file, when both are provided but from different sources (for example, initial from GFS and surface from ERA5)?

Best wishes,
Jia
 
Hi Jia,

Either sst_update is true or false, surface information is always from the initial condition at the initial time.

This has nothing to do with what data you used to create initial and sfc_update data.
 
Hi Ming,

Once I removed the step to create updated surface conditions, my MPAS run works.

As you mentioned that, surface information at the model initial time (T) is taken from the initial condition file. Then in a run with a 6-hourly updated surface file, what is the surface condition between T and T+6h, fixed as values at T or temporally interpolated?

MPAS has both sst and skintemp in the initial condition file. Does MPAS use both?

Best wishes,
Jia
 
Hi Jia,
Please see my answers below:
JiaWang said:
Hi Ming,

Once I removed the step to create updated surface conditions, my MPAS run works.
Click to expand...
Thanks for letting me know. I am glad it works for you.
JiaWang said:
As you mentioned that, surface information at the model initial time (T) is taken from the initial condition file. Then in a run with a 6-hourly updated surface file, what is the surface condition between T and T+6h, fixed as values at T or temporally interpolated?
Click to expand...
If you set sst_update = false, then SST will remain the same as in the initial condition. This is because for short-term run, for example 3-5 days, we won't expect large changes in SST. SST evolves very slowly. However, for sub-seasonal to seasonal and multi-year simulations, we nee to turn on sst update. In this case, SST will be updated frequently (at 6-h interval for your case).
JiaWang said:
MPAS has both sst and skintemp in the initial condition file. Does MPAS use both?
Click to expand...
Yes MPAS use both.
JiaWang said:
Best wishes,
Jia
Click to expand...
 
Hi Ming,

Does this mean in the case of sst_update=true, sst between 0-6h is constant and from initial condition, and sst between 6-12h is constant and from the surface file? So no temporal interpolation?

In the MPAS tutorial, the surface file contains skin temperature instead of SST. Is there a MPAS document explaining how skin temperature and SST are used?

Best wishes,
Jia
 
Hi Jia,

You are right that when sst_update is turned on, the first 6-h SST is from the initial condition, and SST will be updated at 6h with the new SST from the surface input file. There is no temporal interpolation.

Would you please tell me where in the MPAS tutorial it is stated that the surface file contains skin temperature instead of SST? Thanks.
 
Hi Jia,
Thanks for the information. The logic behind using SKINTEMP to replace SST is that, SKINTEMP and SST are identical over open water points, and they are different over sea ice area. In MPAS LSM, sea ice is treated as 'land' point, where TSK (SKINTENP) is calculated based on sea ice physics. In this case, skintemp other than sst is required.
Hope this is helpful for you. Please let me know if you have more questions.
 
You must log in or register to reply here.
Back
Top