Running a "F_4xCO2" simulation similar to the predefined "F1850" compset

[mneedham]

Hello,

I am attempting to run a simulation forced by 4xCO2 SSTs/Sea Ice/GHG. This would be very similar to the basic F1850 compset, but I don't believe there is a pre-defined F_4XCO2 compset, so I would probably need to build it up from scratch.

My understanding is that I would need to specify only three things:

• SSTs
• Sea ice area
• CO2 mixing ratio

Is this correct? And if so, I would appreciate a little advice on the first two items:

• Is there a way to quickly generate the SSTICE data file from existing restart files or publicly available input data?
• Or should I download SSTs and sea ice area from the cmip6 archive to make those files?

Lastly, for CO2, I assume I can just specify the value of co2vmr in the namelist, and set scenario_ghg to "FIXED"

Note: I am running the isotope-enabled version of the model (iCESM, hosted on github), but don't think that should make too much of a difference with setting up the forcings.

 

[mneedham]

This page of the CIME documentation about DOCN has partially answered my question (specifically, section 9.7), but it is still unclear the best way to get the SST and Sea Ice data in the first place.

 

[hannay]

Hi Michael,

I want to make sure I understand what you are planning to do.

I understand that you are changing the CO2 -> 4xCO2 in a F1850 compset.
Are you also trying to change the SSTs/sea-ice with your name file (say my_sst_ice.nc)?

In that case, you don't necessarily need a new compset (except if you want to create one).

You create create your case with a F1850 with something likke

./create_newcase --compset F1850 --res $your_res --case $your_4xco2_run
cd your_4xco2_run

Then you can edit user_nl_cam and set the co2 value. For instance:

co2vmr = 1138.8e-6

And you can udpate the SST/ICE file with:

./xmlchange SSTICE_DATA_FILENAME=my_sst_ice.nc

 

[mneedham]

Hi Cecile,

Thank you for your response. You are correct, it is not necessary that I create a new compset per se.

Instead, I am more asking about the "best practice" way to generate that sst/ice file - especially in a systematic way if I need to do this several times with SST and sea ice fields from different cmip6 runs.

Would it be most efficient to use the restart files that are currently in the UCAR campagin storage:

/glade/campaign/collections/cmip/CMIP6/restarts

to generate the my_sst_ice.nc file as mentioned above? If so, is there a CESM tool/script that could do this automatically?

 

[hannay]

Ok. I understand what you are trying to do now.
There is no official scripts to create SSTs from a CESM runs.

I have a set of scripts in ncl that I used to create SSTs from a coupled run. I have not used these for a while but I am happy to share them if it can help.
Disclaimer: These are my own scripts for my personal use. I didn't write them to be distributed, so you may need to adapt them.

step0_concat_file.txt => concatenate the TEMP and aice (it depends the format your output is)
step1_create_timeseries.ncl => create timeseries in the format expected by the sstice tools + do some consistency checking between sst and sea-ice
step2_regrid_bcgen.txt => build CAM-readable SST and sea ice concentration boundary datasets using the icesst tools
step3_add_metadata.ncl => add some metadata to the file

The icesst tool is part of the cesm distribution: iCESM1.2/models/atm/cam/tools at master · NCAR/iCESM1.2

 

[mneedham]

Thank you for these resources - I can see that I will need to make a few adaptations for my specific situation (e.g., different netcdf variable names between the CMIP and CESM defaults), but this should help me hack together a working SSTICE file.

 

[ohmpawat]

Hi, Cecile,

I'd like to download the files that you mention in your scripts. Where can I to get them? Thanks!

Best,
ohmpwat

Ok. I understand what you are trying to do now.
There is no official scripts to create SSTs from a CESM runs.

I have a set of scripts in ncl that I used to create SSTs from a coupled run. I have not used these for a while but I am happy to share them if it can help.
Disclaimer: These are my own scripts for my personal use. I didn't write them to be distributed, so you may need to adapt them.


The icesst tool is part of the cesm distribution: iCESM1.2/models/atm/cam/tools at master · NCAR/iCESM1.2