changing 'default' sea ice thickness using ice_ic='default'

[s_l_bradley@sheffield_ac_uk]

Hello

I am running the following compset and wish to initial the sea ice thickness in the NH and SH to be values different from what is set by ice_ic=default

1850_CAM60_CLM50%BGC_CICE_POP2_RTM_CISM2%NOEVOLVE_SWAV

With the latest version of cesm2.1.1
My case is for the last glacial maximum; I am running at fv0.9x1.25 and gx1v7 for the ocean and sea ice.

My understanding is that with this user_nl it sets sea ice thickness = 2m in NH and 1m NH where the SST below freezing. However I want to change this to a smaller values, 0.5m

I think this is calculated within the code - but I am unsure how to edit this,
~/components/cice/src/source/ice_init.F90

Can you suggest how I can do this.
Thanks

Sarah Bradley

 

[dbailey]

You should be able to change ice_ic to a filename and make sure this is in your local directory. In a fully-coupled run, ice_ic should not have 'default' as the default value. :) If you are doing a hybrid start, it will use the RUN_REFCASE in env_run.xml.

 

[s_l_bradley@sheffield_ac_uk]

Hi David
We are doing a hybrid run so combining a series of restart files from different earlier runs for CAM/POP etc.. The restart file we have for the CICE model was created on a different land/sea mask (kmt) file so will not run. It fails with Picard acceleration failure: suggesting we have Nans. I have tried a few different CAM/POP restart file, incase there was issues with the forcing - but they all fail.

Therefore, we are wanting to restart the cice and need to decide an ice_ic to use. Rather than use a default of 2m or 1 m, we wanted to go for a smaller value.
Is this possible?

Sara

 

[tcraig]

@s_l_bradley@sheffield_ac_uk, have you sorted this out? I believe the default initial condition is defined in subroutine set_state_var in ice_init.F90. I believe you should be able to modify the initial state there. It appears hinit and ainit will define the initial state and icells, indxi, and indxj the location. If you have managed to make something work, could you summarize your approach for others. Thanks.