Is it possible to run CESM2 with a 1 degree atmosphere and a 0.25 degree ocean ?

[m_kliphuis@uu_nl]

Dear all,

For a project in which I use CESM 2.1.3 I tried to set up a Pre-Industrial case with a 1 degree atmosphere and a 0.25 degree ocean.

When I look for the available grids with command:

./query_config --grids

the model claims that there is such a resolution given by the alias f09_n0253

However when I create a new case with this resolution by typing:

./create_newcase --run-unsupported --case b.e21.B1850.f09_n0253.001 --compset B1850 --res f09_n0253 --output-root /projects/0/prace_imau/prace_2013081679/cesm2_1_3/

everything is indeed set up but when I build the case I run into all sorts of errors.

The first error is in:

../cesm/cesm2_1_3/my_cesm_sandbox/components/pop/bld/generate_pop_decomp.pl

Here the model tries to determine the blocksizes POP_BLCKX en POP_BLCKY based on information about the grid in the belonging .xml file

../cesm/cesm2_1_3/my_cesm_sandbox/components/pop/bld/generate_pop_decomp.xml

but this xml file only contains information for grids gx3v7, gx1v6, gx1v7, tx1v1, tx0.1v2 and tx0.1v3 and not for tn0.25v3

I tried to add the grid myself by adding:

<size res="tn0.25v3">
<nlons>1440</nlons>
<nlats>1200</nlats>
</size>

and later

<!-- tn0.25v3 resolution -->

<decomp nproc="4096" res="tn0.25v3" >
<maxblocks >1</maxblocks>
<bsize_x >16</bsize_x>
<bsize_y >16</bsize_y>
<decomptype>spacecurve</decomptype>
</decomp>

but then I run into problems later since the tn0.25v3 values are also missing in: ../cesm/cesm2_1_3/my_cesm_sandbox/components/pop/bld/namelist_files/namelist_defaults_pop.xml

containing all sorts of values like horiz_grid_file, topography_file etc.

I tried to find such files via for instance:

svn list - Revision 69156: /trunk/inputdata/ocn/pop

but again the 0.25 degree resolution is not there, I only see directories with files for the 3 degree, 1 degree and 0.1 degree resolutions.

I could not find any information about where to find the files either on this forum or elsewhere on the internet.

Basically my question is: Is it possible to run CESM2 with a 1 degree atmosphere and a 0.25 degree ocean ?

Kind regards,
Michael

 

[dbailey]

This is possible, but more complicated. You would have to create a 0.25 ocn / ice grid. Then create mapping files from f09 to the 0.25 grid making sure to account for the land / ocean mask to be updated with the 0.25 coastline. Then there is the question of how to initialize the ocean and ice models. POP does not support this grid. It might have been a special HYCOM grid.

 

[dbailey]

Moving this to the POP forum just to make sure.

 

[mlevy]

I'm less optimistic about getting this to work than @dbailey -- tn0253 is not a POP grid, and out-of-the-box we only support CESM at the 3 degree, 1 degree, and 0.1 degree resolutions. I've looked through the POP documentation, and don't see anything about the process for creating a new grid for the model.

For CESM3, the ocean component will be MOM instead of POP, and we are working on a 0.25 degree grid but I don't think it will make the CESM 3.0 release that is tentatively scheduled for early next year... it'll probably be a number of months into the CESM 3.1 development cycle before a quarter-degree grid is available.

 

[m_kliphuis@uu_nl]

Dear David and Michael,

Thank you for your prompt response. This explains why I wasn’t able to set it up properly.

It’s good to hear that there might be a 0.25-degree ocean (MOM) resolution available in CESM3 in the future. While there may be a reason for it, I think it would be better to ensure that non-existent grids are not included in the output of ./query_config --grids. :)

Thanks again!

Kind regards,
Michael