CESM1 B1850 spinup to steady-state

[Camille013]

I'm running the B1850 compset with CESM1 (T31_g37 grid), and was surprised to find that the initial state was very far from being equilibrated (see attached figure). Doing a little more reading, it seems that this is because the ocean component is initialized with Levitus initial conditions, and that it's possible to start it with spun-up conditions but somewhat complicated and not recommended (https://www.cesm.ucar.edu/models/cesm1.0/cesm/cesm_doc_1_0_6/x2690.html). I'm surprised as I would think many people want to run experiments starting from (at least a quasi-) steady-state, so I'm wondering if there are any easier/more accepted ways to do so? Thanks in advance.

 

[dbailey]

Are you using CESM1.0? The truth is we don't really support T31_g37 any longer. We do this for the nominal one-degree runs. It is actually not that hard to start POP in hybrid mode with a spunup initial condition. I am moving this to the POP forum.

 

[Camille013]

I'm using CESM1.2.2.1. I could certainly runs it in hybrid mode with spunup initial conditions if anyone knows where to find these for the T31_g37 grid? Thanks again

 

[dbailey]

There are initial files for T31_g37 under:

https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/ccsm4_init/

I believe these will work fine with CESM1.2.2.1.

 

[Camille013]

Thank you! Based on the name of the files corresponding to the t31_g37 grid (i.e. 'b40.t31x3.037c.pop.r.0507-01-01-00000') should I assume these are spunup conditions from a 507-year-long run?