[MITgcm-support] very warm climates and aim_v23

Rose, Brian brose at albany.edu
Thu May 4 09:08:28 EDT 2017

Hi Alexandre,
We have used aim_v23 to simulate warm climates before, see for example Rose and Ferreira (2013) J. Clim.<http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-11-00547.1> or Rose, Ferreira and Marshall (2013) J. Clim.<http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-12-00175.1>
So there is no particular limitation of aim_v23 in this regard. It looks like you are getting a numerical instability that often occurs at the cube corners on the CS grid. Reducing the timestep should help this occur less frequently. But I often got around this problem using a trick I learned from David Ferreira:

Restart from the last pickup file before the crash, and change one parameter in data.shap (Shapiro filter), run the model for a short time, then reset the parameter to its normal value. This should just set the storms on a slightly different track, which is often all it takes to avoid the instability.

- Brian


  Brian E. J. Rose
     Assistant Professor
     Atmospheric & Environmental Sciences
     University at Albany

     Earth Sciences 315
     (518) 442-4477

On May 3, 2017, at 11:48 AM, Alexandre Pohl <pohl at cerege.fr<mailto:pohl at cerege.fr>> wrote:

Hi all,

I am using the coupled model with aim_v23 to simulate the warm climate of the Cretaceous. Under high atmospheric forcing levels, the model systematically stops with the following error message:
(PID.TID 0021.0001) *** ERROR *** AIM_DYN2AIM: Temp= 9.8192364560861E+01 out of range  100 400
(PID.TID 0021.0001) *** ERROR *** AIM_DYN2AIM: Pb in i,j,k,bi,bj,myThid,I2,X,Y=   1  30   5  1  1  1   233 -169.687   -0.930
i.e., temperature drops below the minimum temperature defined for the LW radiation scheme of aim_v23.

I tried to reduce the time step but this does not seem to solve the issue. Does anyone know how to overcome this ?

Is it a limitation of aim_v23 ?

Thanks !


Alexandre POHL
Postdoctoral Researcher

MITgcm-support mailing list
MITgcm-support at mitgcm.org<mailto:MITgcm-support at mitgcm.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20170504/36353049/attachment-0001.htm>

More information about the MITgcm-support mailing list