MPAS Init crashed generating Static file with New mesh

[Akinleye]

I encountered problem running init_atmosphere. I have attached my namelist.init_atmosphere, most recent log.init_atmosphere.0000.out, and the python script used to create the mesh file. You can see the created mesh file in my path on derecho (/glade/derecho/scratch/afolorunsho/mpas_practice/test_300m_tex). Though, I attempted much coarser resolution, I encountered the same problem with my created mesh. I did check also the mesh_triangles.nc but no NaN data was found.

The printed cycle in the log file was added in the mpas_init_atm_static.F file to trace where the model crashes, most likely a bad point in the created mesh.

Has anyone successfully created a new MPAS mesh using the jigsaw tool?

 

[Ming Chen]

@Akinleye,​

Would you please download the latest mpas-tools and re-create your mesh? Specifically, a bug in the computation of angleEdge was fixed. I am not sure whether this bug is related to your problem. Anyway, please try and let me know whether the new version of mpas-tools works for you.
Thank you.

 

[Akinleye]

Thank you for the response. The problem has been fixed. Apparently the sphere radius generated from the JIGSAW was slightly different from the static files creating a problem with the kdtree. This is solved applying a scale factor in the mpas_init_atm_static.F.

 

[Ming Chen]

Thank you for the update. Would you please clarify how you apply the scale factor in mpas_init_atm_static.F ? Did you still run the code in files.zip to create your mesh? Thanks !