[MITgcm-support] Solution blows up

[Shevchenko, Igor]

Dear MITgcm users,

I have two solutions computed at 1/12 and 1/6 degree resolution, respectively. MITgcm works perfectly fine for both. 
Now, I would like to run the 1/6 model starting from an initial condition which is computed as an interpolation of a 1/12 snapshot onto the 1/6 grid.
However, the initial condition computed in such a way leads to the solution which blows up after several time steps. I tried different time steps but it didn't change the result.
Any hint of what can potentially go wrong is very welcome. 

Best regards,
Igor