mask_psi changing between input and output

[swarner]

My model is crashing after about only 20 time steps and I'm in the process of debugging. In the output netCDF files, a number of points in mask_psi have been changed from 0 to 2. The other masks (mask_rho, mask_u and mask_v) all seem to be unaltered from the way I set them to be in my input netCDF files. Any suggestions or thoughts as to how I might fix this?

thank you,
Sally

[kate]

Are you sure this is causing you grief? It is being done quite intentionally in metrics.F:

!-----------------------------------------------------------------------
! Set slipperiness (no-slip) mask at PSI-points.
!-----------------------------------------------------------------------
!
! Set no-slip boundary conditions on land-mask boundaries regardless of
! supplied value of gamma2.

[swarner]

Hi Kate,

Thanks for your response. I thought that my problem might be due to the mask because I have some very anomalous temperatures and salinities (negative and/or order 10^3 magnitude) in grid boxes near the mask. But no, I am definitely not sure that the mask is the source of my errors. I'll keep debugging and look for some other reason that it's crashing.

thanks,
Sally