[MITgcm-devel] masks in obcs_apply_ts.F and pkg/seaice

[Dimitris Menemenlis]

> Using maskC is not as universal as we do care about the topographpy inside 
> the domain and extrapolate it across the boundary. That cannot be changed.

I do not fully understand why T/S needs to be set to zero and why maskC should
not know about it but I will take your word for it.

> The open boundary information is all in pkg/obcs, so there where you should
> take it from. To fix your problem you just need to reset the thermodynamics
> along the open boundary ( HEFF=AREA=HSNOW=0).

OK.

D.