[MITgcm-support] 2D domain and icefront

[Menemenlis, Dimitris (3248)]

Carl, here is answers, as best I can.
I copy your questions to Yun Xu, who wrote the icefront package,
and to MITgcm-support list, in case I make a mistake and someone can correct me.

Cheers

Dimitris Menemenlis

On Feb 9, 2012, at 6:26 AM, Carl Gladish wrote:

> OK, one question is:  when you use a domain of size (nx,ny,nz)  where
> ny=1, does the model automatically ensure that the equations being
> solved are the 2-d boussinesq equations?

If you have ny=1, the model assumes infinitely wide in the y-direction.
this is because the model is doubly periodic in x and y directions.
So yes this is equivalent to solving 2-d equations, I think.
Whether or not they are boussinesq depends on your model configuration,
that is, whether you use z or p coordinates.

> Also, to make an icefront appear in my xz 2-dimensional domain, I did
> the following:
> 
> Ocean bathymetry d(x) is < 0 for all x except for x=end where I made
> d(end)=0.   The icefront length is 0 for all x except for x=end-1,
> where length=1.  The icefront depth is 0 everywhere except at x=end-1
> where icefront depth is the same as d(end-1).
> 
> It seems to work, but I thought I'd double check that this is reasonable.

This sounds correct, but please double check with Yun Xu also.
I've been meaning to check in a small verification experiment for icefront
but keep getting distracted by other things.