[MITgcm-support] nudging techniques, OBCS and freezing

[Martin Losch]

On Aug 15, 2004, at 5:15 PM, Ed Hill wrote:

> On Sun, 2004-08-15 at 10:30, prosperi at wicc.weizmann.ac.il wrote:
>> 2) I am confused about something very basic:
>>    in a finite volume integration scheme u have the advantage that
>> increaing the horizontal resolution, u automatuically resolve better 
>> even
>> the topography since u are working with volume units. When I set the 
>> size
>> of the grid and the domain decomposition, I have, basically (...very
>> basically), two files to change:
>>
>> SIZE.h (nSx, nSy, nr) and in data delX = nSx*size of the grid.
>> hFacMin gives me  the actual thickness of the cell.
>> The question is:
>> is it eventually by hFacMin that I can increase the spatial 
>> resolution of
>> my domain?
>
I may misunderstand your question, but hFacMin sets the minimum allowed 
thickness of cell. This concerns the partial or looped cells. hFacMin=1 
by default (no partial cells). If for example hFacMin=0.1, then 
fraction of a cell occupied by water (instead of bottom/land) cannot 
become smaller than 10 percent of the thickness of the cell, no matter 
what the actual topography is, you can see what the model uses as 
actual topography in Depth.00?.00?.data which will differ from you 
original topography file according to hFacMin. The vertical extend 
(thickness) of the cells is set by delR/delZ/delP.

Martin