Clarification on ghost cells and boundary conditions in bc.dat

Question

Having read through the AEM3D manual and the forums, there are still a few gaps in my understanding on how boundary conditions are set in AEM3D.  Could you please clarify my questions below?

A. If ghost cells can't be wet, then, are they only placed within the domain containing land values (e.g., with 999 values)?

B. If (A) is true, then how would it be possible to specify the face of a cell in bc.dat for an outflow in the middle of a lake, where there are no land cells, and presumably, where no ghost cells can be placed?

C. If xp, xm, etc. are only specifying the cell faces where flow can enter/leave and does NOT have an impact on the direction of flow, then is there a difference if I run a simulation with a particular cell face set to "flow_xp" versus running the same simulation with the same cell face set to "flow_xm" in bc.dat?

Apologies if these questions seem silly and appreciate your time.

Answer 1

A.  No, they are also located under wet cells.  

B. So when there is a drop in the bathymetry across z-layers there will be a ghost-wet boundary.

C. The bc.dat must always specify the wet cell.  If you change the face then the water will enter/leave from a different side of the wet cell (assuming this face is also a ghost-wet cell boundary).