phreatic surface crossing layers in steep topography

Hi,
I am modelling a layered heterogeneous system in steep topography and thus the water table crosses multiple slices. 
The top slice is phreatic and underlying slices are unspecified. 
Is there a way to determine the resulting water table surface elevation?  (i.e. as it crosses the layers?).  Does the head in the top slice represent the water table surface?  Or does it represent the residual water table in that slice where it has gone "almost dry" and the water table is actually the head at the underlying slice?
Laurie

I'd typically do an isosurface for a pressure of 0 (or close to 0) in a 3D view to determine the water table location visually. The hydraulic head in slice 1 approximately is the water table elevation, but not exactly. In case that groundwater recharge is applied, this has to be passed down to the water level, which will result in slightly higher hydraulic head in slice 1 than at the water table.

Thanks Peter.
Another question.  I am using phreatic for the top slice and unspecified below.  The results look fine (I have flux out of the top slice where the seepage face should be, my velocity vectors look good and the head contours and water table are reasonable).  My problem is that the budget analyzer indicates I have too much water going in. 
I have set the specific options to unconfined, and the constraints are set to unconstrained as sseepage face if falling dry at bottom or reaching the op surface.  I have tried decreasing the tolerance and refining the grid but I still get the same imbalance.  Is this because my water table crosses many slices? and the slices are not allowed to go completely dry? (I am running saturated mode).
Can I determine how much water is flowing through the system by subtracting the imbalance from the inflow?
Any ideas?