determining inflow of water in to model

im working on 3d coastal aquifer flow model. eastern part of model boundary is constant. north and south is no flow. west is variable boundary condition.
which boundary type will be better to use in western boundary, ie, neumann or fluid transfer. only data i have is water level  near to that boundary which are inside and outside domain. anyone please help to  detemine how to calculate flux for neumann. or in case of fluid transfer condition, can i use use near by well data as reference head ? if so wat should be in/out transfer rate in material proprty?

The flux q for NeumannBC's can be approximated by q = -K*grad(h). K is the hydraulic conductivity and grad(h) the gradient of the hydraulic head. The in- and out-transfer rates can be estimated by phi = K / d. d is the resistance between the aquifer and the surface water body. As I am not deep involved in the project I find it difficult to give recommendation on which boundary condition to assign. Therefore a rather general hint. If you use a Fluid-TransferBC, the flow directions (inflow and outflow) may change depending on the dynamics of the system. In contrast, if you use NeumannBC's need to impose a priori the time-constant or time-variable in-/outflow.

It is very good information offline.

I suggest you to use Cauchy bc for variable head because the use of Neuman BC is more related to specific flux like inflow or outflow, this can be applied to rechargue zones.
Using Cauchy you can extent your model domain a little to take a portion of the sea and use it like Cauchy Bc.
However if I have to do that model I would prefer use constant head bc in the coastal because is the simplest way to do, but it depends of the extent of your model.
I hope this can help you