aquifer elevation wrt MSL for groundwater modelling

Aquifer top elevation can be easily estimated based on the elevation of the top ground surface but how to estimate bottom elevation of the aquifer. It is not possible to physically determine the bottom elevation everywhere in the area under study. So it has to be estimated. But now the question arises whether the bottom elevation will follow the top elevation (i.e. top elevation minus fixed aquifer thickness) or the bottom of the aquifer will be flat (i.e. the variable aquifer thickness). How to deal with this problem. Can anybody help?

In the typical case, the bottom elevation is estimated based on borehole logs showing the stratigraphy. From this, the hydrogeological system can be conceputalized, and the model can follow this concept. There will be no general rule whether the bottom elevation follows the shape of the top or not. This highly depends on the geological setting you're working in.