The main difference between a Type 1 and Type 3 boundary is this:
[list]
[li]At a Type 1 boundary, head is specified for those nodes and the modelled heads will comply with specified values exactly. Constant or time varying head values can be specified. A Type 1 boundary could also be called a constant or variable specified head boundary. This boundary type is typically used where a strong hydraulic control on groundwater levels exists, such as from the sea or a large lake or river.[/li]
[li]At a Type 3 boundary, a target head is specified along with a conductance term which is used to calculate flows to/from the boundary. Modelled heads at the boundary will not necessarily comply with the BC head values. Note also the conductance terms in FEFLOW are set as material properties to elements. Note that a type 3 boundary must be defined across element faces in order to function. Type 3 boundaries are commonly used to represent streams or small rivers which have weaker hydraulic controls on groundwater and have a limited recharge or discharge capacity (e.g. due to channel dimensions, bed sediment characteristics, water availability). I suggest referring to the FEFLOW material for a proper description of this BC and its maths.[/li]
[/list]
A general rule of thumb is that Type 1 (specified head) should be used to represent the coastline.
For Q2, could you just limit your model domain to exclude the mountainous area?