Suppose I am given a 3-manifold as a double branched cover over a link. Let a null-homologus knot in this space be given as a lift of an arc with endpoints on the link (it is automatically null-homologous if the double branched cover is a homology sphere, i.e. the link has determinant 1). Is there a way to calculate the Alexander polynomial of this knot simply from the link and the arc?
If not, can we say at least something about the Alexander polynomial, e.g. whether it is non-trivial, or what the determinant of the knot is?
Are there any conditions on the link and the arc that ensure that the Alexander polynomial is non-trivial?