Proof: The following argument is quite close to the proof of the Sturmfels–Tevelev formula given by Baker, Payne and Rabinoff (see [BPR11], Theorem 8.2).
Let be the closure of in and let be a generic homomorphism onto a split torus of rank where generic is meant in the same way as in 7.4. Since removing lower dimensional subvarieties does not change and the tropical multiplicity functions, we may assume that is a finite morphism and then is affine.
Let be the closure of in . Let be a regular point of , i.e. is contained in the relative interior of an -dimensional polytope . We may choose for an integral -affine polytope. We set and . We consider the affinoid subdomains in and in . By finiteness of , the set is an affinoid subdomain of and restricts to a finite morphism . Let be the canonical formal affine -models of associated to the algebra of power bounded elements in the corresponding affinoid algebra. Moreover, let be the closure of in . Then we have canonical morphisms
|
|
|
(4) |
of admissible formal affine schemes over in the sense of Bosch, Lütkebohmert and Raynaud (see [BL93], §1). We claim that all these morphisms are finite and surjective. Obviously, the generic fibres of the first and second morphism are finite and surjective. To see that the generic fibre of the third morphism is finite, we note first that is finite by construction of and hence is in the relative interior of an affinoid subdomain of which is contained in . We conclude that is a proper map (see the proof of Theorem 4.31 in [BPR11] for more details about the argument). Since is the disjoint union of the finitely many affinoids , , we conclude that induces a proper morphism of affinoids. By Kiehl’s direct image theorem ([BGR84], Theorem 9.6.3/1), this morphism is finite and hence also surjective using dimensionality arguments. We conclude that all three morphisms in (4) are surjective and finiteness follows from [BPR11], Proposition 3.13.
The degree of over the affinoid torus is well-defined as is irreducible (see [BPR11], Section 3, for a discussion of degrees). Since the degree does not change by passing to an affinoid subdomain of (see [BPR11], Proposition 3.30), we get
|
|
|
(5) |
The projection formula ([BPR11], Proposition 3.32) shows
|
|
|
(6) |
where ranges over all irreducible components of . We conclude from (5) and (6) that
|
|
|
(7) |
where ranges over all irreducible components of and ranges over all irreducible components of mapping onto . Since the special fibre of is isomorphic to the initial degeneration , all irreducible components are isomorphic to the torus (see [BPR11], Theorem 4.29) proving
|
|
|
(8) |
Using (7) and (8), we get
|
|
|
(9) |
Since is the preimage of the affinoid subdomain of , we deduce from [BPR11], Proposition 3.30, that is of pure degree over and hence the projection formula again shows the equality
|
|
|
(10) |
of cycles in . Inserting (10) in (9) by using that the special fibre of is reduced, we get
|
|
|
where is the multiplicity of the irreducible component in the special fibre of . By definition, the right hand side is equal to which proves the claim for -rational points in . An obvious density argument finishes the proof.