A Shafarevich-Faltings theorem for rational functions (Q998625)

The authors consider non-constant morphisms \(\varphi: {\mathbb{P}}_K^1\to {\mathbb{P}}_K^1\) where \(K\) is a number field. For a finite place \(v\) of \(K\), denote by \(O_v\) its local ring and by \(k_v\) it residue field. The morphism \(\varphi\) is said to have \textit{simple good reduction} at a finite place \(v\) of \(K\) if \(\varphi\) extends to a map \({\mathbb{P}}_{O_v}^1\to {\mathbb{P}}_{O_v}^1\). In other words, if we represent \(\varphi\) as \((x:y)\mapsto (P(x,y):Q(x,y))\) where \(P,Q\) are two binary forms of equal degree with coefficients in \(O_v\) not both divisible by a non-unit of \(O_v\), then \(\varphi\) has simple good reduction at \(v\) if \(P,Q\) do not have a common factor in the algebraic closure of the residue field of \(O_v\). The authors argue that this notion of reduction is not restrictive enough and therefore they introduce another type of reduction, called \textit{critically good reduction}. There main result is a finiteness theorem for morphisms with critically good reduction outside a finite set of places \(S\) of \(K\). Denote by \(O_S\) the ring of \(S\)-integers in \(K\). The authors fix an isomorphism from \({\mathbb{P}}_K^1\) to \({\mathbb{P}}_{O_S}^1\times_{\text{Spec}\, O_S} \text{Spec} K\) and use this to define a reduction map from \({\mathbb{P}}^1(\overline{K})\) to \({\mathbb{P}}^1(\overline{k_v})\) for every \(v\not\in S\). Loosely speaking, denote by \(R_{\varphi}\) the ramification divisor of \(\varphi\) over \(\overline{K}\). Then \(\varphi\) is said to have critically good reduction at a place \(v\not\in S\) if after reduction mod \(v\), distinct points of \(\text{supp} (R_{\varphi})\) remain distinct, and also, distinct points of \(\varphi (\text{supp} (R_{\varphi}))\) remain distinct. Further, two morphisms \(\varphi ,\psi :\, {\mathbb{P}}_K^1\to {\mathbb{P}}_K^1\) are called equivalent if there are \(\sigma ,\gamma\in \text{Aut}({\mathbb{P}}_{O_S}^1)\) such that \(\psi =\sigma\varphi\gamma\). The authors show that in many cases, critically good reduction implies simple good reduction, but they give also examples of morphisms with simple good reduction but not critically good reduction and vice versa. The authors' main result is that for any positive integer \(n\), there are only finitely many equivalence classes of morphisms \(\varphi :{\mathbb{P}}_K^1\to {\mathbb{P}}_K^1\) of degree \(n\), such that \(\varphi\) has critically good reduction outside \(S\) and such that \(|\text{supp}( R_{\varphi})|\geq 3\). Their main tools are a finiteness result of \textit{B. J. Birch} and \textit{J. R. Merriman} [Proc. Lond. Math. Soc., III. Ser. 24, 385--394 (1972; Zbl 0248.12002)] and \textit{K. Győry} and the reviewer [Compos. Math. 79, No. 2, 169--204 (1991; Zbl 0746.11020)] for equivalence classes of binary forms of given discriminant, and a structure theorem, due to \textit{A. Grothendieck} [Lecture Notes in Mathematics. 224. Berlin-Heidelberg-New York: Springer-Verlag (1971; Zbl 0234.14002)] and \textit{S. Mori} [Ann. Math. (2) 110, 593--606 (1979; Zbl 0423.14006)] for the set of scheme morphisms \(f: A\to B\) that extend a given morphism \(p: Z\to B\) where \(Z\) is given closed subscheme of \(A\).