No label defined (Q4350108)

An important tool in diophantine geometry are heights. The height of a point measures the arithmetic complexity of its coordinates. Let us briefly recall its definition. Let \(P\) be a point of projective space with rational coordinates. We may assume that the coordinates \(x_0,\dots,x_n\) are entire and have no common divisor. Then the height of \(P\) is defined by \(h(P):= \max_{j=0,\dots,n}| x_j|\). Easily, this is generalized to points with coordinates in the algebraic closure \(\overline\mathbb{Q}\) of \(\mathbb{Q}\). For a projective variety \(X\) over a number field, we use an embedding (or more generally a morphism) \(\varphi\) into some projective space to define the height \(h_\varphi:=h\circ\varphi\) on \(X(\overline\mathbb{Q})\). The basic question is how \(h_\varphi\) depends on the choice of \(\varphi\). This is answered by the following theorem of \textit{A. Weil} [Ann. Math., II. Ser. 53, 412-444 (1951; Zbl 0043.27002)]: The height \(h_\varphi\) depends only on the isomorphy class of \(\varphi^*{\mathcal O}_{\mathbb{P}^n}(1)\) up to bounded functions. Any element of the Picard group of \(X\) may be written as the difference of two elements of the form above. Therefore we get a homomorphism from the Picard group to the real functions on \(X(\overline\mathbb{Q})\) modulo bounded functions. Let \(Y\) be a subvariety of projective space. Then there is a canonical multi-homogeneous polynomial associated to \(Y\) called the Chow form. Nesterenko and Philippon have defined the height of \(Y\) as the height of the coefficient vector of the Chow form. Using arithmetic intersection theory, Faltings defined equivalently the height of \(Y\) as an arithmetic degree of \(Y\). This idea leads to many important properties of heights of subvarieties by translating the corresponding properties of the usual degree. As for points, we can define the height of a subvariety of any projective variety. In a previous paper [\textit{W. Gubler}, Math. Ann. 298, No. 3, 427-455 (1994; Zbl 0792.14012)], Weil's theorem was generalized to heights of subvarieties. It says that the height \(h_\varphi\) on subvarieties depends only on the isomorphy class of \(\varphi^*{\mathcal O}_{\mathbb{P}^n} (1)\) up to functions bounded by a multiple of the degree. The theory of heights of points works not only in the number field case. We may use any ground field with a set of absolute values satisfying the product formula. Moreover, Vojta pointed out that in Nevanlinna theory the characteristic function is an analogue of the height of points and the first main theorem in Nevanlinna theory corresponds to Weil's theorem [cf. \textit{P. Vojta}, ``Diophantine approximations and value distribution theory'', Lect. Notes Math. 1239 (1987; Zbl 0609.14011)]. This leads to far reaching results and conjectures for number fields and in Nevanlinna theory. Since Faltings' definition of the height of a subvariety uses arithmetic intersection theory, Weil's theorem for subvarieties was proved in the number field case. Here, we omit this restriction and we prove the theorem over very general ground fields. We do not assume that the product formula is satisfied. In particular, we get a generalization of the first main theorem in Nevanlinna theory. The paper is organized in the following way. In the first section, we introduce local heights of complete varieties \(X\) over \(K\). In the second section, we introduce the notion of \(M\)-fields. This is a field \(K\) together with a measure space \(M\). Moreover, a family \(|\;|_v \), \(v\in M\), is given which may be viewed as absolute values on \(K\). The difference with the usual absolute values is, that for given arguments, they are defined and satisfy the axioms only for almost all \(v\in M\). It is assumed that for non-zero \(f\in K\), \(\log| f|_v\) is integrable on \(M\). The prototype of an \(M\)-field is the field of meromorphic functions on \(\mathbb{C}\) with measure space \(M_R\). It will be proved that the algebraic closures of \(\mathbb{Q}\) and of the field of meromorphic functions on \(\mathbb{C}\) are canonically \(M\)-fields. The third section deals with global heights of subvarieties over an \(M\)-field \(K\). If \(K\) is a number field, then the global heights agree with the absolute version of Faltings' definition. Finally, the generalization of the first main theorem in Nevanlinna theory is given. For the entire collection see [Zbl 0864.00054].