No label defined (Q4252164)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1305295
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 1305295

      Statements

      20 August 2000
      0 references
      hyperelliptic curve
      0 references
      Jacobian
      0 references
      endomorphisms
      0 references
      period matrix
      0 references
      0 references
      Let \(n\) be a positive integer and \(\sigma =(\alpha (1),\beta (1),\dots ,\alpha (n),\beta (n))\in \mathbb{Z}^{2n}\) with 0\(\leq \alpha (1)<\beta (1)<\dots <\alpha (n)<\beta (n).\) For an affine variable \(z\) consider the hyperelliptic curve \(C_{\sigma }(z)\) defined by the equation \( y^{2}=x(x-z^{\alpha (1)})(x-z^{\beta (1)})\dots (x-z^{\alpha (n)})(x-z^{\beta (n)})\) over the field \(\mathbb{Q(}z\mathbb{).}\) This is a smooth curve of genus \(n,\) so its Jacobian \(\operatorname {Jac}C_{\sigma }(z)\) is an abelian variety of dimension \(n\), also defined over \(\mathbb{Q}(z)\), thus has a one-parameter family of abelian varieties. In order to apply the specialization theorems one has to know the endomorphism ring \(\operatorname {EndJac}C_{\sigma }(z).\) The element \(\sigma \) as above is ``good'' if the differences \(\beta (1)-\alpha (1),\dots ,\beta (n)-\alpha (n)\) are distinct. Theorem. Suppose that \(\sigma \) is good. \ Then for the algebraic closure \(F\) of \(\mathbb{C}(z)\) we have \(\operatorname {End}_{F} \operatorname {Jac}C_{\sigma }(z)=\mathbb{Z}.\) In other words, the generic fibre has only trivial endomorphisms. For \(\varsigma \) in \(\mathbb{C}\) not a root of unity obtain through \( z=\varsigma \) a specialized curve \(C_{\sigma }(\varsigma)\) and Jacobian, both defined over \(\mathbb{Q}(\varsigma).\) If \(\varsigma \) is an \ algebraic number, then write \(d(\varsigma)\) for its degree, and write \( h(\varsigma)\) for its absolute logarithmic height; \ also say that \( \varsigma \) (not a root of unity) is ``exceptional'' if \(\operatorname {End}_{\mathbb{C} }\operatorname {Jac}C_{\sigma }(\varsigma)\neq \mathbb{Z.}\) For example, this happens if \(\operatorname {Jac}C_{\sigma }(\varsigma)\) is not simple. Corollary 1. Suppose that \(\sigma \) is good. Then there are constants \( c_{1},k_{1},\) depending only on \(n\) and \(\sigma ,\) such that for any \(d\geq 1,\) \(h\geq 1,\) the number of exceptional \(\varsigma \) with \(d(\varsigma)\leq d\) and \(h(\varsigma)\leq h\) \(\;\)is at most \(c_{1}(\max \{d,h\})^{k_{1}}.\) Say that \(\varsigma \) is ``very exceptional'' if \(\operatorname {End}_{\mathbb{C} }\operatorname {Jac}C_{\sigma }(\varsigma)\) is not isomorphic to a subring of the ring of rational matrices of order \(n\). Corollary 3. (André-Masser). Suppose that \(n\geq 3\) is odd, and \(\sigma \) is good. Then there are constants \(c_{3},k_{3},\) depending only on \(n\) and \( \sigma \), such that for any \(\;d\geq 1\) the number of very exceptional \( \varsigma \) with \(d(\varsigma)\leq d\) is at most \(c_{3}d^{k_{3}}.\) Proofs are purely analytic in character and compute the ``period matrix'' of \( \operatorname {Jac}C_{\sigma }(\varsigma)\) rather accurately as \(\varsigma \rightarrow \infty \) by real values. For the entire collection see [Zbl 0911.00025].
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references