No label defined (Q4805271)

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 1908542
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 1908542

      Statements

      0 references
      11 May 2003
      0 references
      elliptic curves
      0 references
      root numbers
      0 references
      Let \(E/\mathbb Q\) be an elliptic curve. If the Birch and Swinnerton-Dyer conjecture is true for \(E/\mathbb Q\) then the root number of \(E/\mathbb Q\) is equal to \((-1)^{\text{rank}(E(\mathbb Q))}\). It is expected that for a ``generic'' family of elliptic curves over \(\mathbb Q\) that the number of curves having odd rank is the same as the number of those having even rank [\textit{D. Zagier} and \textit{G. Kramarz}, J. Indian Math. Soc., New Ser. 52, 51-69 (1987; Zbl 0688.14016)]. If the family is one of twists of a given curve (hence the \(j\)-invariant is constant), under the assumption of Selmer's conjecture, \textit{J. W. S. Cassels} and \textit{A. Schinzel} [Bull. Lond. Math. Soc. 14, 345-348 (1982; Zbl 0474.14010)] show that \((7+7t^4)y^2=x^3-x\) has odd rank for every \(t\in\mathbb Q\). If \(E/\mathbb Q\) is an elliptic curve admitting everywhere good reduction over some Abelian extension of \(\mathbb Q\), then \textit{D. Rohrlich} [Compos. Math. 87, 119-151 (1993; Zbl 0791.11026)] proves that the root number \(W(E^{f(t)})\) of the twist \(E^{(f(t))}\) of \(E\) by \(f\in\mathbb Q[t]\) is equal to \(W(E)\text{sgn}(f(t))\). The author previously proved [Proc. Am. Math. Soc. 127, 1597-1603 (1999; Zbl 0944.11019)] that the set \(\{\text{Av}_{\mathbb Q}(W(E^{(f(t))}))\) of average root numbers for \(t\) running through \(\mathbb Q\) is dense in the interval \([-1,1]\). It was suggested by \textit{J. Silverman} [J. Reine Angew. Math. 504, 227-236 (1998; Zbl 0923.11087)] that this kind of phenomenon would only take place on constant families. In the paper under review the author gives two counter-examples to this suggestion, both with \(t\) running through \(\mathbb Z\). The first one is the family \(E_t:y^2=x^3+tx^2-(t+3)x+1\) for which the \(j\)-invariant \(j(t)=256(t^2+3t+9)\); in this case \(W(E_t)=-1\) for every \(t\in\mathbb Z\). The second is the family \(E_t=y^2=x^3+\frac t4x^2-\frac{36t^2}{t-1728}x-\frac{t^3}{t-1728}\) for \(t\) running through \(\mathbb Z\). In this case \(j(t)=t\) and \(W(E_t)=0.0037182\ldots\).
      0 references

      Identifiers