2
$\begingroup$

Let $p$ be a rational prime number, and let $K$ be a finite extension of $\mathbb Q_p$. Let $A$ be an abelian variety over $K$. For any rational prime number $\ell$, let $K(A[\ell])$ be the field of definition of the $\ell$-torsion points of $A$.

Suppose $A$ and $A'$ are $K$-isogenous abelian varieties over $K$.

Is there an odd prime number $\ell\neq p$ such that $K(A[\ell])=K(A'[\ell])$?

Are there even infinitely many such $\ell$?

$\endgroup$
4
  • 1
    $\begingroup$ If $f:A \to A'$ is an isogeny, then $f$ induces an isomorphism from $A[l]$ onto $A'[l]$ for any $l$ prime to the degree of $f$. $\endgroup$
    – naf
    Aug 11, 2013 at 10:59
  • $\begingroup$ Thanks for pointing this out. A priori this isomorphism is an isomorphism of abstract groups, is this isomorphism in addition compatible with Galois actions? If yes, this would imply $K(A[\ell])=K(A'(\ell))$ for all $\ell\nmid \deg(f)$? $\endgroup$ Aug 11, 2013 at 11:04
  • 1
    $\begingroup$ The isogeny $f$ will be compatible with the Galois actions if it is a $K$-isogeny. $\endgroup$ Aug 11, 2013 at 11:08
  • $\begingroup$ Yes, it's induced by f, and so compatible with Galois. $\endgroup$
    – abz
    Aug 11, 2013 at 11:08

1 Answer 1

4
$\begingroup$

Let $f:A\to A′$ be an isogeny over $K$. Then there exists an isogeny $g:A'\to A$ such that the composite of $f$ with $g$ is multiplication by the degree of $f$ (see any book on abelian varieties). It follows that, for every prime $l$ not dividing its degree, $f$ defines an isomorphism from $A[l]$ onto $A'[l]$. This isomorphism automatically commutes with the action of the Galois group.

$\endgroup$
1
  • $\begingroup$ Ok, now everything is clear. Thank you all very much for the answers. $\endgroup$ Aug 16, 2013 at 10:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.