### November 27, 2006

2000 Mathematics Subject Classification. Primary 11G05.
<ph f="cmbx">Open conditions for infinite multiplicity eigenvalues on elliptic curves</ph>

### Bo-Hae Im and Michael Larsen

Department of Mathematics, University of Utah, Salt Lake City, Utah 84112, USA E-mail address : im@math.utah.edu Department of Mathematics, Indiana University, Bloomington, Indiana 47405, USA E-mail address : larsen@math.indiana.edu
• Abstract. Let $E$  be an elliptic curve defined over a number field $K$  . We show that for each root of unity $\zeta$  , the set ${\Sigma }_{\zeta }$  of $\sigma \in Gal\left(\overline{K}/K\right)$  such that $\zeta$  is an eigenvalue of infinite multiplicity for $\sigma$  acting on $E\left(\overline{K}\right)\otimes \mathbb{C}$  has non-empty interior.
For the eigenvalue $-1$  , we can show more: for any $\sigma$  in $Gal\left(\overline{K}/K\right)$  , the multiplicity of the eigenvalue $-1$  is either $0$  or $\infty$  . It follows that ${\Sigma }_{-1}$  is open.

1 Introduction

Let $K$  be a number field, $\overline{K}$  an algebraic closure of $K$  , and ${G}_{K}:=Gal\left(\overline{K}/K\right)$  the absolute Galois group of $\overline{K}$  over $K$  . Let $E$  be an elliptic curve defined over $K$  . There is a natural continuous action of ${G}_{K}$  on the countably infinite-dimensional complex vector space ${V}_{E}:=E\left(\overline{K}\right)\otimes \mathbb{C}$  . The resulting representation decomposes as a direct sum of finite-dimensional irreducible representations in each of which ${G}_{K}$  acts through a finite quotient group.
In particular, the action of every $\sigma \in {G}_{K}$  on ${V}_{E}$  is diagonalizable, with all eigenvalues roots of unity. In [3, the first-named author showed that for generic $\sigma$  , every root of unity appears as an eigenvalue of countably infinite multiplicity. This is true both in terms of measure and of Baire category. However, there exist $\sigma$  for which the spectrum is quite different: trivially, the identity and complex conjugation elements; less trivially, examples which can be constructed for an arbitrary set $S$  of primes, such that $\zeta$  is an eigenvalue if and only if every prime factor of its order lies in $S$  .
Throughout this paper, we will write ${\Sigma }_{\zeta }$  for the subset of ${G}_{K}$  consisting of elements $\sigma$  acting as $\zeta$  on an infinite-dimensional subspace of ${V}_{E}$  ( $E$  and $K$  being fixed). For $\zeta =1$  , a good deal is known. In [2, it is proved that whenever $1$  appears as an eigenvalue of $\sigma$  at all, we have $\sigma \in {\Sigma }_{1}$  . It follows that ${\Sigma }_{1}$  is open. By [4, when $K=\mathbb{Q}$  , ${\Sigma }_{1}$  is all of ${G}_{K}$  , and quite possibly this may be true without restriction on $K$  . We have already observed that ${\Sigma }_{\zeta }\ne {G}_{K}$  for $\zeta \ne 1$  . We can still hope for positive answers to the following progression of increasingly optimistic questions:
Question 1.1. Does ${\Sigma }_{\zeta }$  have non-empty interior for all $\zeta$  ?
Question 1.2. Is ${\Sigma }_{\zeta }$  open for all $\zeta$  ?
Question 1.3. Do all eigenvalues of $\sigma$  acting on ${V}_{E}$  appear with infinite multiplicity?
In this paper, we give an affirmative answer to Question  1.1 for all $\zeta$  and an affirmative answer to all three questions for $\zeta =-1$  .
The difficulty in proving such theorems is that placing $\sigma$  in a basic open subset $U$  of ${G}_{K}$  amounts to specifying the action of $\sigma$  on a finite Galois extension $L$  of $K$  . By the Mordell-Weil theorem, $E\left(L\right)\otimes \mathbb{C}$  cannot provide an infinite eigenspace for $\zeta$  . Thus, theintersection of eigenspaces ${\bigcap }_{\sigma \in U}{V}_{E}^{\sigma -\zeta }$  is finite-dimensional. Thus, the behavior of a finite collection of rational points must be enough to guarantee the existence of infinitely many linearly independent points on the curve with specified $\sigma$  -action.
We would like to thank L. Moret-Bailly for correcting an error in the proof of Proposition  3.1 in an earlier draft of this manuscript.

2 Multiplicity of the eigenvalue -1

In this section, we answer Questions  1.2  and  1.3 for $\zeta =-1.$
Proposition 2.1. Let $E/K$  be an elliptic curve over $K$  . Suppose $-1$  is an eigenvalue of the action of $\sigma \in {G}_{K}$  on ${V}_{E}$  . Then the $-1$  -eigenspace of $\sigma$  is infinite-dimensional.
• Proof. As $-1$  is an eigenvalue of $\sigma$  acting on ${V}_{E}$  , it is an eigenvalue of $\sigma$  acting on $E\left(\overline{K}\right)\otimes \mathbb{Q}$  . Clearing denominators, there exists a non-torsion $P\in E\left(\overline{K}\right)$  such that $\sigma \left(P\right)+P\in E\left(\overline{K}{\right)}_{tor}$  . Replacing $P$  by a suitable positive integral multiple, $\sigma \left(P\right)=-P$  .
Let ${y}^{2}=f\left(x\right)$  be a fixed Weierstrass equation of $E/K$  . Let $P=\left(\alpha ,\sqrt{f\left(\alpha \right)}\right)$  . As $\sigma \left(P\right)=-P$  , we have $\alpha \in {\overline{K}}^{\sigma }$  but $\sigma \left(\sqrt{f\left(\alpha \right)}\right)=-\sqrt{f\left(\alpha \right)}$  so $\sqrt{f\left(\alpha \right)}/\in {\overline{K}}^{\sigma }$  . Then, $\sqrt{f\left(\alpha \right)}/\in K\left(\alpha \right)$  , since $K\left(\alpha \right)\subseteq {\overline{K}}^{\sigma }$  Note that $f\left(\alpha \right)\in K\left(\alpha \right)\subseteq {\overline{K}}^{\sigma }$  . Let $c=f\left(\alpha \right)\in K\left(\alpha \right)$  . We still have $\sigma \in$  Gal $\left(\overline{K}/K\left(\alpha \right)\right)$  and $\sigma \left(\sqrt{c}\right)=-\sqrt{c}$  .
Let ${E}^{\prime }/K\left(\alpha \right)$  denote the twist ${y}^{2}=cf\left(x\right)$  . Then, ${E}^{\prime }$  has a rational point ${P}^{\prime }=\left(\alpha ,f\left(\alpha \right)\right)$  over $K\left(\alpha \right)$  . The $\overline{K}$  -isomorphism $\phi :E\to {E}^{\prime }$  mapping $\left(x,y\right)↦\left(x,\sqrt{f\left(\alpha \right)y}\right)$  sends $P$  to ${P}^{\prime }$  , so ${P}^{\prime }$  is of infinite order on ${E}^{\prime }$  . By ([2,Theorem5.3, ${E}^{\prime }\left({\overline{K}}^{\sigma }\right)$  has infinite rank. Let $\left\{{P}_{i}^{\prime }=\left({x}_{i},\sqrt{cf\left({x}_{i}\right)}\right){\right\}}_{i=1}^{\infty }$  be an infinite sequence of linearly independent points of ${E}^{\prime }$  generating the infinite dimensional eigenspace of $1$  of $\sigma$  in ${E}^{\prime }\left(\overline{K}\right)\otimes \mathbb{C}$  . Then, $\sigma \left({x}_{i}\right)={x}_{i}$  and $\sigma \left(\sqrt{f\left({x}_{i}\right)}\right)=-\sqrt{f\left({x}_{i}\right)}$  for all $i$  , since $\sigma \left(\sqrt{c}\right)=-\sqrt{c}$  .
Let ${P}_{i}={\phi }^{-1}\left({P}_{i}^{\prime }\right)=\left({x}_{i},\sqrt{f\left({x}_{i}\right)}\right)$  . These are points of the given elliptic curve $E$  such that $\sigma \left({P}_{i}\right)=-{P}_{i}$  for all $i$  , since $\sigma \left({x}_{i}\right)={x}_{i}$  and $\sigma \left(\sqrt{f\left({x}_{i}\right)}\right)=-\sqrt{f\left({x}_{i}\right)}$  .
The points ${P}_{i}$  are linearly independent because the ${P}_{i}^{\prime }$  are so. Therefore, $\left\{{P}_{i}\otimes 1{\right\}}_{i=1}^{\infty }$  generates an infinite dimensional subspace of the $-1$  -eigenspace of $\sigma$  on ${V}_{E}$  . This completes the proof.
Theorem 2.2. Let $E/K$  be an elliptic curve over $K$  . Then, ${\Sigma }_{-1}$  is open.
• Proof. We have already seen that if $\sigma \in {\Sigma }_{-1}$  , we can choose a point $P\in E\left(\overline{K}\right)$  of infinite order such that $\sigma \left(P\right)=-P$  . By Proposition  2.1 , $\tau \left(P\right)=-P$  implies $\tau \in {\Sigma }_{-1}$  It follows that ${\Sigma }_{-1}$  contains the open neighborhood $\left\{\tau \in {G}_{K}|\tau \left(P\right)=\sigma \left(P\right)\right\}$  of $\sigma$  .
Remark 2.3. The same argument shows that Questions  1.2  and  1.3 have an affirmative answer for $\zeta =\omega$  (resp. $\zeta =i$  ) when $E$  has complex multiplication by $\mathbb{Z}\left[\omega \right]$  (resp. $\mathbb{Z}\left[i\right]$  ).

3 Interior Points

In this section, we show that for every root of unity $\zeta$  , the set ${\Sigma }_{\zeta }$  contains a non-empty open subset. We assume that the order of $\zeta$  is $n\ge 3$  , the case $n=1$  having been treated in [2, and the case $n=2$  in Theorem  2.2 .
Our strategy will be to find points ${Q}_{i}\in E\left(\overline{K}\right)$  such that the $\sigma$  -orbit of ${Q}_{i}$  has length $n$  . For each such point ${Q}_{i}$  , we set
 $\begin{array}{c}{R}_{i}:={\sum }_{i=0}^{n-1}{\sigma }^{i}\left({Q}_{i}\right)\otimes {\zeta }^{-i}\end{array}$ (1)
and observe that ${R}_{i}$  is a $\zeta$  -eigenvector of $\sigma$  provided that it is non-zero.
We therefore begin with the following proposition:
Proposition 3.1. Let $X$  be a Riemann surface of genus $g>1$  with an automorphism $\sigma$  of order $n$  . Then $X$  contains a non-empty open set $U$  such that $x\in U$  implies that ${\sum }_{i=0}^{n-1}\left[{\sigma }^{i}x\right]\otimes {\zeta }^{-i}\ne 0$  in $PicX\otimes \mathbb{C}$  .
• Proof. We can regard $X$  as the group of complex points of a non-singular projective curve whose Picard scheme has complex locus $PicX$  . Then $PicX\otimes \mathbb{Z}\left[\zeta \right]$  is the group of complex points of a group scheme whose identity component $Pi{c}^{0}X\otimes \mathbb{Z}\left[\zeta \right]$  is isomorphic to the $\phi \left(n\right)$  th power of the Jacobian variety of this curve. The action of $\sigma$  on $X$  defines an action on $PicX$  , and the map $\psi :PicX\to PicX\otimes \mathbb{Z}\left[\zeta \right]$  given by $\psi \left(y\right)={\sum }_{i=0}^{n-1}{\sigma }^{i}y\otimes {\zeta }^{-i}$  then comes from a morphism of group schemes. The image of $\psi$  actually lies in $Pi{c}^{0}X\otimes \mathbb{Z}\left[\zeta \right]$  , and its kernel ${P}_{\zeta }^{0}$  is Zariski-closed in $PicX$  .
The set ${P}_{\zeta }$  of $y$  such that $\psi \left(y\right)$  maps to 0 in $PicX\otimes \mathbb{C}$  is the union of all translates of ${P}_{\zeta }^{0}$  by torsion points of $PicX$  . Applying Raynaud's theorem [7(i.e., the proof of theManin-Mumford conjecture) to the image of $X$  in $PicX/{P}_{\zeta }^{0}$  , the intersection $X\cap {P}_{\zeta }$  is finite whenever ${P}_{\zeta }^{0}$  does not contain the identity component of $PicX$  . It therefore suffices to prove that the Lie algebra of ${P}_{\zeta }^{0}$  is a proper subspace of the Lie algebra of $PicX$  . We identify the latter Lie algebra in the usual way [1,Ch. 2, §6with ${H}^{1}\left(X,{\mathcal{O}}_{X}\right)={H}^{0,1}\left(X\right)$  .
As ${V}_{E}=\left(E\left(\overline{K}\right)\otimes \mathbb{Q}\right){\otimes }_{\mathbb{Q}}\mathbb{C}$  , the eigenvalues of $\sigma$  form a Galois-invariant set. Therefore, it suffices to show some primitive $n$  th root of unity is an eigenvalue of $\sigma$  acting on ${H}^{0,1}$  .
The Hodge decomposition ${H}^{1}\left(X\left(\mathbb{C}\right),\mathbb{C}\right)={H}^{1,0}\oplus \overline{{H}^{1,0}}$  implies that it suffices to show that $\zeta$  is an eigenvalue of $\sigma$  acting on ${H}^{1}\left(X\left(\mathbb{C}\right),\mathbb{C}\right)$  .
Let ${R}_{\mathbb{C}}\left(G\right)$  denote the ring of complex (virtual) representations of $G$  . For any subgroup $H$  of $G:=〈\sigma 〉$  , let ${R}_{G/H}$  denote the regular representation of $G/H$  regarded as an element of ${R}_{\mathbb{C}}\left(G\right)$  , and let ${I}_{H}:={R}_{G}-{R}_{G/H}$  . In particular, ${I}_{\left\{1\right\}}=0$  . Regarded as an element of ${R}_{\mathbb{C}}\left(G\right)$  , the $G$  -representation ${H}^{1}\left(X\left(\mathbb{C}\right),\mathbb{C}\right)\right)$  is $2g+\left(2h-2\right){I}_{G}+{\sum }_{\left[x\right]\in X/G}{I}_{Sta{b}_{G}\left(x\right)},$  where $h$  is the genus of $X/G$  , and $Sta{b}_{G}\left(x\right)$  is the stabilizer of any element of $X$  representing the $G$  -orbit $\left[x\right]$  . This is worked out in the case that $h=0$  in [6,Prop. 2.2), but the method (in which the character of ${H}^{1}\left(X\left(\mathbb{C}\right),\mathbb{C}\right)\right)$  as a representation of $G$  is deduced from the Hurwitz formula and the Lefschetz trace formula) works in general.
The multiplicity of $\zeta$  as an eigenvalue of $\sigma$  on ${H}^{1}\left(X\left(\mathbb{C}\right),\mathbb{C}\right)\right)$  is therefore $2g-2$  plus the number of ramification points of the cover $X\to X/G$  . This is positive except in two cases: the cyclic cover ${\mathbb{P}}^{1}\to {\mathbb{P}}^{1}$  of degree $n$  (necessarily ramified over two points) and a degree $n$  isogeny of elliptic of elliptic curves. As $g\ge 2$  , neither case applies.
Theorem 3.2. Let $E/K$  be an elliptic curve over a number field $K$  . For each root of unity $\zeta$  , there exists a nonempty open subset ${\Sigma }_{\zeta }$  of Gal $\left(\overline{K}\right)$  suchthat the multiplicity of the eigenvalue $\zeta$  for $\sigma \in {\Sigma }_{\zeta }$  acting on $E\left(\overline{K}\right)\otimes \mathbb{C}$  is infinite.
• Proof. Let $\zeta$  be an $n$  th root of unity. Let ${\lambda }_{1},{\lambda }_{2},{\lambda }_{3},\infty$  be the ramification points of a double cover $E\to {\mathbb{P}}^{1}$  , and let $\lambda$  denote the cross-ratio of $\left({\lambda }_{1},{\lambda }_{2},{\lambda }_{3},\infty \right)$  . Choose $a,b\in \overline{K}$  such that the ordered quadruple ( $a,b,\zeta a,\zeta b$  ) satisfies $\frac{\left(\zeta a-a\right)\left(\zeta b-b\right)}{\left(\zeta b-a\right)\left(\zeta a-b\right)}=\lambda$  This is always possible; for instance, setting $a=1$  , we get a non-trivial quadratic equation for $b$  , and since $\lambda$  is not $1$  or $\infty$  , we have $b,\zeta b\notin \left\{a,\zeta a\right\}$  . Thus the elliptic curves ${X}_{i}:{y}^{2}=\left(x-{\zeta }^{i-1}a\right)\left(x-{\zeta }^{i-1}b\right)\left(x-{\zeta }^{i}a\right)\left(x-{\zeta }^{i}b\right),\text{for}i=1,\dots ,n.$  all have the same $j$  -invariant as $E$  .
Let $L=K\left(a,b,\zeta \right)$  . Fix $q\in K$  such that $L\left(\sqrt[n]{q}\right)$  is a Galois $\mathbb{Z}/n\mathbb{Z}$  -extension of $L$  . We claim that ${\Sigma }_{\zeta }$  contains the open set ${U}_{\zeta }:=\left\{\sigma \in Gal\left(\overline{K}/L\right)|\sigma \left(\sqrt[n]{q}\right)=\zeta \sqrt[n]{q}\right\}.$  Let $M=L\left(\sqrt[n]{q}\right)$  . For $N$  any number field containing $M$  , let ${C}_{N}$  denote the affine curve over $N$  $SpecN\left[x,{y}_{1},\dots ,{y}_{n}\right]/\left({P}_{1}\left(x,{y}_{1}\right),\dots ,{P}_{n}\left(x,{y}_{n}\right),{y}_{1}\cdots {y}_{n}-\left({x}^{n}-{a}^{n}\right)\left({x}^{n}-{b}^{n}\right)\right)$  where ${P}_{i}\left(x,y\right)={y}^{2}-\left(x-{\zeta }^{i-1}a\right)\left(x-{\zeta }^{i}a\right)\left(x-{\zeta }^{i-1}b\right)\left(x-{\zeta }^{i}b\right).$  Note that the equation ${y}_{1}\cdots {y}_{n}-\left({x}^{n}-{a}^{n}\right)\left({x}^{n}-{b}^{n}\right)=0$  merely selects one of the two irreducible components of the 1-dimensional affine scheme cut out by the other equations.
Let $X$  denote the compact Riemann surface which is the compactification of ${C}_{N}\left(\mathbb{C}\right)$  .
By the Hurwitz genus formula, the genus of $X$  is $\left(n-2\right){2}^{n-2}+1$  , which is greater than $1$  since $n\ge 3$  . For any $n$  -tuple $\left({k}_{1},\dots ,{k}_{n}\right)\in \left\{0,1{\right\}}^{n}$  with even sum, the map  $\begin{array}{c}\left(x,{y}_{1},\dots ,{y}_{n}\right)↦\left(\zeta x,\left(-1{\right)}^{{k}_{1}}{\zeta }^{2}{y}_{n},\left(-1{\right)}^{{k}_{2}}{\zeta }^{2}{y}_{1},\left(-1{\right)}^{{k}_{3}}{\zeta }^{2}{y}_{2},\dots ,\left(-1{\right)}^{{k}_{n}}{\zeta }^{2}{y}_{n-1}\right)\end{array}$ (2)
defines an automorphism $\sigma$  of ${C}_{N}$  and therefore of $X$  . As the ${k}_{i}$  have even sum, $\sigma$  is of order $n$  . If $x\in \sqrt[n]{q}{L}^{*}$  and $\sigma \in {U}_{\zeta }$  , then $\sigma \left(x\right)=\zeta x$  , so $\sigma \left({y}_{i}{\right)}^{2}={\zeta }^{4}{y}_{i-1}^{2},$  and so there exists an $n$  -tuple $\left({k}_{1},\dots ,{k}_{n}\right)$  with even coordinate sum such that $\sigma$  acts on $Q:=\left(x,{y}_{1},\dots ,{y}_{n}\right)$  by ( 2 ). By Proposition  3.1 , for all but finitely many values of $x$  , $R:={\sum }_{i=0}^{n-1}{\sigma }^{i}\left(Q\right)\otimes {\zeta }^{-i}$  is a non-zero eigenvector of $\sigma$  with eigenvalue $\zeta$  .
Assume now that $N$  is a finite Galois extension of $M$  . Consider the morphism from ${C}_{N}$  to the affine line over $M$  given by $\left(x,{y}_{1},\dots ,{y}_{n}\right)↦x$  . This is a branched Galois cover with Galois group $Gal\left(N/M\right)×\left(\mathbb{Z}/2\mathbb{Z}{\right)}^{n-1}$  . There exists a Hilbert set of values $t\in M$  such that the geometric points lying over $x=\sqrt[n]{q}t$  in ${C}_{M}$  consists of a single $Gal\left(\overline{K}/M\right)$  -orbit or, equivalently, $Gal\left(M\left({y}_{1},\dots ,{y}_{n}\right)/M\right)\sim =\left(\mathbb{Z}/2\mathbb{Z}{\right)}^{n-1}$  and $M\left({y}_{1},\dots ,{y}_{n}\right)$  is linearly disjoint from $N$  over $M$  . As a Hilbert set of a finite extension of $L$  always contains some Hilbert set of $L$  ([5,Ch. 9, Prop. 3.3), it follows that there exists $t\in L$  such that setting $x=\sqrt[n]{q}t$  , relative to $M$  , the extension $M\left({y}_{1},\dots ,{y}_{n}\right)$  is linearly disjoint from $N$  and has Galois group $\left(\mathbb{Z}/2\mathbb{Z}{\right)}^{n-1}$  . We can therefore iteratively construct a sequence ${t}_{1},{t}_{2},\dots \in {L}^{*}$  such that the extensions
${M}_{i}:=M\left(\sqrt{\left(\sqrt[n]{q}{t}_{i}-a\right)\left(\sqrt[n]{q}{t}_{i}-b\right)\left(\sqrt[n]{q}{t}_{i}-\zeta a\right)\left(\sqrt[n]{q}{t}_{i}-\zeta b\right)},\dots ,\sqrt{\left(\sqrt[n]{q}{t}_{i}-{\zeta }^{n-1}a\right)\left(\sqrt[n]{q}{t}_{i}-{\zeta }^{n-1}b\right)\left(\sqrt[n]{q}{t}_{i}-a\right)\left(\sqrt[n]{q}{t}_{i}-b\right)}\right)$
are all linearly disjoint over $M$  . Let ${Q}_{i}$  be a point with $x$  -coordinate $\sqrt[n]{q}{t}_{i}$  , and ${R}_{i}$  the corresponding $\zeta$  -eigenvector of $\sigma$  given by ( 1 ). We claim that the ${R}_{i}$  span a space of infinite dimension. The ${Q}_{i}$  do so by [2,Lemma3.12, and as the ${\zeta }^{-j}$  are linearly independent over $\mathbb{Q}$  , it follows that the ${R}_{i}$  do so as well.
We conclude with a question that does not seem to be directly amenable to the methods of this paper:
Question 3.3. Does the set ${\cap }_{\zeta \in {\mathbb{C}}_{tor}^{*}}{\Sigma }_{\zeta }$  of elements of ${G}_{K}$  having generic spectrum on ${V}_{E}$  always have an interior point?
References

1. P. Griffiths and J. Harris: Principles of Algebraic Geometry, John Wiley & Sons, New York, 1978.
2. B. Im: Mordell-Weil groups and the rank over large fields of elliptic curves over large fields, arXiv: math.NT/0411533, to appear in Canadian J. Math.
3. B. Im: Infinite multiplicity of roots of unity of the Galois group in the representation on elliptic curves, preprint, 2003.
4. B. Im: Heegner points and Mordell-Weil groups of elliptic curves over large fields, preprint, arXiv: math.NT/0411534.
5. S. Lang: Fundamentals of Diophantine Geometry, Springer-Verlag, New York, 1983.
6. M. Larsen: Rigid lattices are Mordell-Weil, preprint, arXiv: math.NT/0401209.
7. M. Raynaud: Courbes sur une variété abélienne et points de torsion. Invent. Math. 71 (1983), no. 1, 207–233.

Department of Mathematics, University of Utah, Salt Lake City, Utah 84112, USA E-mail address : im@math.utah.edu Department of Mathematics, Indiana University, Bloomington, Indiana 47405, USA E-mail address : larsen@math.indiana.edu