Rational Points on Curves

Similar documents
Geometry of points over Q of small height Part II

MANIN-MUMFORD AND LATTÉS MAPS

Rational Points on Curves of Genus 2

Vojta s conjecture and level structures on abelian varieties

Introduction to Arithmetic Geometry

Bjorn Poonen. MSRI Introductory Workshop on Rational and Integral Points on Higher-dimensional Varieties. January 17, 2006

MORDELL-LANG CONJECTURE

A reduction of the Batyrev-Manin Conjecture for Kummer Surfaces

Some. Manin-Mumford. Problems

CYCLES OF QUADRATIC POLYNOMIALS AND RATIONAL POINTS ON A GENUS 2 CURVE

Geometry dictates arithmetic

Algebraic Geometry: Elliptic Curves and 2 Theorems

Di erentially Closed Fields

Dynamics and Canonical Heights on K3 Surfaces with Noncommuting Involutions Joseph H. Silverman

On symmetric square values of quadratic polynomials

Jean-Pierre Serre. Lectures on the. Mordell-Weil Theorem

LECTURE 2 FRANZ LEMMERMEYER

MORDELL-LANG PLUS BOGOMOLOV. 1. Introduction Let k be a number field. Let A be a semiabelian variety over k; i.e., an extension

Proof of the Shafarevich conjecture

Irrationality of 2 and Arakelov Geometry

Structure of elliptic curves and addition laws

Elliptic curves over function fields 1

On the distribution of rational points on certain Kummer surfaces

Modern Number Theory: Rank of Elliptic Curves

LECTURE 22, WEDNESDAY in lowest terms by H(x) = max{ p, q } and proved. Last time, we defined the height of a rational number x = p q

Reciprocity laws and integral solutions of polynomial equations

this to include the explicit maps, please do so!

On the Torsion Subgroup of an Elliptic Curve

The Birch & Swinnerton-Dyer conjecture. Karl Rubin MSRI, January

Riemann surfaces with extra automorphisms and endomorphism rings of their Jacobians

Heights. Lectures by Fabien Pazuki. Spring semester 2014

Isogeny invariance of the BSD conjecture

Computations with Coleman integrals

Arithmetic of Hyperelliptic Curves

Rational Points on Curves in Practice. Michael Stoll Universität Bayreuth Journées Algophantiennes Bordelaises Université de Bordeaux June 8, 2017

Abstracts of papers. Amod Agashe

Higher genus curves and the inverse Galois problem

Why Should I Care About Elliptic Curves?

A p-adic Birch and Swinnerton-Dyer conjecture for modular abelian varieties

Introduction to Elliptic Curves

Galois Theory of Several Variables

Introduction to Arithmetic Geometry Fall 2013 Lecture #23 11/26/2013

Arithmetic applications of Prym varieties in low genus. Nils Bruin (Simon Fraser University), Tübingen, September 28, 2018

l-adic Representations

On Siegel s lemma outside of a union of varieties. Lenny Fukshansky Claremont McKenna College & IHES

MA 254 notes: Diophantine Geometry

Elliptic curves, Néron models, and duality

FIVE SQUARES IN ARITHMETIC PROGRESSION OVER QUADRATIC FIELDS

Outline of the Seminar Topics on elliptic curves Saarbrücken,

The Sato-Tate conjecture for abelian varieties

MORDELL EXCEPTIONAL LOCUS FOR SUBVARIETIES OF THE ADDITIVE GROUP

The average size of the 2-Selmer group of Jacobians of hyperelliptic curves having a rational Weierstrass point

Height Functions. Michael Tepper February 14, 2006

Torsion points on modular curves and Galois Theory

THE MANIN-MUMFORD CONJECTURE: A BRIEF SURVEY

Elliptic Curves: An Introduction

0.1 Valuations on a number field

TORSION POINTS ON CURVES

On Mordell-Lang in Algebraic Groups of Unipotent Rank 1

Lecture 2: Elliptic curves

Curves with many points Noam D. Elkies

arxiv: v2 [math.nt] 23 Sep 2011

FINITE GROUPS AND EQUATIONS OVER FINITE FIELDS A PROBLEM SET FOR ARIZONA WINTER SCHOOL 2016

Computing a Lower Bound for the Canonical Height on Elliptic Curves over Q

Hanoi lectures on the arithmetic of hyperelliptic curves

DIOPHANTINE EQUATIONS AND DIOPHANTINE APPROXIMATION

A short proof of Klyachko s theorem about rational algebraic tori

ABC Triples in Families

Algorithmic Number Theory in Function Fields

COMPLEX MULTIPLICATION: LECTURE 15

Topics in Number Theory: Elliptic Curves

A p-adic Birch and Swinnerton-Dyer conjecture for modular abelian varieties

The ABC Conjecture and its Consequences on Curves

GALOIS THEORY AND TORSION POINTS ON CURVES

Algebraic dynamics, function fields and descent

TORSION POINTS ON CURVES. Andrew Granville Université de Montréal Zeev Rudnick Tel-Aviv University. 1. Introduction

Cover Page. The handle holds various files of this Leiden University dissertation.

Infinite rank of elliptic curves over Q ab and quadratic twists with positive rank

Elliptic Curves and Mordell s Theorem

UNIFORM BOUNDS FOR THE NUMBER OF RATIONAL POINTS ON HYPERELLIPTIC CURVES OF SMALL MORDELL-WEIL RANK

BSD and the Gross-Zagier Formula

p-jets and Uniform Unramified Manin-Mumford

LARGE TORSION SUBGROUPS OF SPLIT JACOBIANS OF CURVES OF GENUS TWO OR THREE

How to compute regulators using Arakelov intersection theory

The arithmetic of Drinfeld modules. Dragos Florin Ghioca. M.S. (University of Pittsburgh) 2000

Introduction to Arithmetic Geometry Fall 2013 Lecture #17 11/05/2013

A Generalization of Wilson s Theorem

CONICS - A POOR MAN S ELLIPTIC CURVES arxiv:math/ v1 [math.nt] 18 Nov 2003

Diophantine geometry and non-abelian duality

Equations for Hilbert modular surfaces

The hyperbolic Ax-Lindemann-Weierstraß conjecture

Pseudo-abelian varieties

Elliptic Curves. Akhil Mathew (Department of Mathematics Drew UniversityElliptic MathCurves 155, Professor Alan Candiotti) 10 Dec.

Lecture Notes on Arithmetic Dynamics Arizona Winter School March 13 17, 2010

Some applications of p-adic uniformization to algebraic dynamics

Does There Exist an Elliptic Curve E/Q with Mordell-Weil Group Z 2 Z 8 Z 4?

On Hrushovski s proof of the Manin-Mumford conjecture

ON THE KERNELS OF THE PRO-l OUTER GALOIS REPRESENTATIONS ASSOCIATED TO HYPERBOLIC CURVES OVER NUMBER FIELDS

TORSION AND TAMAGAWA NUMBERS

Elliptic Curves, Factorization, and Cryptography

Transcription:

Rational Points on Curves 1

Q algebraic closure of Q. k a number field. Ambient variety: Elliptic curve E an elliptic curve defined over Q. E N = E E. In general A an abelian variety defined over Q. For A defined over k, we denote by A(k) the k-rational points of A, (all coordinates in k). We identify A = A(Q). 2

Tori G m = Q the multiplicative group of Q G n m = Q Q, Torsion subgroup of G n m Tor Gm = Roots of unity = {ζ G m : N N,ζ N = 1} Tor G n m = {(ζ 1,,ζ n ) G n m : ζ i Tor Gm } Finitely generated subgroup Γ of G n m 3

Algebraic subgroups of G n m An algebraic subgroup B of dimension n s is the kernel of a matrix φ B Mat s,n (Z) of rank s b 11... b 1n φ B =... : G n m G s m b s1... b sn φ B : (x 1,...,x n ) (x b 11 1... x b 1n n,...,x b s1 1... x b sn n ). Up to constants, degb is the maximal Minor of φ B. Choosing φ B appropriately, degb is up to constant b 1... b s. Remark There are only finitely many algebraic subgroups of bounded degree. Tor G n m = dimb=0 B. dimb=0 B dimb 1 B dimb n 1 B. 4

Cuves in Tori Let C be an algebraic curve embedded in G n m. Questions When is C Tor G n m finite? When is C Γ finite? When is C dimb r B finite? Note that if C = G m ζ 2 ζ n then none of the above sets is finite!! These are special cases of the toric Manin-Mumford Conjecture, Mordell-Lang Conjecture, Torsion Anomalous Conjecture 5

Elliptic Case Let E be an elliptic curve. Consider E N for some integer N. Torsion subgroups Tor E = {P E : m N + with mp = 0} Tor E N = (Tor E ) N Finitely generated subgroup Γ of E N Theorem (Mordell-Weil Theorem) Let k be a number field and let E be defined over k. Then E(k) is finitely generated. Let A be an abelian variety defined over k. Then A(k) is finitely generated. 6

Algebraic subgroups of E N An algebraic subgroup B of dimension n s is the kernel of a matrix φ B Mat s,n (End(E)) of rank s b 11... b 1,n φ B =... : E N E s b s,1... b s,n φ B : (x 1,...,x N ) (b 11 x 1 +... + b 1N x N,...,b s1 x 1 +... + b sn x N ). Up to constants, degb is the square of the maximal Minor of φ B. Choosing φ B appropriately, degb is up to constants b 1 2... b s 2. Remark There are only finitely many algebraic subgroups of bounded degree. Tor E N = dimb=0 B. dimb=0 B dimb 1 B dimb n 1 B. 7

Cuves in E N Let C be an algebraic curve embedded in E N. Questions When is C Tor E N finite? When is C(k) finite? When is C dimb r B finite? Note that if C = E + t with t a torsion point, then none of the above sets is finite!! These are special cases of the Manin-Mumford Conjecture, Mordell-Lang Conjecture, Torsion Anomalous Conjecture 8

Some Classical results on Curves Let C P 2 be a curve defined by a homogeneous polynomial with coefficients in k. P(x,y,z) = 0 Geometry Arithmetic C(k) Quadrics C(k) = /0 or g(c) = 0 degp 2 C(k) infinite Elliptic C(k) = curves degp = 3 Z r Z/n 1 Z Z/n 2 Z g(c) = 1 non-singular (Mordell-Weil 1922) C(k) g(c) 2 degp 4 finite non-singular (Faltings 1983) 9

Effectivity The result of Faltings is not effective, in the sense that it does not give any method for finding the points in C(k). This is due to the non existence of an effective bound for the height of the points in C(k). 10

Height Function is a measure of the complexity of the coordinates of a point. h : E N (Q) R + On P N (Q) we consider the logarithmic Weil height. On an abelian variety we consider the canonical Néron-Tate height. This is the square of a norm. Theorem (Northcott Theorem) A set of bounded (Néron-Tate) height and bounded degree is finite. 11

Effective Methods The method of Chabauty-Coleman provides a bound on the number of rational points on curves with Jacobian of Q-rank strictly smaller than the genus. Example Flynn gives explicit examples: he finds the rational points for a selection of curves of genus 2 with Jacobian of Q-rank 1. 12

The Manin-Dem janenko method applies to curves that admit many Q-independent morphisms towards an abelian variety. Example Kulesz, Girard, Matera, Schost and others find all rational points on some families of curves: these curves have genus 2 (resp. 3) and elliptic Jacobian of Q-rank 1 (resp. 2) with some special properties. For instance with factors given by a Weierstrass equation y 2 = x 3 + a 2 x, with a square-free integer and such that the Mordell-Weil group has rank 1. No explicit height s bound The bounds for the height must be worked out with ad hoc methods case by case and for the technique to be successful the equations of the curve must be of a special shape and small genus. 13

Joint work with S. Checcoli and F. Veneziano Let E be an elliptic curve given in the form y 2 = x 3 + Ax + B. with A, B algebraic integers. Let ĥ be the Néron-Tate height on E N. Let h(c) be the normalised height of C. 14

Theorem Let E be an elliptic curve of Q-rank 1. Let C E N be a curve of genus at least 2. Then P C(Q) has height bounded as ĥ(p) 4 3 N 2 N!degC ( C 1 h(c)(degc) + 4C 1 c 1 (degc) 2 + 2c 1 ). Moreover if N = 2 ĥ(p) C 1 h(c)degc + 4C 1 c 1 (degc) 2 + 4c 1 C 1 = 145 c 1 = c 1 (E) = 2h W (A) + 2h W (B) + 4, 15

Compared with previous bounds In a previous work, for C not contained in any translate of a subgroup of E N with N 3 then ĥ(p) B 1 (N)4(N 1)C 1 h(c)(degc) N 1 + B 2 (N)(N 1)C 2 (degc) N + + N 2 C 3 where B 1 (N) B 2 (N) 10 27 N N2 (N!) N. While here ĥ(p) 4(N 1)C 1 h(c)degc + (N 1)C 2 (degc) 2 + N 2 C 3, 16

Explicit Bounds Assume that E is without CM, defined over a number field k and that E(k) has rank 1. Let (x 1,y 1 ) (x 2,y 2 ) be the affine coordinates of E 2 Corollary Let C be the curve given in E 2 cut by the additional equation p(x 1 ) = y 2, with p(x) k[x] a non-constant polynomial of degree n. Then for P C(k) we have ĥ(p) 2595(h W (p) + logn + 4c 1 (E))(2n + 3) 2 + 4c 1 (E) where h W (p) = h W (1 : p 0 :... : p n ) is the height of the coefficients of p(x) and c 1 = 2log(3 + A + B ) + 4. 17

C is transverse. Moreover degc = 6n + 9 and h(c) 6(2n + 3)(h W (p) + logn + 2c 1 (E)) For almost all E, the C n have genus 4n + 2 and the D n have increasing genus. 18

Explicit Examples Definition Let {C n } n be the family of curves C n E 2 defined for n 1 via the additional equation x n 1 = y 2. Let {D n } n be the family of curves D n E 2 defined for n 1 via the additional equation Φ n (x 1 ) = y 2, where Φ n (x) is the n-th cyclotomic polynomial. 19

Choice of the ambient variety: E 1 : y 2 = x 3 + x 1, E 2 : y 2 = x 3 26811x 7320618, E 3 : y 2 = x 3 675243x 213578586, E 4 : y 2 = x 3 110038419x + 12067837188462, E 5 : y 2 = x 3 2581990371x 50433763600098. These are five elliptic curves without CM and of rank 1 over Q. 20

y1 2 = x1 3 + x 1 1 E 1 C n = y2 2 = x2 3 + x 2 1 E 1 n x 1 = y 2 y1 2 = x1 3 + x 1 1 E 1 D n = y2 2 = x2 3 + x 2 1 E 1 Φ n (x 1 ) = y 2 The C n have genus 4n + 2 and the D n have increasing genus. 21

degc n = 6n + 9, h(c n ) 6(2n + 3)log(3 + A + B ). degd n = 6ϕ(n) + 9, h(d n ) 6(2ϕ(n) + 3) (2 ω 2(n) log2 + 2log(3 + A + B ) ), where ϕ(n) is the Euler function, ω 2 (n) is the number of distinct odd prime factors of n. 22

For every n 1 and every point P C n (Q) we have ĥ(p) 1301(4c 6 (E))(2n + 3) 2 + 4c 6 (E). For every n 2 and every point P D n (Q) we have ( ) ĥ(p) 1302 2 ω 2(n) log2 + 4c6 (E) (2ϕ(n) + 3) 2 + 4c 6 (E) were c 6 (E) = log(3 + A + B ). 23

Explit Examples Example For the curves C n E 1 E 1 we have C n (Q) = {(1,±1) (1,1)}. For the curves C n E i E i with i = 2,3,4,5, we have C n (Q) =. 24

Example For i = 2,3,4,5, the curves D n E i E i have For the curves D n E 1 E 1 we have D n (Q) =. D 1 (Q) = (2,±3) (1,1) D 2 (Q) = (2,±3) (2,3) D 3 k (Q) = (1,±1) (2,3) D 47 k (Q) = (1,±1) (13,47) D p k (Q) = if p 3,47 or p = 2 and k > 1 D 6 (Q) = (1,±1) (1,1) and (2,±3) (2,3) D n (Q) = (1,±1) (1,1) if n 6 has at least two distinct prime factors. 25

Theorem (Manin-Mumford Conjecture) Raynaud 1983 Let A be an abelian variety and Tor A its torsion. Let C A be a curve of genus 2. Then, C Tor A is finite. Theorem ( Mordell-Lang Conjecture) Faltings 1983, Vojta 1996, Hindry 1988 Let C A of genus at least 2. Let Γ be a subgroup of A of finite rank. Then C Γ is finite. Mordell-Lang Conjecture implies Manin-Mumford Conjecture Mordell-Lang Conjecture + Mordell-Weil Theorem imply C(k) is non dense. 26

Theorem (Torsion Anomalous Conjecture) Let C be weak-transverse in A. Then the set C dimb N 2 B is finite. Here B ranges over all algebraic subgroups of dimension N 2. Show that C dimb N 2 B has bounded height. C dimb N 2 B has bounded degree. Remark This theorem implies the Mordell-Lang conjecture for curves. 27

Effective/Explicit Methods For Curves The proof of the theorem is effective only if C is transverse in E N (i.e. not contained in any translate of an algebraic subgroup of E N ) (Use the geometry of numbers and properties of the height) In principle effective For N 3 and C weak-transverse in E N, E with CM and C dimb=1 B. (Use a Lehmer type bound) explicit For N 2 and C in E N of genus at least 2, E without CM and C dimb=1 B. IMPLEMENTABLE FOR N = 2 explicit (preprint 2016) For N 2 and C transverse in E N and C dimb N 1 B. This explicit result implies new cases of the Effective Mordell Conjecture. Let C be transverse in E N and rank of E(k) N 1. Then C(k) has height bounded by an explicit constant. 28

Normalized Height The height of a subvariety V of A is its normalized height. The normalized height of an algebraic subgroup is 0. Theorem (Arithmetic Bézout Theorem) Let V and W be irreducible subvarieties of P m. Let Z 1,...,Z n be the irreducible components of V W. Then n i=1 h(z i ) degv h(w) + degw h(v) + c(m)degv degw. 29

Essential Minimum Definition: h : V R + 0 } {{ } µ(v) } {{ } Essential Minimum h 1 non dense h 1 dense µ(v) = sup{ε : h 1 [0,ε) non dense in V} Theorem (Zhang Inequality) Let X be an irreducible subvariety of P m, then 1 h(x) (1 + dimx) degx µ(x) h(x) degx. 30