Finite covers of a hyperbolic 3-manifold and virtual fibers.



Similar documents
Surface bundles over S 1, the Thurston norm, and the Whitehead link

TOPOLOGY OF SINGULAR FIBERS OF GENERIC MAPS

On The Existence Of Flips

Math 231b Lecture 35. G. Quick

RIGIDITY OF HOLOMORPHIC MAPS BETWEEN FIBER SPACES

Fiber sums of genus 2 Lefschetz fibrations

FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES

LECTURE NOTES ON GENERALIZED HEEGAARD SPLITTINGS

1 if 1 x 0 1 if 0 x 1

6.2 Permutations continued

HYPERBOLIC GROUPS WHICH FIBER IN INFINITELY MANY WAYS

8.1 Examples, definitions, and basic properties

4. Expanding dynamical systems

Metric Spaces Joseph Muscat 2003 (Last revised May 2009)

On the existence of G-equivariant maps

How To Find Out How To Build An Elliptic Curve Over A Number Field

The Tangent Bundle. Jimmie Lawson Department of Mathematics Louisiana State University. Spring, 2006

A NOTE ON TRIVIAL FIBRATIONS

FIBER PRODUCTS AND ZARISKI SHEAVES

ON GALOIS REALIZATIONS OF THE 2-COVERABLE SYMMETRIC AND ALTERNATING GROUPS

Holomorphic motion of fiber Julia sets

Ideal Class Group and Units

A NEW CONSTRUCTION OF 6-MANIFOLDS

THE THURSTON METRIC ON HYPERBOLIC DOMAINS AND BOUNDARIES OF CONVEX HULLS

Notes on metric spaces

Rotation Rate of a Trajectory of an Algebraic Vector Field Around an Algebraic Curve

I. GROUPS: BASIC DEFINITIONS AND EXAMPLES

GEOMETRIC PROPERTIES OF PROJECTIVE MANIFOLDS OF SMALL DEGREE

SMOOTH CASE FOR T or Z

FIBRATION SEQUENCES AND PULLBACK SQUARES. Contents. 2. Connectivity and fiber sequences. 3

Surfaces of minimal complexity in low-dimensional topology

Singular fibers of stable maps and signatures of 4 manifolds

Introduction to Topology

Galois Theory III Splitting fields.

Cycles and clique-minors in expanders

11 Ideals Revisiting Z

THE MINIMAL GENUS PROBLEM

Quasi-static evolution and congested transport

( ) = ( ) = {,,, } β ( ), < 1 ( ) + ( ) = ( ) + ( )

THE RELATIVE REIDEMEISTER NUMBERS OF FIBER MAP PAIRS. Fernanda S. P. Cardona Peter N.-S. Wong. 1. Introduction

MA4001 Engineering Mathematics 1 Lecture 10 Limits and Continuity

Pro-p Groups with Few Normal Subgroups

1. (First passage/hitting times/gambler s ruin problem:) Suppose that X has a discrete state space and let i be a fixed state. Let

Symplectic structures on fiber bundles

ON FIBER DIAMETERS OF CONTINUOUS MAPS

Continued Fractions and the Euclidean Algorithm

FIXED POINT SETS OF FIBER-PRESERVING MAPS

1. Prove that the empty set is a subset of every set.

SOME PROPERTIES OF FIBER PRODUCT PRESERVING BUNDLE FUNCTORS

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES

Metric Spaces. Chapter Metrics

HOMEWORK 5 SOLUTIONS. n!f n (1) lim. ln x n! + xn x. 1 = G n 1 (x). (2) k + 1 n. (n 1)!

1. Introduction. PROPER HOLOMORPHIC MAPPINGS BETWEEN RIGID POLYNOMIAL DOMAINS IN C n+1

1 = (a 0 + b 0 α) (a m 1 + b m 1 α) 2. for certain elements a 0,..., a m 1, b 0,..., b m 1 of F. Multiplying out, we obtain

ON ISOMORPHISMS OF BERS FIBER SPACES

INDISTINGUISHABILITY OF ABSOLUTELY CONTINUOUS AND SINGULAR DISTRIBUTIONS

GENERIC COMPUTABILITY, TURING DEGREES, AND ASYMPTOTIC DENSITY

SINTESI DELLA TESI. Enriques-Kodaira classification of Complex Algebraic Surfaces

COMBINATORIAL PROPERTIES OF THE HIGMAN-SIMS GRAPH. 1. Introduction

NOTES ON CATEGORIES AND FUNCTORS

How To Prove The Dirichlet Unit Theorem

Introduction to Algebraic Geometry. Bézout s Theorem and Inflection Points

Reference: Introduction to Partial Differential Equations by G. Folland, 1995, Chap. 3.

Mathematical Research Letters 1, (1994) MAPPING CLASS GROUPS ARE AUTOMATIC. Lee Mosher

INVARIANT METRICS WITH NONNEGATIVE CURVATURE ON COMPACT LIE GROUPS

Sign changes of Hecke eigenvalues of Siegel cusp forms of degree 2

(0, 0) : order 1; (0, 1) : order 4; (0, 2) : order 2; (0, 3) : order 4; (1, 0) : order 2; (1, 1) : order 4; (1, 2) : order 2; (1, 3) : order 4.

Invariant Metrics with Nonnegative Curvature on Compact Lie Groups

10.2 Series and Convergence

Sets of Fibre Homotopy Classes and Twisted Order Parameter Spaces

SHINPEI BABA AND SUBHOJOY GUPTA

Geometrical Characterization of RN-operators between Locally Convex Vector Spaces

Notes V General Equilibrium: Positive Theory. 1 Walrasian Equilibrium and Excess Demand

Online publication date: 11 March 2010 PLEASE SCROLL DOWN FOR ARTICLE

Row Ideals and Fibers of Morphisms

88 CHAPTER 2. VECTOR FUNCTIONS. . First, we need to compute T (s). a By definition, r (s) T (s) = 1 a sin s a. sin s a, cos s a

CHAPTER 1 BASIC TOPOLOGY

Kyoto University. Note on the cohomology of finite cyclic coverings. Yasuhiro Hara and Daisuke Kishimoto

How To Know The Limit Set Of A Fuchsian Group On A Boundary Of A Hyperbolic Plane

arxiv: v3 [math.at] 10 May 2012

A CONSTRUCTION OF THE UNIVERSAL COVER AS A FIBER BUNDLE

GROUPS ACTING ON A SET

ON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING. 0. Introduction

G = G 0 > G 1 > > G k = {e}

ON THE CLASSIFICATION OF SURFACES OF GENERAL TYPE WITH p g = q = 2. Contents 0. Introduction 1 1. Product-Quotient Surfaces with p g = q = 2 3

Lectures 5-6: Taylor Series

Analytic cohomology groups in top degrees of Zariski open sets in P n

The Ideal Class Group

FACTORING POLYNOMIALS IN THE RING OF FORMAL POWER SERIES OVER Z

THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS

A Short Proof that Compact 2-Manifolds Can Be Triangulated

CONTINUED FRACTIONS AND PELL S EQUATION. Contents 1. Continued Fractions 1 2. Solution to Pell s Equation 9 References 12

BP-cohomology of mapping spaces from the classifying space of a torus to some p-torsion free space

Samuel Omoloye Ajala

How To Know If A Domain Is Unique In An Octempo (Euclidean) Or Not (Ecl)

minimal polyonomial Example

A RIGOROUS AND COMPLETED STATEMENT ON HELMHOLTZ THEOREM

NIELSEN METHODS AND GROUPS ACTING ON HYPERBOLIC SPACES

ADDITIVE GROUPS OF RINGS WITH IDENTITY

The cover SU(2) SO(3) and related topics

Transcription:

Claire Renard Institut de Mathématiques de Toulouse November 2nd 2011

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Some conjectures. Let M be a hyperbolic 3-manifold, connected, closed and oriented. Theorem (Kahn, Markovic) The fundamental group π 1 M contains a surface subgroup. Conjectures (1) (Virtually Haken.) There exists a finite cover M M containing an incompressible surface, i.e. an embedded surface T in M such that the map induced by the embedding ι : T M on fundamental groups ι : π 1 T π 1 M is injective. (2) (Virtually positive first Betti number.) There exists a finite cover M M with b 1 (M ) > 0. (3) () For each n N, there exists a finite cover M n M with b 1 (M n) n. (4) (Thurston.) There exists a finite cover M M which fibers over the circle S 1.

Definition A 3-manifold M is said to be virtually fibered if it admits a finite cover which fibers over the circle. An embedded surface S in M is a virtual fiber if there is a finite cover of M in which the preimage of S is a fiber. T {1/2} T {0} T {1} ϕ M = T I / (x,0) ~ ( ϕx,1) S M Question: Let M M be a finite cover of M. Find conditions for M to contain an embedded surface which is a fiber, or at least a virtual fiber?

Definition A 3-manifold M is said to be virtually fibered if it admits a finite cover which fibers over the circle. An embedded surface S in M is a virtual fiber if there is a finite cover of M in which the preimage of S is a fiber. T {1/2} T {0} T {1} ϕ M = T I / (x,0) ~ ( ϕx,1) S M Question: Let M M be a finite cover of M. Find conditions for M to contain an embedded surface which is a fiber, or at least a virtual fiber?

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Introduction. Theorem (1, main theorem.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k(ɛ, Vol(M)) such that: If M M is a cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g ln g < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M contains an embedded surface T of genus g(t ) g which is a virtual fiber. In particular, M virtually fibers over the circle and M is Haken. Conjecture ( ) The technical assumption (1) is not necessary. Remark If Vol(M) is fixed, lim ɛ 0 k(ɛ, Vol(M)) +. If ɛ is fixed, lim Vol(M) + k(ɛ, Vol(M)) +.

Ideas of the proof of the main theorem. Suppose that the ratio g ln g/ ln ln(d/q) is small enough. Proof in two steps.

First step: Construct an embedded long and thin product T [0, m] in M, satisfying the following properties. M T [0,m] < K(g) > r T 0 T 1 T 2 T m 1 T m The surface T is orientable and closed, with genus g(t ) g. The number m = m( d, g) is large. q The surfaces T j := T {j} have their diameters uniformly bounded from above by K = K (g). Two surfaces T j and T j+1 are at distance at least r > 0.

First step: Construct an embedded long and thin product T [0, m] in M, satisfying the following properties. M T [0,m] < K(g) > r T 0 T 1 T 2 T m 1 T m The surface T is orientable and closed, with genus g(t ) g. The number m = m( d, g) is large. q The surfaces T j := T {j} have their diameters uniformly bounded from above by K = K (g). Two surfaces T j and T j+1 are at distance at least r > 0.

Second step: Use this product to construct a virtual fibration of M. P 2 P 1 T T T 1 1 T T 0 2 3 Choose D, a Dirichlet fundamental polyhedron for M in H 3. For each surface T j, consider the pattern of fundamental domains P j that is the union of the fundamental domains meeting T j. Find two patterns: 1 disjoint from each other, 2 isometric to the same model pattern P, 3 containing parallel surfaces T 1 and T 1. Cut along T 1 and T 1, glue them together to get a finite fibered cover N of M.

Second step: Use this product to construct a virtual fibration of M. P 2 P 1 T T T 1 1 T T 0 2 3 Choose D, a Dirichlet fundamental polyhedron for M in H 3. For each surface T j, consider the pattern of fundamental domains P j that is the union of the fundamental domains meeting T j. Find two patterns: 1 disjoint from each other, 2 isometric to the same model pattern P, 3 containing parallel surfaces T 1 and T 1. Cut along T 1 and T 1, glue them together to get a finite fibered cover N of M.

Second step: Use this product to construct a virtual fibration of M. P 2 P 1 T T T 1 1 T T 0 2 3 Choose D, a Dirichlet fundamental polyhedron for M in H 3. For each surface T j, consider the pattern of fundamental domains P j that is the union of the fundamental domains meeting T j. Find two patterns: 1 disjoint from each other, 2 isometric to the same model pattern P, 3 containing parallel surfaces T 1 and T 1. Cut along T 1 and T 1, glue them together to get a finite fibered cover N of M.

Second step: Use this product to construct a virtual fibration of M. P 2 P 1 T T T 1 1 T T 0 2 3 Choose D, a Dirichlet fundamental polyhedron for M in H 3. For each surface T j, consider the pattern of fundamental domains P j that is the union of the fundamental domains meeting T j. Find two patterns: 1 disjoint from each other, 2 isometric to the same model pattern P, 3 containing parallel surfaces T 1 and T 1. Cut along T 1 and T 1, glue them together to get a finite fibered cover N of M.

Second step: Use this product to construct a virtual fibration of M. P 2 P 1 T T T 1 1 T T 0 2 3 Choose D, a Dirichlet fundamental polyhedron for M in H 3. For each surface T j, consider the pattern of fundamental domains P j that is the union of the fundamental domains meeting T j. Find two patterns: 1 disjoint from each other, 2 isometric to the same model pattern P, 3 containing parallel surfaces T 1 and T 1. Cut along T 1 and T 1, glue them together to get a finite fibered cover N of M.

P 2 P 1 T T T 1 1 T T 0 2 3 ϕ 1 ψ ϕ 2 P T

S 1 W F M T 1 T 1 S 1 T 1 N M

E E + E+ 1 1 E 2 2 P 2 x 0 γ 0 P 1 x 1 γ 1 x 2 γ 2 x 3 γ 3 x 4 x 5 γ 4 T T T 1 T T 0 1 2 3 ϕ 1 ψ ϕ 2 E E + P T

The regular case. Theorem (2, regular case.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k (ɛ, Vol(M)) such that: If M M is a regular cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g 2 < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M is a fiber bundle over the circle, and a fiber can be obtained from a component of F, possibly after some surgeries.

The regular case. Theorem (2, regular case.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k (ɛ, Vol(M)) such that: If M M is a regular cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g 2 < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M is a fiber bundle over the circle, and a fiber can be obtained from a component of F, possibly after some surgeries.

The regular case. Theorem (2, regular case.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k (ɛ, Vol(M)) such that: If M M is a regular cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g 2 < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M is a fiber bundle over the circle, and a fiber can be obtained from a component of F, possibly after some surgeries.

The regular case. Theorem (2, regular case.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k (ɛ, Vol(M)) such that: If M M is a regular cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g 2 < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M is a fiber bundle over the circle, and a fiber can be obtained from a component of F, possibly after some surgeries.

The regular case. Theorem (2, regular case.) Fix ɛ Inj(M)/2. There exists an explicit constant k = k (ɛ, Vol(M)) such that: If M M is a regular cover of finite degree d, with an embedded, closed, orientable, pseudo-minimal surface F, which splits M into q compression bodies C 1,..., C q with the following properties: 1 Every simple closed curve embedded in C j of length ɛ is nul-homotopic in C j. 2 k g 2 < ln ln d/q, with g = max j {g(c j )}. Then the finite cover M is a fiber bundle over the circle, and a fiber can be obtained from a component of F, possibly after some surgeries.

Application to Heegaard splittings. ϕ F F H1 H2 If F is a surface of genus at least 1, χ (F ) = 2g(F ) 2. Definition The Heegaard characteristic: χ h (M) = 2g(M) 2. The strong Heegaard characteristic: χ sh (M) = max F {χ (F )} where F is a strongly irreducible Heegaard surface for M. Remark If M M is a cover of finite degree d, χ h (M ) dχ h (M).

Application to Heegaard splittings. ϕ F F H1 H2 If F is a surface of genus at least 1, χ (F ) = 2g(F ) 2. Definition The Heegaard characteristic: χ h (M) = 2g(M) 2. The strong Heegaard characteristic: χ sh (M) = max F {χ (F )} where F is a strongly irreducible Heegaard surface for M. Remark If M M is a cover of finite degree d, χ h (M ) dχ h (M).

Application to Heegaard splittings. ϕ F F H1 H2 If F is a surface of genus at least 1, χ (F ) = 2g(F ) 2. Definition The Heegaard characteristic: χ h (M) = 2g(M) 2. The strong Heegaard characteristic: χ sh (M) = max F {χ (F )} where F is a strongly irreducible Heegaard surface for M. Remark If M M is a cover of finite degree d, χ h (M ) dχ h (M).

Heegaard gradient and conjectures of Lackenby. Definition (Lackenby) Heegaard gradient: h (M) = inf i { } χ h (M i ). d i Strong Heegaard gradient: sh (M) = inf i { χ sh (M i ) d i }. Conjecture (Lackenby) (1) The Heegaard gradient of M is zero if and only if M virtually fibers over the circle. (2) The strong Heegaard gradient is always strictly positive.

Heegaard gradient and conjectures of Lackenby. Definition (Lackenby) Heegaard gradient: h (M) = inf i { } χ h (M i ). d i Strong Heegaard gradient: sh (M) = inf i { χ sh (M i ) d i }. Conjecture (Lackenby) (1) The Heegaard gradient of M is zero if and only if M virtually fibers over the circle. (2) The strong Heegaard gradient is always strictly positive.

The sub-logarithmic version is true. Definition Let η (0, 1). η-sub-logarithmic Heegaard gradient: { } χ h h (M i ) log,η (M) = inf i (ln ln d i ) η. Strong η-sub-logarithmic Heegaard gradient: { } χ sh sh (M i ) log,η (M) = inf i (ln ln d i ) η. Proposition (3, Sub-logarithmic version of Lackenby s conjectures.) Suppose conjecture ( ) is true. Let η (0, 1). (1) The η-sub-logarithmic Heegaard gradient of M is zero if and only if M virtually fibers over the circle. (2) The strong η-sub-logarithmic Heegaard gradient is always strictly positive.

The sub-logarithmic version is true. Definition Let η (0, 1). η-sub-logarithmic Heegaard gradient: { } χ h h (M i ) log,η (M) = inf i (ln ln d i ) η. Strong η-sub-logarithmic Heegaard gradient: { } χ sh sh (M i ) log,η (M) = inf i (ln ln d i ) η. Proposition (3, Sub-logarithmic version of Lackenby s conjectures.) Suppose conjecture ( ) is true. Let η (0, 1). (1) The η-sub-logarithmic Heegaard gradient of M is zero if and only if M virtually fibers over the circle. (2) The strong η-sub-logarithmic Heegaard gradient is always strictly positive.

The homological viewpoint. Let α H 1 (M, Z) be a non-trivial element. Definition A α -minimizing surface R is an embedded surface with homology class Poincaré-dual to α, and minimizing Thurston s norm: χ (R) = α. Question: Find conditions to ensure that R is the fiber of a fibration over the circle?

The homological viewpoint. Let α H 1 (M, Z) be a non-trivial element. Definition A α -minimizing surface R is an embedded surface with homology class Poincaré-dual to α, and minimizing Thurston s norm: χ (R) = α. Question: Find conditions to ensure that R is the fiber of a fibration over the circle?

The homological viewpoint. Let α H 1 (M, Z) be a non-trivial element. Definition A α -minimizing surface R is an embedded surface with homology class Poincaré-dual to α, and minimizing Thurston s norm: χ (R) = α. Question: Find conditions to ensure that R is the fiber of a fibration over the circle?

Circular decompositions. Definition Let M R be the 3-manifold obtained from M by removing a regular neighborhood of R diffeomorphic to R ( 1, 1). The circular characteristic of α, denoted by χ c (α), is the minimum over all α -minimizing surfaces R of the Heegaard characteristic of the cobordism (M R, R {1}, R { 1}). M M R S 1 S R {1} R R { 1} f 1~( 1) Remark χ c (α) = α + h(α), where h(α) is the minimum over all α -minimizing surfaces R of the minimal number of critical points of index 1 and 2 of a Morse function M R [ 1, 1].

Circular decompositions. Definition Let M R be the 3-manifold obtained from M by removing a regular neighborhood of R diffeomorphic to R ( 1, 1). The circular characteristic of α, denoted by χ c (α), is the minimum over all α -minimizing surfaces R of the Heegaard characteristic of the cobordism (M R, R {1}, R { 1}). M M R S 1 S R {1} R R { 1} f 1~( 1) Remark χ c (α) = α + h(α), where h(α) is the minimum over all α -minimizing surfaces R of the minimal number of critical points of index 1 and 2 of a Morse function M R [ 1, 1].

Applications to circular decompositions. Theorem (4, Adapted from a result of Lackenby) There exists an explicit constant l = l (ɛ, Vol(M)) such that: Fix α H 1 (M) a non-trivial cohomology class and R a α -minimizing surface. Let M M be a d-sheeted regular cover and α H 1 (M, Z) the Poincaré-dual class associated to a connected component R of the preimage of R in M. If l χ c (α ) 4 d, then the manifold M fibers over the circle and the surface R is a fiber.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Question: Find a tower of finite covers... M i+1 M i... M such that lim i + b 1 (M i ) = +? Theorem (5) Suppose that there exists an infinite tower... M i+1 N i+1 M i N i... N 1 M of finite covers of M such that for all i 1, M i N i is regular, with Galois group H i (Z/ 2Z ) r i. If inf i N χ h (M i )[π 1 M : π 1 N i ]/( 2) r i = 0, Then lim i + b 1 (M i ) = +. Corollary (6) Let... M i M i 1... M 1 M be the tower of finite covers corresponding to the lower mod 2 central series. For all i 1, set r i = b 1,F2 (M i 1 ), with M 0 = M, and R i = r 1 + r 2 +... + r i. Suppose that r 1 4. If inf i N χ h (M i )2 R i 1 /( 2) r i = 0, Then lim i + b 1 (M i ) = +.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Link with results of Lackenby. Theorem (Lackenby) Let... M i M i 1... M 1 M a tower of finite covers of M such that for all i 1, M i M i 1 is regular, of group isomorphic to (Z/ 2Z ) r i (with the convention that M 0 = M). Set R i = r 1 + r 2 +... + r i. Suppose that one of the following assumptions is satisfied: (a) π 1 M does not have property (τ) with respect to the family {π 1 M i } i N (for example r if lim i + h(m i ) = 0) and inf i+1 i N > 0, or 2 R i r (b) each cover M i M is regular and lim i+1 i + = +. 2 R i Then π 1 M is virtually large. In particular, the first Betti number of M is virtually infinite. Corollary (7) Let... M i M i 1... M 1 M be an infinite tower of finite covers of M as in the previous theorem. Then the first Betti number of M is virtually infinite if neither of the following properties is satisfied. 1 inf i N h(m i ) > 0 and the sequence ( r i+1 ) 2 R i i N admits a bounded subsequence. 2 inf i N h(m i+1 )4 R i ( 2) r i+1 > 0 and inf i N r i+1 2 R i = 0.

Thank you for your attention!!!! Merci pour votre attention!!!!