Documentation

TauCeti.LowDimTopology.Geometrization

The geometrization conjecture #

Thurston's geometrization conjecture (Kirby's problem 3.45), proved by Perelman, describes every closed orientable irreducible 3-manifold M: it can be cut along finitely many disjoint incompressible tori into pieces each of which is Seifert fibred or hyperbolic of finite volume. This is the form in which Bessières, Besson, Boileau, Maillot and Porti state it. Thurston's original formulation instead asks that the pieces carry geometric structures of finite volume modelled on his eight geometries (TauCeti.HasGeometricStructure).

A geometric torus decomposition of M, TauCeti.IsGeometricTorusDecomposition T, is a finite family of maps T i : S¹ × S¹ → M that are

such that every connected component of the complement M \ ⋃ i, T i (S¹ × S¹) is Seifert fibred (TauCeti.IsSeifertFibered) or carries a complete hyperbolic metric of finite volume (TauCeti.IsFiniteVolumeHyperbolic). The components are the interiors of the pieces of M cut along the tori; they are open subsets of M, so they inherit its smooth structure, which is the one their hyperbolic metrics are smooth for. A family of tori is allowed to be empty: a closed Seifert-fibred or hyperbolic 3-manifold is its own single piece (TauCeti.isGeometricTorusDecomposition_of_isEmpty_of_isSeifertFibered for the Seifert case).

The decomposition is only asked to exist. The canonical choice, the JSJ decomposition, and the uniqueness of a minimal family of tori up to isotopy are separate theorems.

TauCeti.GeometrizationConjecture asserts that every closed orientable irreducible smooth 3-manifold has a geometric torus decomposition. Irreducibility is the condition TauCeti.IsSphereBoundsBall that every locally flat 2-sphere bounds a ball. A closed orientable prime 3-manifold is irreducible unless it is S¹ × S², which is Seifert fibred, so this also covers the prime manifolds. The conjecture is stated, not proved.

Main definitions #

Main results #

References #

theorem TauCeti.exists_opens_eq_connectedComponentIn_compl_iUnion_range {M : Type u_1} [TopologicalSpace M] [ChartedSpace (EuclideanSpace ℝ (Fin 3)) M] {ι : Type u_2} [T2Space M] [Finite ι] (T : ι → C(Circle × Circle, M)) (x : M) :
∃ (U : TopologicalSpace.Opens M), ↑U = connectedComponentIn (⋃ (i : ι), Set.range ⇑(T i))ᶜ x

For finitely many tori in a Hausdorff 3-manifold, every connected component of the complement of their images is open.

A family of maps T i : S¹ × S¹ → M is a geometric torus decomposition of the 3-manifold M when there are finitely many maps, they are disjoint locally flat incompressible embeddings of the torus, and every connected component of the complement of their images is Seifert fibred or hyperbolic of finite volume. For finitely many tori the components are open (TauCeti.exists_opens_eq_connectedComponentIn_compl_iUnion_range), so they are 3-manifolds with the structure inherited from M.

Instances For

    A connected Seifert-fibred 3-manifold has a geometric torus decomposition with no tori: it is its own single piece.

    The geometrization conjecture (Thurston, Kirby's problem 3.45; a theorem of Perelman): every closed orientable irreducible 3-manifold M, that is a compact connected orientable smooth 3-manifold without boundary in which every locally flat 2-sphere bounds a ball, has a geometric torus decomposition: finitely many disjoint locally flat incompressible tori whose complement has Seifert-fibred and finite-volume hyperbolic components.

    Instances For

      A proof of the geometrization conjecture gives every closed orientable irreducible 3-manifold a geometric torus decomposition.

      The defining characterization of the geometrization conjecture.