Polynomial bounds for surfaces in cusped 3-manifolds

Basic Information

• Paper ID: 2311.08567
• Title: Polynomial bounds for surfaces in cusped 3-manifolds
• Authors: Jessica S. Purcell, Anastasiia Tsvietkova
• Classification: math.GT (Geometric Topology)
• Publication Date: November 2023 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2311.08567

Abstract

Studying the number of homotopy classes of embedded essential surfaces in a given 3-manifold is a natural question. Early bounds obtained using normal surface theory are exponential. Recently, Hass, Thompson, and Tsvietkova obtained polynomial bounds for finite classes of surfaces (closed and spanning surfaces) by restricting to alternating link complements in the 3-sphere. This paper completes the full picture for classical alternating links and extends the results to other classes of cusped 3-manifolds. The article provides explicit polynomial bounds for all embedded essential surfaces, including closed or arbitrary boundary slopes, orientable or non-orientable surfaces. The 3-manifolds studied are link complements with alternating diagrams in a broad family of 3-manifolds, including all alternating links in the 3-sphere and many non-alternating links, alternating virtual knots, many torus alternating knots, and most Dehn fillings of these manifolds.

Research Background and Motivation

Importance of the Problem

1. Fundamental Question: How many homotopy classes of embedded essential surfaces exist in a given 3-manifold? This is a fundamental question in 3-manifold topology.
2. Geometric Significance: The existence or non-existence of essential surfaces has important implications for the geometrization of 3-manifolds, following Thurston's work.
3. Theoretical Foundation: According to work by Kahn-Markovic, Haglund-Wise, and Agol, every 3-manifold has a finite cover containing embedded essential surfaces.

Limitations of Existing Methods

1. Exponential Bounds: Using triangulations of 3-manifolds and normal surface theory can bound the number of homotopy classes of fundamental normal surfaces, but the resulting bounds are exponential towers in genus.
2. Limited Results: Hass, Thompson, and Tsvietkova obtained polynomial bounds in alternating link complements, but only for closed and spanning surfaces.
3. Open Problems: Even in the alternating setting, counting surfaces of arbitrary topological type remains an open problem.

Research Motivation

This paper aims to complete the full counting program for alternating links and extend to a broader family of cusped 3-manifolds by developing more general tools.

Core Contributions

1. Completion for Classical Alternating Links: Provides polynomial bounds for all essential surfaces (orientable or non-orientable, with or without boundary) in alternating link complements in the 3-sphere.
2. Extension to Generalized Alternating Links: Generalizes results to weakly generalized alternating links, including virtual alternating knots, torus alternating knots, etc.
3. Explicit Effective Bounds: The provided bounds are universal, explicit, and effective, depending only on the crossing number of the diagram and the Euler characteristic of the surface.
4. Dehn Filling Applications: Extends bounds to 3-manifolds obtained by Dehn fillings of these link complements.

Detailed Methodology

Problem Setup

Given a weakly generalized alternating link L in a 3-manifold Y, compute the number of ways a connected surface Z of fixed topological type can be embedded as an essential surface in Y-N(L) (up to homotopy).

Core Technical Framework

1. Weakly Generalized Alternating Links

Defining Conditions:

• π(L) is alternating on Π
• π(L) is weakly prime
• π(L) intersects each component of the projection surface Π
• Each component of π(L) projects to at least one crossing
• π(L) is checkerboard colorable
• Representation degree r(π(L),Π) ≥ 4

2. Block Decomposition

Decompose the link complement Y-N(L) into blocks:

• Cut along two checkerboard-colored surfaces (white and shaded surfaces)
• Obtain connected components homeomorphic to Y-N(Π) in the interior
• Crossing arcs become ideal edges, link strands become ideal vertices

3. Normal Surface Theory

Place the essential surface Z in normal form relative to the block decomposition:

• Z = ∪ᵢZᵢ, where Zᵢ are normal subsurfaces in blocks
• Each ∂Zᵢ boundary curve is labeled with letters S, B, P
• Use combinatorial area to control the number of subsurfaces

4. Combinatorial Area

Define the combinatorial area of subsurface Zⱼ:

a(Zⱼ) = (π/2)(#S) + (π/2)(#B) + π(#P) - 2πχ(Zⱼ)

satisfying the Gauss-Bonnet formula: a(S) = -2πχ(Z)

Technical Innovations

1. Unified Framework: Combines Menasco's standard form techniques with normal forms from Lackenby, Futer-Guéritaud, and Howie-Purcell.
2. Generalized Setting: Extends from classical alternating links on S² to alternating diagrams on arbitrary closed orientable surfaces.
3. Subsurface Counting:
   • Uses combinatorial area to bound the number of non-zero area subsurfaces
   • Specially handles zero-area subsurfaces (BBBB and BBSS disks)
4. Boundary Curve Counting: Counts possible curves on block boundaries through combinatorial methods.

Main Results

Theorem 6.4 (Classical Alternating Links)

Let π(L) be a prime alternating projection of link L on S² in S³ with n crossings. Let Z be a connected topological surface with Euler characteristic χ(Z). The number of ways Z can be embedded as an essential surface in S³-N(L) is at most:

(6n)^(80χ(Z)²) · 2^(2-4χ(Z)+2)

Theorem 7.5 (Generalized Case)

Let L be a weakly generalized alternating link in 3-manifold Y with alternating diagram π(L) having n crossings. Assume that in each component Σ of Y-N(Π), there is a universal bound X on the number of homotopy classes of incompressible surfaces of fixed genus and fixed boundary curves. Then the number of essential embeddings of fixed topological surface Z is at most:

(2X(g+1))^(-4χ+2) · (6n)^(-800χ³+80χ²)

Corollary 8.6 (Virtual Alternating Knots)

For a weakly prime alternating projection of virtual link L, the number of essential, meridian-incompressible embeddings of connected orientable topological surface Z is at most:

(2(g+1))^(-4χ) · (6n)^(-800χ³+80χ²)

Proof Strategy

1. Subsurface Counting (Lemma 4.1)

For meridian-incompressible surface Z, the number of normal subsurfaces Zᵢ that are not BBBB or BBSS disks is at most -4χ(Z).

Proof Outline:

• Use combinatorial area and Gauss-Bonnet formula
• Negative Euler characteristic subsurfaces contribute at least 2π area
• Zero or positive Euler characteristic subsurfaces contribute at least π/2 area (unless zero area)
• Zero area cases are excluded by the assumption

2. Boundary Curve Counting (Lemma 5.4)

The number of curves on block boundaries that can be boundary components of some subsurface Zᵢ is at most:

C(n,χ(Z)) = (6n)^(-20χ(Z))

Proof Outline:

• Each curve is determined by its intersection pattern with block boundaries
• At most 6n possible intersection choices on Π⁺
• Each boundary component has at most -20χ(Z) intersections

3. Meridian Compression Handling (Lemma 3.1)

Through meridian compression, any essential surface can be converted to a meridian-incompressible surface with at most -4χ(Z)+2 additional boundary components.

Experimental Verification

This is primarily a theoretical work without numerical experiments. However, it provides multiple concrete applications:

Application Examples

1. Classical Alternating Knots: Completes surface counting for alternating link complements in S³
2. Virtual Alternating Knots: Provides bounds for virtual alternating knots in thickened surfaces
3. Torus Alternating Knots: Handles alternating diagrams on Heegaard tori
4. Dehn Fillings: Extends to closed 3-manifolds obtained by fillings

Quality of Bounds

• Authors acknowledge that bounds may be far from optimal
• Goal is to establish polynomial bounds rather than optimal bounds
• Conservative estimates used at multiple steps to maintain polynomial order

Related Work

Historical Development

1. Normal Surface Theory: Exponential bounds by Matveev, Jaco-Oertel, Hass-Lagarias-Pippenger
2. Specialized Results for Alternating Knots: Polynomial bounds by Hass-Thompson-Tsvietkova (closed and spanning surfaces)
3. Immersed Surfaces: Exponential bounds by Masters, Kahn-Markovic in hyperbolic 3-manifolds
4. Algorithmic Approaches: Quasi-polynomial bounds by Dunfield-Garoufalidis-Rubinstein

Innovations in This Paper

• First polynomial bounds for all essential surfaces in alternating setting
• Completely different methodology from previous work
• Bounds are explicit, effective, and computable

Conclusions and Discussion

Main Conclusions

1. Completes polynomial counting of essential surfaces in classical alternating links
2. Generalizes results to a broad family of weakly generalized alternating links
3. Provides explicit, universal polynomial bounds

Limitations

1. Optimality of Bounds: Acknowledges that bounds may be far from optimal
2. Technical Assumptions: Requires multiple technical assumptions (weakly prime, representation degree, etc.)
3. Scope of Applicability: Limited to specific classes of links

Future Directions

1. Improve tightness of bounds
2. Extend to more general 3-manifolds
3. Seek conjectures for sharp bounds

In-Depth Evaluation

Strengths

1. Theoretical Completeness: Completes the full picture of surface counting in alternating links
2. Methodological Innovation: Cleverly combines multiple techniques to create a unified framework
3. Broad Applicability: Results apply to multiple important families of links
4. Explicit Bounds: Provides computable concrete formulas

Weaknesses

1. Tightness of Bounds: Authors acknowledge bounds may be overly conservative
2. Technical Complexity: Requires multiple technical assumptions, limiting applicability
3. Lack of Lower Bounds: No matching lower bounds or optimality discussion provided

Impact

1. Theoretical Contribution: Provides important quantitative results in 3-manifold topology
2. Methodological Value: Developed techniques may apply to other problems
3. Practical Significance: Provides complexity bounds for related algorithmic problems

Applicable Scenarios

The method is particularly suitable for:

• Topological analysis of alternating link complements
• Geometric study of virtual knot theory
• Algorithmic problems on surfaces in 3-manifolds
• Counting quasi-Fuchsian surfaces in hyperbolic geometry

References

The paper cites 54 related references covering multiple fields including 3-manifold topology, knot theory, and hyperbolic geometry, particularly foundational work by Thurston, Menasco, Lackenby, Hass-Thompson-Tsvietkova, and others.