[Paper Review] Mean curvature flow as a tool to study topology of 4-manifolds

Tobias Colding, William P. Minicozzi+1 more|arXiv (Cornell University)|Aug 29, 2012

Geometric Analysis and Curvature Flows13 references3 citations

TL;DR

This paper proposes using mean curvature flow (MCF) as a tool to study the topology of 4-manifolds by evolving smooth, closed 4-dimensional hypersurfaces embedded in ℝ⁵. It shows that under generic initial conditions, MCF produces only generic singularities—shrinking spheres and cylinders—whose structure encodes topological information, and conjectures that these singularities admit canonical neighborhood structures, offering a potential pathway to solving classification problems in 4-dimensional topology via geometric flows.

ABSTRACT

In this paper we will discuss how one may be able to use mean curvature flow to tackle some of the central problems in topology in 4-dimensions. We will be concerned with smooth closed 4-manifolds that can be smoothly embedded as a hypersurface in R^5. We begin with explaining why all closed smooth homotopy spheres can be smoothly embedded. After that we discuss what happens to such a hypersurface under the mean curvature flow. If the hypersurface is in general or generic position before the flow starts, then we explain what singularities can occur under the flow and also why it can be assumed to be in generic position. The mean curvature flow is the negative gradient flow of volume, so any hypersurface flows through hypersurfaces in the direction of steepest descent for volume and eventually becomes extinct in finite time. Before it becomes extinct, topological changes can occur as it goes through singularities. Thus, in some sense, the topology is encoded in the singularities.

Motivation & Objective

• To investigate whether mean curvature flow (MCF) can be used to analyze the topology of smooth, closed 4-manifolds embedded in ℝ⁵.
• To establish that all closed smooth homotopy 4-spheres can be smoothly embedded as hypersurfaces in ℝ⁵, generalizing classical results of Kervaire and Milnor.
• To analyze the nature of singularities that arise during MCF of such hypersurfaces, particularly under generic initial conditions.
• To explore the possibility that only generic singularities—shrinking spheres and cylinders—occur under generic initial data, and that these singularities carry topological information.
• To develop a canonical neighborhood theorem near generic singularities, suggesting that MCF near such points behaves like a positive mean curvature flow, potentially enabling topological classification.

Proposed method

• Use the fact that any closed smooth 4-manifold homotopy equivalent to S⁴ can be smoothly embedded in ℝ⁵, based on classical surgery theory and results of Kervaire and Milnor.
• Apply mean curvature flow as the negative gradient flow of the volume functional, evolving hypersurfaces toward minimal volume configurations.
• Analyze the evolution of hypersurfaces under MCF, focusing on the types of singularities that can occur, especially in generic initial positions.
• Leverage Huisken’s monotonicity formula and the theory of self-similar shrinkers to understand the asymptotic behavior near singularities.
• Use the linearization of the rescaled MCF around self-shrinkers, governed by the operator L defined via the Hessian of the F-functional, to study stability and dynamics near singularities.
• Apply infinite-dimensional Morse theory ideas to model the local dynamics near self-shrinkers using the negative gradient flow of a quadratic functional on L²(Σ, e^{−|x|²/4} dVol), with eigenvalues μ_i of the operator L.

Experimental results

Research questions

• RQ1 Can mean curvature flow be used to extract topological invariants of 4-manifolds through the structure of singularities that arise during the flow?
• RQ2 What types of singularities can occur under mean curvature flow of a generic closed hypersurface in ℝ⁵, and can only the simplest ones—shrinking spheres and cylinders—persist?
• RQ3 Is there a canonical neighborhood structure around generic singularities in MCF, such that the flow near such points has positive mean curvature?
• RQ4 To what extent does the topology of a 4-manifold embedded in ℝ⁵ become encoded in the nature and classification of the singularities of its MCF?
• RQ5 Can the rigidity of generalized cylinders (Sᵏ × ℝⁿ⁻ᵏ) be proven under entropy and mean curvature bounds, implying uniqueness of such singular models?

Key findings

• All closed smooth 4-manifolds homotopy equivalent to S⁴ can be smoothly embedded as hypersurfaces in ℝ⁵, based on surgery theory and classical results of Kervaire and Milnor.
• Under generic initial conditions, mean curvature flow of a closed hypersurface in ℝ⁵ produces only generic singularities—specifically, shrinking spheres and shrinking cylinders—because higher-order singularities can be perturbed away.
• The singularities of MCF are modeled on self-shrinkers, and the local dynamics near a self-shrinker are approximated by the negative gradient flow of a quadratic functional derived from the Hessian of the F-functional.
• The conjecture is supported that if a self-shrinker has non-negative mean curvature on a large ball and bounded entropy, then it must be a generalized cylinder Sᵏ × ℝⁿ⁻ᵏ, implying strong rigidity.
• A canonical neighborhood theorem is proposed: if a singularity is modeled on a generalized cylinder, then in a full space-time neighborhood of the singularity, the flow has positive mean curvature, suggesting a universal local structure.
• The work in progress with Ilmanen confirms that such canonical neighborhoods exist under mild assumptions, and full proofs of the conjectures appear within reach.