Lectures on curve shortening flow

Abstract: The curve shortening flow is a geometric heat equation for curves and provides an accessible setting to illustrate many important concepts from nonlinear partial differential equations, including maximum principle estimates, monotonicity formulas, Harnack inequalities and blowup analysis. All these techniques will be combined to give an exposition of Huisken's proof of Grayson's beautiful theorem that the curve shortening flow shrinks any closed embedded curve in the plane to a round point.

• The paper presents a rigorous analysis of curve shortening flow for embedded planar curves, culminating in a complete proof of Grayson's theorem.
• It employs nonlinear parabolic PDE techniques, deriving a priori estimates via the maximum principle, Harnack inequalities, and monotonicity formulae.
• The study highlights practical implications, from imaging and shape smoothing to insights for higher-dimensional geometric flows.

Expert Synopsis of "Lectures on Curve Shortening Flow" (2604.01981)

Introduction

This monograph delivers a comprehensive and technically rigorous account of the curve shortening flow (CSF) for embedded planar curves, emphasizing the interplay between geometric analysis and nonlinear parabolic PDEs. Building upon seminal results—including Grayson's theorem and the convergence to round points—it systematically introduces the maximum principle, derivative estimates, monotonicity formulae, Harnack inequalities, and intricate blowup analyses, situating the subject within the larger landscape of geometric flows.

Core Formulation and Analytical Foundations

The CSF evolves an embedded curve $\Gamma_t$ by its curvature vector, with the velocity $\partial_t p = \vec{\kappa}(p)$. Classical solutions, such as shrinking circles and the grim reaper, are detailed, illustrating explicit dynamics and reinforcing the nonlinear, parabolic nature of the flow. The PDE, while locally resembling the heat equation when recast in arclength coordinates, exhibits non-strict parabolicity due to its geometric invariance under reparametrization.

Existence, uniqueness, and instant regularization are treated via quasilinear parabolic reduction, with the maximal time of existence signaled by curvature blowup. The treatment extends well beyond smooth initial data, as established flows become instantaneously smooth even from Jordan curves of finite length.

A Priori and Derivative Estimates

The manuscript provides meticulous derivations of geometric a priori bounds via the maximum principle and integral formulas. Notably, the area $A(t)$ decreases linearly, achieving extinction at $T = A(0)/2\pi$, while the length $L(t)$ decreases according to the $L^2$-norm of the curvature, framing CSF as the gradient flow of the length functional.

Crucial to later blowup analysis, the derivative estimates (Theorem 1.8) are proved inductively, furnishing pointwise control of higher order spatial derivatives in terms of curvature bounds. Convexity is sharply preserved, and convex initial data yield explicit lower bounds for the minimal curvature through time.

Monotonicity and Singularity Analysis

Harnessing Huisken's monotonicity formula—based on integrals against the backward heat kernel—the text establishes scale-invariant quantities that decay monotonically except for soliton-like solutions. This formula is central for classifying singularity models via blowup limits.

Using parabolic rescaling and the type I curvature blowup assumption, any blowup sequence at a singular time converges (in the smooth Cheeger-Gromov sense) to the family of round shrinking circles, underlining the rigidity and universality of singularity formation for embedded CSF.

A local regularity theorem (adapted from Brakke and White) is established: if the Gaussian density is sufficiently close to 1 in a parabolic neighborhood, curvature is quantitatively controlled. This not only precludes nontrivial singularities below the critical density, but provides a powerful mechanism to propagate regularity away from concentration points.

Harnack Inequality and Soliton Classification

The manuscript proves Hamilton's Harnack inequality for convex solutions, delivering lower bounds on the evolution of the curvature function and ruling out various types of degeneracy. The strong form of the Harnack estimate $$\frac{\kappa_t}{\kappa} - \frac{\kappa_s^2}{\kappa^2} + \frac{1}{2t} \geq 0$$ is derived with careful use of the commutator structure and maximum principle.

Ancient nontrivial convex solutions are shown to be translating solitons, culminating in the complete classification: any eternal strictly convex ancient CSF solution with a curvature critical point is, up to rigid motions and scaling, the grim reaper.

Embeddedness, Distance Comparison, and Quantitative Geometry

Huisken's distance comparison theorem is proved in detail, establishing monotonicity of a refined isoperimetric ratio $R(t)$ that blends intrinsic and extrinsic geometric data. This result quantitatively encodes the preservation of embeddedness and rules out limiting behavior by noncompact or multiply covered grim reaper-type solutions.

Grayson's Theorem: Convergence to Round Points

Synthesizing the developed theory, the proof of Grayson's theorem is delivered in full analytic generality. Any embedded, closed planar curve under CSF shrinks smoothly to a round point, with area and extinction time explicitly related. Blowup analyses distinguish between type I (round-point) and type II (soliton) scenarios, excluding the latter by integrating the maximum principle, Hamilton's Harnack inequality, and Huisken's distance comparison. The limiting behavior is thus uniquely the round shrinking circle.

Implications and Directions

This exposition not only codifies the technical machinery for analyzing CSF, but also serves as a template for handling higher-dimensional mean curvature flows and other geometric evolution equations. The techniques—maximum principle-based a priori bounds, monotonicity, blowup classification, Harnack estimates, and quantitative embeddedness criteria—are foundational in the study of dispersive and parabolic geometric PDEs.

On a practical level, these results underpin algorithmic approaches in imaging (e.g., active contours), shape smoothing, and geometric quantization; theoretically, they scaffold the program of understanding singularities and long-time behavior in geometric flows.

Looking forward, the interplay between local regularity, monotonicity formulae, and blowup analysis encapsulated here suggests fruitful extensions to anisotropic flows, networks and higher codimension. The universal emergence of round points as singularity models hints at a pervasive role for entropy minimization and rigidity phenomena in geometric analysis.

Conclusion

The lecture notes systematically present the core analytic and geometric frameworks for the theory of curve shortening flow, culminating in a complete, modern proof of Grayson's theorem for embedded curves. The text underscores how sophisticated PDE techniques yield precise dynamical and singularity descriptions for geometric evolution and lay a foundation for further theoretical and applied research in geometric flows.