The Ricci flow equation is a geometric evolution equation on Riemannian manifolds:

gijt=2Rij\frac{\partial g_{ij}}{\partial t} = -2 R_{ij}

Here gijg_{ij} is the Riemannian metric and RijR_{ij} is its Ricci curvature tensor. Introduced by Richard Hamilton in 1982, Ricci flow became the technical foundation for Perelman’s 2002–2003 proof of the Poincaré conjecture and the stronger Thurston geometrization conjecture.

1. The idea

Given a Riemannian manifold (M,g)(M, g), Ricci flow deforms the metric in time by its own Ricci curvature. Intuitively, the flow contracts regions of positive Ricci curvature (like the round sphere) and expands regions of negative curvature (like hyperbolic space). Over time, the metric is pushed toward a canonical geometry.

The equation is a non-linear parabolic PDE on the space of metrics. It is the gradient flow of Perelman’s F\mathcal{F}-functional:

F(g,f)=M(R+f2)efdVg\mathcal{F}(g, f) = \int_M (R + |\nabla f|^2)\, e^{-f}\, dV_g

where RR is the scalar curvature and ff is an auxiliary function.

2. Hamilton’s program

Hamilton proposed in 1982 that Ricci flow could be used to prove the Poincaré conjecture (every simply-connected closed 3-manifold is homeomorphic to S3S^3) by deforming any such manifold’s metric until it became round — i.e., revealed as a sphere.

He proved the program for 3-manifolds with positive Ricci curvature (1982): they flow to constant-curvature metrics. But the general case hit a wall — Ricci flow can develop singularities in finite time, and Hamilton could not control them.

3. Perelman’s breakthrough

Grigori Perelman, in three arXiv preprints (2002–2003), solved the outstanding technical problems:

  1. Monotonicity formulas — Perelman introduced reduced volume and F\mathcal{F}- and W\mathcal{W}-entropy functionals, proving they are monotone along Ricci flow. These gave global control.

  2. Analysis of singularities — Perelman classified all possible singularities of Ricci flow on 3-manifolds, showing they arise from “neck pinches” and topological cap-offs.

  3. Ricci flow with surgery — the crowning technical contribution: cut out neighborhoods of incipient singularities, glue in standard caps, continue the flow. After finitely many surgeries, the manifold decomposes canonically — proving both the Poincaré and the Thurston geometrization conjectures.

The proof totals only about 70 pages across the three papers, but required several years of additional verification by teams of mathematicians before it was universally accepted.

4. The geometric picture

For a manifold with positive curvature that flows to a round sphere, the normalized Ricci flow

gijt=2Rij+2nRgij\frac{\partial g_{ij}}{\partial t} = -2 R_{ij} + \frac{2}{n} R g_{ij}

(where nn is the dimension and RR is the scalar curvature) preserves the volume and converges to an Einstein metric — one with RijR_{ij} proportional to gijg_{ij}. In dimension 3, the only compact Einstein metrics are constant-curvature, so the manifold becomes locally symmetric.

Thurston’s geometrization conjecture refined this: every compact 3-manifold decomposes into pieces each carrying one of the eight Thurston geometries. Perelman’s proof established this decomposition rigorously via Ricci flow with surgery.

5. Extensions and variants

Ricci flow has spawned a productive industry in geometric analysis:

  • Kähler-Ricci flow on complex manifolds
  • Extended Ricci flow for connections and bundles
  • Mean curvature flow for hypersurfaces (intensely studied independently)
  • Inverse mean curvature flow (Huisken-Ilmanen, used in Penrose-inequality work)
  • Q-curvature flow and related conformal flows

Each is a geometric PDE evolution whose fixed points or long-time behavior encode deep topological or metric information.

6. Context: the solved Millennium Prize

The Poincaré conjecture, proved via Ricci flow, is the only one of the seven Millennium Prize Problems to have been solved. The Clay Mathematics Institute awarded Perelman the million-dollar prize in 2010; he declined it, as he had declined the 2006 Fields Medal.

Further reading

  • Hamilton, “Three-manifolds with positive Ricci curvature” (1982).
  • Perelman, arXiv:math/0211159, 0303109, 0307245 (2002–2003).
  • Morgan & Tian, Ricci Flow and the Poincaré Conjecture (2007) — full exposition of the proof.
  • Kleiner & Lott, “Notes on Perelman’s papers” (2008) — line-by-line verification.

Frequently asked

Why is Ricci flow called a 'geometric heat equation'?

Because the equation has the same structure as the heat equation ∂u/∂t = Δu — the metric 'diffuses' by its curvature in a way that smooths out local irregularities. Just as heat redistributes to reach a uniform temperature, Ricci flow tries to redistribute curvature toward a canonical, more symmetric geometry.

What are the singularities of Ricci flow, and why do they matter?

Ricci flow can develop singularities in finite time — regions where curvature blows up. Hamilton's original program could not handle these; Perelman's key innovation was surgery: a rigorous procedure to cut out the singular region, cap the holes, and continue the flow on the modified manifold. Controlling the singularities is the technical heart of the Poincaré proof.

How does Ricci flow connect to Einstein's equations?

Einstein's vacuum field equations require Rij = 0. Ricci flow is, in a precise sense, a gradient flow for an action related to the Einstein–Hilbert action. Fixed points of Ricci flow with zero scalar curvature satisfy Einstein's equations, linking the flow to general relativity and differential geometry.