Abstract: In this paper, we establish the rigidity of the generalized cylinder $N^n \times \mathbb R^{m-n}$, or a quotient thereof, in the space of Ricci shrinkers equipped with the pointed-Gromov-Hausdorff topology. Here, $N$ is a stable Einstein manifold that has an obstruction of order $3$. The proof is based on a quantitative characterization of the rigidity of compact Ricci shrinkers, a rigidity inequality of mixed orders on generalized cylinders, and the method of contraction and extension. As an application, we prove the uniqueness of the tangent flow for general compact Ricci flows under the assumption that one tangent flow is a generalized cylinder.
Introduction
In the study of Ricci flow—a geometric process that deforms a manifold's metric to smooth out its curvature—singularities often develop where the curvature becomes infinite. To understand these "blow-ups," mathematicians analyze Ricci shrinkers, which represent the self-similar limit shapes of the flow near a singularity.
A fundamental question in this field is the uniqueness of the tangent flow: if one "zooms in" on a singularity at different rates or along different sequences of times, does the resulting geometric model remain consistent? This property depends heavily on whether the model itself is rigid. A shape is considered rigid if it is geometrically isolated within the space of all possible shrinkers. While rigidity has been established for simpler cases like spheres or Gaussian solitons, proving it for more complex structures like "generalized cylinders" ($N^n \times \mathbb{R}^{m-n}$) has remained a significant challenge.
Methodology
The authors establish rigidity through a sophisticated blend of spectral geometry and analysis on non-compact manifolds. Their approach is built upon three main pillars:
- Quantitative Rigidity Inequalities: Moving beyond qualitative descriptions, the authors prove that a compact Ricci shrinker is rigid if and only if it satisfies a specific inequality of order $k$. This inequality bounds the distance between a perturbed metric and the rigid metric using the Ricci shrinker operator $\Phi$.
- Mixed-Order Analysis: For generalized cylinders, rigidity is not uniform across the different factors of the product space. The authors derive a "rigidity inequality of mixed orders" to account for the distinct contributions of the Einstein manifold $N$ and the Euclidean factor $\mathbb{R}^{m-n}$ to the overall stability of the shrinker.
- Contraction and Extension: This iterative technique allows the authors to show that if two Ricci shrinkers are sufficiently close on a large compact set, they must be identical. By "contracting" the region to improve error estimates and then "extending" it back out, they prove the two shrinkers are isometric.
Key Insights and Results
A central technical contribution is the identification of the "obstruction of order 3" on the Einstein manifold $N$. This condition ensures that $N$ cannot be part of a continuous family of Einstein metrics, a necessary prerequisite for the entire cylinder's rigidity. Furthermore, the authors bridge the gap between compact and non-compact manifolds by utilizing weighted Sobolev spaces and elliptic estimates that account for the quadratic growth of the potential function $f$ at infinity.
Major Findings
- Rigidity of Generalized Cylinders: The paper establishes that the generalized cylinder $N^n \times \mathbb{R}^{m-n}$ is rigid in the pointed-Gromov-Hausdorff topology, provided $N$ is a stable Einstein manifold with a 3rd-order obstruction.
- Classification of 4D Shrinkers: As a corollary, the authors prove that any four-dimensional Ricci shrinker with a non-negative curvature operator is rigid, providing a definitive answer for a large class of 4D models.
- Uniqueness of Tangent Flow: The research demonstrates that for general compact Ricci flows, if one tangent flow at a singularity is a generalized cylinder of the type described, then that tangent flow is unique. This represents a major step toward a complete stratification of Ricci flow singularities.
Conclusion and Future Work
While this work provides a robust framework for understanding rigidity in generalized cylinders, several avenues for future research remain. The current proof relies on a 3rd-order obstruction; manifolds with higher-order obstructions or those that are not H-stable remain an open problem. Additionally, extending these techniques to non-cylindrical models, such as those with conical singularities, or to Type-II singularities where curvature grows faster than the reciprocal of time, remains a significant goal in the field of geometric analysis.
Summary updated