Rigidity and gap theorems for Ricci shrinkers

Pak-Yeung Chan, Yongjia Zhang · 2026

Read Rigidity and gap theorems for Ricci shrinkers by Pak-Yeung Chan, Yongjia Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-08884.txt.

Introduction

Pak-Yeung Chan and Yongjia Zhang prove local gap theorems for Ricci shrinkers, special solutions of the Ricci flow that model how singularities form. Gap theorems say that if curvature or entropy sits close enough to the flat case, the space must be exactly flat nearby. Their local versions generalize earlier global results, and the crucial gain is that the gap size depends only on dimension, not on any global entropy or other whole space data.

As a payoff, they obtain a local test for when a Type I singularity of the Ricci flow can be removed. Geometric analysis gains tools that judge a shrinker from nearby information alone, a far more usable criterion for studying singularity neighborhoods.

Catalog record

CREATORPak-Yeung Chan, Yongjia Zhang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.08884

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Rigidity and gap theorems for Ricci shrinkers Pak-Yeung Chan, Yongjia Zhang Abstract We prove local versions of the Ricci curvature and ν \nu -entropy gap theorems for Ricci shrinkers, which respectively generalize a previous result of Munteanu-Wang [ MW11 ] and a prior result of the authors with Ma [ CMZ25 ] . The key point is that these local gaps depend only on the dimension and not on the global entropy or any other geometric information of the Ricci shrinker. As an application, we provide a local criterion for removable Type I singularities of the Ricci flow. 1 Introduction A Ricci shrinker is a tuple (M^{n},g,f) of a complete Riemannian manifold (M,g) and a smooth potential function f:M\to\mathbb{R} , satisfying the equation \displaystyle\Ric+\nabla^{2}f=\frac{1}{2}g. (1.1) In this article, we always normalize the potential function such that the following equation holds: \displaystyle|\nabla f|^{2}+R=f. (1.2) By [ CZ10 ] , the potential function f always admits a minimum point p . Since this point plays a central role in the study of Ricci shrinkers, we will use the tuple (M,g,f,p) to represent a shrinker. The most trivial example of shrinkers is the gaussian shrinker (\mathbb{R}^{n},g_{e},|x|^{2}/4,0) , where g_{e} is the standard flat metric on the Euclidean space. The rigidity of Gaussian shrinkers has attracted considerable attention in the literature. For instance, Carrillo-Ni [ CN09 ] and Yokota [ Yo12 ] proved a gap theorem for the gaussian shrinker with respect to Perelman’s \mu -functional [ Per02 ] . This result was later localized by the authors and Ma [ CMZ25 ] . On the other hand, Li-Wang [ LW20 ] studied the rigidity of the gaussian shrinker with respect to the Gromov-Hausdorff distance. They proved that a normalized shrinker, if close enough to the gaussian shrinker in the Gromov-Hausdorff sense, must be isometric to the gaussian shrinker. For other rigidity results of shrinkers, especially of the cylindrical shrinkers, see [ LW24 , CM25 , LZ23 , WW25 , Zh18 , Zh20 ] . In this article, we continue studying the rigidity of the gaussian shrinker. We shall find conditions as local and weak as possible that are sufficient to force a shrinker to be flat. Our work is inspired by [ LLW21 ] . Indeed, though not explicitly stated, the arguments of [ LLW21 ] imply the following strong rigidity result of the gaussian shrinker. Theorem 1.1 ( [ LLW21 ] ) . There is a positive constant \varepsilon=\varepsilon(n)>0 with the following property. Let (M^{n},g,f,p) be a Ricci shrinker, where p is a minimum point of f . Suppose \displaystyle d_{\operatorname{GH}}\Big(B(p,1),B_{e}(1)\Big)<\varepsilon, where d_{\operatorname{GH}} is the Gromov-Hausdorff distance, and B_{e}(1) is the unit ball in the Euclidean space, then (M^{n},g,f,p) is the gaussian shrinker (\mathbb{R}^{n},g_{e},|x|^{2}/4,0) . It is interesting to ask, whether we can find other local conditions that are sufficient to force a shrinker to be gaussian. Naber [ Na10 , Theorem 8.1] proved that a complete Ricci shrinker must be gaussian if \frac{1}{2}>\Ric\geq 0 everywhere. Munteanu-Wang [ MW11 , Corollary 1.1] showed that a shrinker is flat if |{\Ric}|\leq\frac{1}{100n} globally on M . The authors and Ma [ CMZ25 ] proved that a shrinker with sufficiently large curvature radius (depending only on the dimension) at a minimum point of f must be flat. In light of these results, as well as those of Li-Li-Wang [ LLW21 ] , it is natural to ask whether the same conclusion holds under a small local Ricci curvature bound near a minimum point of f . We show that this is indeed the case. Theorem 1.2 . There is a positive constant \varepsilon=\varepsilon(n)>0 with the following property. Let (M^{n},g,f,p) be a Ricci shrinker, where p is a minimum point of f . Suppose \displaystyle\Ric\leq\varepsilon g\quad\text{ on }\quad B(p,1), then (M^{n},g,f,p) is the gaussian shrinker (\mathbb{R}^{n},g_{e},|x|^{2}/4,0) . Remarks. 1. Since on a shrinker, we always have R>0 by a result of Chen [ Che09 ] , it makes no difference to assume |{\Ric}|\leq\varepsilon or \Ric\leq\varepsilon g ; we will freely implement the former assumption in our argument. 2. According to our argument, the theorem can be improved straightforwardly to the following: For any r>0 , there is an \varepsilon=\varepsilon(r,n)>0 , such that if \displaystyle\Ric\leq\varepsilon g\quad\text{ on }\quad B(p,r), then the shrinker is gaussian. 3. Li-Li-Wang [ LLW21 , Theorem 1.3] obtained a local gap theorem for the scalar curvature of closed shrinkers, where the gap also depends on the entropy. Wang-Wang [ WW25 ] established a rigidity theorem for shrinkers in terms of the volume ratio on a large ball centered at p . In [ CMZ25 ] , it is also proved that a shrinker with sufficiently small local \mu -functional in a sufficiently large ball at the minimum point of the potential function must be flat. We will also provide an improvement of this theorem. Theorem 1.3 . There is a positive constant \varepsilon=\varepsilon(n)>0 with the following property. Let (M^{n},g,f,p) be a Ricci shrinker, where p is the minimum point. Suppose \displaystyle\nu\Big(B(p,1),g,1\Big)\geq-\varepsilon, then (M^{n},g,f,p) is the gaussian shrinker (\mathbb{R}^{n},g_{e},|x|^{2}/4,0) . Since, by [ EMT11 ] , smooth Ricci shrinkers are singularity models (fixed-point blow-up limits) of Type I Ricci flows, and since a flat blow-up limit implies that the base point is not a singular point [ EMT11 , Theorem 1.1] , we may apply Theorem 1.2 to obtain the following result concerning removable singularities for Type I Ricci flows. Theorem 1.4 (A local condition for removable Type I singularities) . There is a dimensional constant \delta=\delta(n)>0 with the following property. Let (M^{n},g_{t})_{t\in[-T,0)} be a Type I Ricci flow (that is, |{\Rm_{g_{t}}}|\leq C/|t| for some C ) on a close manifold M^{n} . Assume x\in M is a point such that r_{c}(x,t)\geq 6n\sqrt{|t|} for all t\in[-T,0) , where \displaystyle r_{c}(x,t):=\sup\{r>0\,:\,\Ric_{g_{t}}\leq\delta r^{-2}g_{t}\ \text{ on }\ B_{g_{t}}(x,r)\}. Then |{\Rm_{g_{t}}}|(x) remains bounded as t\to T . In other words, x is not a singular point according to the definition of [ EMT11 ] . Remarks. 1. Note that the constant \delta=\delta(n) is a dimensional constant, and it is in particular independent of the Type-I-coefficient C . On the other hand, a global version of the theorem can be deduced from [ MW11 ] , namely, if r_{c}(x,t)\geq\sqrt{|t|} for all (x,t)\in M\times[-T,0) , then the flow does not admit any singularity at t=0 . 2. One may also apply Theorem 1.3 to obtain a condition of removable singularities formulated with the local \nu -functional. However, this result is not different from that which one can deduce from Perelman’s pseudolocality theorem [ Per02 , Theorem 10.1] . Acknowledgement. The second author would like to thank Professor Lei Ni, Professor Liang Cheng, and Professor Meng Zhu for some enlightening discussions. The authors would like to thank Professor Bennett Chow for providing some helpful unpublished manuscripts. 2 Preliminaries In this section, we summarize the main techniques for the proof of our main theorems. 2.1 Harmonic radius estimate First of all, we introduce an \varepsilon -regularity due to Anderson [ And90 ] . Let us recall the definition of harminic radius. Definition 2.1 (Harmonic radius) . Let (M^{n},g) be a complete Riemannian manifold and p\in M . The (Q,\alpha) -harmonic radius r_{H}(p) at p , where Q>1 and \alpha\in(0,1) , is defined

Read the complete text