Abstract
In this paper, we prove an n-dimensional radially flat gradient shrinking Ricci solitons with \(div^2W(\nabla f,\nabla f)=0\) is rigid. Moreover, we show that a four-dimensional radially flat gradient shrinking Ricci soliton with \(\text {div}^2W^\pm (\nabla f,\nabla f)=0\) is either Einstein or a finite quotient of \({\mathbb {R}}^4\), \({\mathbb {S}}^2\times {\mathbb {R}}^2\) or \({\mathbb {S}}^3\times {\mathbb {R}}\).
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
A complete Riemannian manifold \((M^n,g,f)\) is called a gradient shrinking Ricci soliton if there exists a smooth function f on \(M^n\) such that the Ricci tensor Ric of the metric g satisfies the equation
for some constant \(\lambda >0\). By scaling the metric g, one customarily assumes \(\lambda =\frac{1}{2}\). In this paper, we focus on the geometric structures of gradient shrinking Ricci solitons.
A gradient soliton is rigid if it is a flat bundle \(N\times _{\Gamma }{\mathbb {R}}^k\), where \(\Gamma \) acts freely on an Einstein manifold N and by orthogonal transformations on \({\mathbb {R}}^k\) (see Peterson–Wylie [11] for details).
Some classification theorems for gradient shrinking Ricci solitons have been proved in the recent years by various authors. An n-dimensional locally conformally flat gradient shrinking Ricci soliton was classified (cf. [6, 12, 17]). Moreover, an n-dimensional gradient shrinking Ricci soliton with harmonic Weyl tensor is rigid (see [7, 9]). More generally, Cao and Chen [1] proved that an n-dimensional \((n\ge 5)\) complete non-compact Bach-flat gradient shrinking Ricci soliton is a finite quotient of \({\mathbb {R}}^n\) or \({\mathbb {R}}\times N^{n-1}\), where N is an \((n-1)\)-dimensional Einstein manifold. In 2016, Catino et al. [3] showed that a gradient shrinking Ricci soliton with fourth-order divergence-free Weyl tensor (i.e. \(\text {div}^4W=0\)) is rigid.
In the case of dimension 4, Naber [10] classified four-dimensional non-compact shrinking Ricci solitons with bounded nonnegative Riemannian curvature. Cao and Chen [1] proved that a four-dimensional complete non-compact Bach-flat gradient shrinking Ricci soliton is a finite quotient of \({\mathbb {R}}^4\) or \({\mathbb {R}}\times {\mathbb {S}}^3\). Under the condition of \(W^+=0\), Chen and Wang [5] showed that a four-dimensional complete gradient shrinking Ricci soliton is isometric to finite quotients of \({\mathbb {R}}^4\), \({\mathbb {S}}^3\times {\mathbb {R}}\), \({\mathbb {S}}^4\), or \({\mathbb {C}}P^2\). More generally, Wu et al. [13] proved that a four-dimensional gradient shrinking Ricci soliton with half harmonic Weyl tensor (i.e. \(\text {div}W^\pm =0\)) is either Einstein or a finite quotient of \({\mathbb {R}}^4\), \({\mathbb {R}}^2\times {\mathbb {S}}^2\) or \({\mathbb {R}}\times {\mathbb {S}}^3\).
Under radial conditions, Peterson and Wylie [11] proved that a gradient soliton is rigid if and only if it has constant scalar curvature and is radially flat, that is, \(sec(E,\nabla f)=0\), where sec stands for the sectional curvature operator and E is an arbitrary vector field. Moreover, they [12] showed that a complete gradient Ricci soliton with constant scalar curvature and \(W(\nabla f,\cdot ,\cdot ,\nabla f)=o(|\nabla f|^2)\) is a flat bundle of rank 0, 1, or n over an Einstein manifold. Moreover, the authors [14] proved that a gradient shrinking Ricci soliton is rigid if \(Rm(\cdot ,\nabla f)\nabla f=0\) and \(\text {div}^2B\le 0\). They [14] also showed that a complete non-compact gradient expanding Ricci soliton with nonnegative Ricci curvature is rigid if \(Rm(\cdot ,\nabla f)\nabla f=0\) and \(\text {div}^2B\ge 0\).
In order to state our results precisely, we introduce the following definitions for the Riemannian curvature:
For the Weyl tensor, we define:
We explore the geometric structures of gradient shrinking Ricci solitons with some radially conditions. The main results of this paper are the following theorems.
For n-dimensional gradient shrinking Ricci solitons, we have the following rigid result.
Theorem 1.1
Let \((M^n,g,f)\) be a radially flat gradient shrinking Ricci soliton with \(\text {div}^2W(\nabla f,\nabla f)=0\). Then \((M^n,g,f)\) is rigid.
For 4-dimensional gradient shrinking Ricci solitons, we have the following classification theorem.
Theorem 1.2
Let \((M^4,g,f)\) be a four-dimensional radially flat gradient shrinking Ricci soliton. If \(\text {div}^2W^\pm (\nabla f,\nabla f)=0\). Then \((M^4,g,f)\) is either Einstein or a finite quotient of \({\mathbb {R}}^4\), \({\mathbb {S}}^2\times {\mathbb {R}}^2\) or \({\mathbb {S}}^3\times {\mathbb {R}}\).
Remark 1.3
It is clear from the proof that the assumptions on the vanishing of \(\text {div}^2W(\nabla f,\nabla f)\) and \(\text {div}^2W^\pm (\nabla f,\nabla f)\) in the above theorems can be trivially relaxed to a (suitable) inequality. To be precise, the condition of \(\text {div}^2W(\nabla f,\nabla f)=0\) in Theorem 1.1 can be relaxed to \(\int \text {div}^2W(\nabla f,\nabla f)e^{-f}\ge 0\), while the the condition of \(\text {div}^2W(\nabla f,\nabla f)=0\) in Theorem 1.2 can be relaxed to \(\int \text {div}^2W^\pm (\nabla f,\nabla f)e^{-f}\ge 0\).
We arrange this paper as follows. In Sect. 2, we fix our notations and recall some known results about gradient Ricci solitons that we shall need in the proof of Theorems 1.1 and 1.2. We prove the rigid result for n-dimensional gradient shrinking Ricci solitons (Theorem 1.1) in Sect. 4. Before we classify the four-dimensional gradient shrinking Ricci soliton, we present some useful formulas of the curvature in Sect. 5. Finally, we finish the proof of Theorem 1.2 in Sect. 6.
2 Preliminaries
First of all, we recall that on any n-dimensional \((n\ge 3)\) Riemannian manifold, the Weyl tensor is given by
The Cotton tensor is defined as
The relation between the Cotton tensor and the divergence of the Weyl tensor is
In the case of dimension 4, denote \((i'j')\) to be the dual of any pair (ij) with \(1\le i\ne j\le 4\). That is the pair such that \(e_i\wedge e_j\pm e_{i'}\wedge e_{j'}\in \wedge ^{\pm }M^4\). In other words, \((i'j'ij)=\sigma (1234)\) for some even permutation \(\sigma \in S_4\), i.e.
For any (0, 4)-tensor T of an 4-dimensional manifold, its (anti-)self-dual part is
On four-manifolds, the Weyl tensor has sufficiently exotic symmetries.
Proposition 2.1
(Wu et al. [13]) Let (M, g) be a four-dimensional Riemannian manifold. Then
therefore,
It follows that
Therefore, we have
and
Next, we recall some basic facts about complete gradient shrinking Ricci solitons.
Proposition 2.2
(Yang and Zhang [15]) Let \((M^n,f,g)\) be a gradient Ricci soliton satisfying (1.1). Then we have
and
Lemma 2.3
(Cao and Zhou [2]) Let \((M^n,g)\) be a complete gradient shrinking soliton with (1.1). Then,
-
1.
the potential function f satisfies the estimates
$$\begin{aligned} \frac{1}{4}(r(x)-c_1)^2\le f(x)\le \frac{1}{4}(r(x)+c_2)^2, \end{aligned}$$(2.3)where \(r(x)=d(x_0,x)\) is the distance function from some fixed point \(x_0\in M\), \(c_1\) and \(c_2\) are positive constants depending only on n and the geometry of g on the unit ball \(B(x_0,1)\);
-
2.
there exists some constant \(C>0\) such that
$$\begin{aligned} Vol(B(x_0,s))\le Cs^n \end{aligned}$$(2.4)for \(s>0\) sufficiently large.
Lemma 2.4
(Peterson and Wylie [11]) A gradient soliton is rigid if and only if it has constant scalar curvature and is radially flat, that is, \(\text {sec}(E,\nabla f)=0\).
3 Two Formulas for Gradient Ricci Solitons
In this section, we prove two identities for gradient Ricci solitons. First of all, we give a identity of \(\text {div}Rm^2(\nabla f,\nabla f)\).
Proposition 3.1
Let \((M^n,g,f)\) be a gradient Ricci soliton. Then, we have
Proof
It follows from (2.1) that
where we used \(Ric(\nabla f,\cdot )=\frac{1}{2}\nabla R\) and (1.1) in the third equality. \(\square \)
Lemma 3.2
Let \((M^n,g,f)\) be a gradient radially flat Ricci soliton. Then, we have
Proof
From (2.2) and Proposition 3.1, we have
Note that \(sec(E,\nabla f)=0\) implies that \(Ric(\nabla f,\nabla f)=0\). In the following, we show that \(R_{ijkl}R_{jl}\nabla _if\nabla _kf=0\) on \(M^n\).
\(\forall x\in M^n\), we divide the arguments into two cases:
-
Case 1: \(\nabla f(x)=0\). It is clear that (3.2) holds at x.
-
Case 2: \(\nabla f(x)\ne 0\). In this case, we denote by \(\{e_i\}_{i=1}^n\) a local orthonormal frame of \(M^n\) with \(e_1=\frac{\nabla f}{|\nabla f|}\). We use \(\{\alpha _i\}_{i=1}^n\) to represent eigenvalues of the Ricci tensor with corresponding orthonormal eigenvectors \(\{e_i\}_{i=1}^n\), respectively. Since \(sec(E,\nabla f)=0\), at x we have
$$\begin{aligned} R_{ijkl}R_{jl}\nabla _if\nabla _kf=\sum _{s=2}^nR_{1s1s}\alpha _s|\nabla f|^2=0. \end{aligned}$$(3.4)
This completes the proof of Lemma 3.2. \(\square \)
4 Proof of Theorem 1.1
For gradient shrinking Ricci soliton satisfying the equation
integrating the right side of (3.2), we have the following lemma.
Lemma 4.1
Let \((M^n,g,f)\) be a gradient shrinking Ricci soliton. Then, we have
Proof
We divide the arguments into the compact case and the complete non-compact case:
-
The compact case:
Integrating by parts,we have
where we used (4.1) and \(\Delta f=\frac{n}{2}-R\) in the second equality. Moreover, we used \(Ric(\nabla f,\cdot )=\frac{1}{2}\nabla R\) in the third equality.
Integrating by parts, we obtain
where we used (4.1) in the second equality and \(Ric(\nabla f,\cdot )=\frac{1}{2}\nabla R\) in the third.
Combining (4.3) and (4.4), we obtain (4.2).
-
The complete non-compact case:
Let \(\phi (t)=\frac{s-t}{s}\) on (0, s), \(\phi =0\) on \([s,\infty )\) for any fixed constant \(s>0\). From Lemma 2.3, we know that f is of quadratic growth and the volume of a geodesic ball is at most Euclidean growth. Moreover, the support of \(\phi (f)\) is a closed subset of the compact set \(\{x|\frac{1}{4}(r(x)-c_1)^2\le s\}\) for any fixed \(s>0\). Therefore, \(\phi (f)\) has compact support on \(M^n\).
Integrating by parts, we have
where we used (4.1) and \(\Delta f=\frac{n}{2}-R\) in the second equality. Moreover, we used \(Ric(\nabla f,\cdot )=\frac{1}{2}\nabla R\) in the third equality.
Integrating by parts, we obtain
where we used (4.1) in the second equality and \(Ric(\nabla f,\cdot )=\frac{1}{2}\nabla R\) in the third.
Combining (4.5) and (4.6), we obtain
where we used \(\nabla R=2Ric(\nabla f,\cdot )\).
Note that \(R\ge 0\) (see Chen [4]), \(R+|\nabla f|^2- f=Const.\) and f is at most quadratic growth, we know that \(|\nabla f|\) is at most linear growth and R is at most quadratic growth. It follows from Lemma 2.2 that
and
where c is a constant in (0, 1]. Moreover, we have
(4.2) follows by taking \(s\rightarrow +\infty \) in (4.7).
This completes the proof of Lemma 4.1. \(\square \)
Now we are ready to prove Theorem 1.1.
Theorem 4.2
Let \((M^n,g,f)\) be a radially flat gradient shrinking Ricci soliton with \(\text {div}^2W(\nabla f,\nabla f)=0\). Then \((M^n,g,f)\) is rigid.
Proof
Integrating (3.2) and using the condition of \(\text {div}^2W(\nabla f,\nabla f)=0\), we obtain
Note that \(sec(E,\nabla f)=0\) implies \(Ric(\nabla f,\nabla f)=0\). Applying Lemma 4.1 to (4.11), we have
It follows that \(|\nabla R|=0\) a.e. on \(M^n\). Since any gradient shrinking Ricci soliton is analytic in harmonic coordinates, we have \(|\nabla R|\equiv 0\) on \(M^n\), i.e. \(M^n\) has constant scalar curvature.
From Lemma 2.4, we know that \((M^n,g,f)\) is rigid. \(\square \)
5 Curvatures of Four-Dimensional Gradient Ricci Solitons
In this section, we present a formula (Proposition 5.3) on curvatures that are essential in the proof of Theorem 1.2. First of all, we recall two propositions that are proved by the authors [16]
Proposition 5.1
(Yang and Zhang [16]) On a four-dimensional Riemannian manifold, we have
and
Proposition 5.2
(Yang and Zhang [16]) Let \((M^4,f,g)\) be a four-dimensional gradient Ricci soliton. Then we have
Proof
The second Bianchi identity implies that
Using (1.1), we have
By direct computation, we obtain
where we used (5.5) in the first identity, (1.1) in the third and Proposition 5.1 in the last.
By the same arguments, we have
Applying (5.6) and (5.7) to (5.4), we have
where we used (5.2).
This completes the proof of proposition 5.2. \(\square \)
Proposition 5.3
Let \((M^4,f,g)\) be a four-dimensional gradient Ricci soliton. Then we have
Proof
Both sides of (5.8) are zero at the critical point of f. In the following, we consider regular points of f. Let x be a regular point of f, we denote by \(\{e_i\}_{i=1}^4\) a local orthonormal frame with \(e_1=\frac{\nabla f}{|\nabla f|}\). We use \(\{\alpha _i\}_{i=1}^4\) to represent eigenvalues of the Ricci tensor with corresponding orthonormal eigenvectors \(\{e_i\}_{i=1}^4\), respectively.
Note that
where we used (1.1).
Moreover, we have
where we used the second Bianchi identity. It follows that
where we used Proposition 5.2 in the second equality and (1.1) in the third. Moreover, we used (5.1) in the last equality. \(\square \)
6 Proof of Theorem 1.2
We finish the proof of Theorem 1.2 in this section. First of all, we prove a integral identity for four-dimensional gradient shrinking Ricci solitons with (4.1).
Lemma 6.1
Let \((M^4,f,g)\) be a four-dimensional gradient shrinking Ricci soliton, then we have
Proof
Since \(W^\pm _{ijkl}=\frac{1}{2}(W_{ijkl}\pm W_{i'j'kl})\), we only need to show that
For the complete non-compact case, let \(\phi (t)=\frac{s-t}{s}\) on (0, s), \(\phi =0\) on \([s,\infty )\) for any fixed constant \(s>0\). Moreover, the support of \(\phi (f)\) is a closed subset of the compact set \(\{x|\frac{1}{4}(r(x)-c_1)^2\le s\}\) for any fixed \(s>0\). Then we have \(\phi (f)\) having compact support in \(M^4\).
Using Proposition 5.3, we have
Next, we prove that \(\nabla _j(R_{i'j'kl}\nabla _i f\nabla _k f\phi (f)e^{-f})\nabla _lf\equiv 0\) on \(M^4\). We divide the arguments into two cases:
-
Case 1: \(|\nabla f|^2=0\) on some nonempty open set. In this case, since any gradient shrinking Ricci soliton is analytic in harmonic coordinates, it follows that \(|\nabla f|^2=0\) on \(M^4\). It is clear that \(\nabla _j(R_{i'j'kl}\nabla _i f\nabla _k f\phi (f)e^{-f})\nabla _lf\equiv 0\) on \(M^4\).
-
Case 2: The set \(\Theta :=\{x\in M^4|\nabla f(x)\ne 0\}\) is dense in \(M^4\). \(\forall x\in \Theta \), we denote by \(\{e_i\}_{i=1}^4\) a local orthonormal frame of with \(e_1=\frac{\nabla f}{|\nabla f|}\). Note that
$$\begin{aligned} \nabla _j(R_{i'j'kl}\nabla _i f\nabla _k f\phi (f)e^{-f})\nabla _lf=\nabla _j(R_{1'j'11}|\nabla f|^2\phi (f)e^{-f})|\nabla f|=0 \end{aligned}$$on \(\Theta \), the continuity implies that \(\nabla _j(R_{i'j'kl}\nabla _i f\nabla _k f\phi (f)e^{-f})\nabla _lf\equiv 0\) on \(M^4\).
It follows from (6.3) that
$$\begin{aligned} \int \nabla _j\nabla _lW_{i'j'kl}\nabla _if\nabla _kf\phi (f)e^{-f}=0 \end{aligned}$$(6.4)
From the proof of the complete non-compact case, we know that (6.2) also holds for the compact case.
This completes the proof of Lemma 6.1. \(\square \)
Next, we present a result that will be needed in the proof of Theorem 1.2.
Lemma 6.2
(Fernández-López et al. [8]) No complete gradient shrinking Ricci soliton may exist with \(R=\lambda \).
Now we are ready to finish the proof of Theorem 1.2.
Theorem 6.3
Let \((M^4,g,f)\) be a four-dimensional radially flat gradient shrinking Ricci soliton. If \(\text {div}^2W^\pm (\nabla f,\nabla f)=0\). Then \((M^4,g,f)\) is either Einstein or a finite quotient of \({\mathbb {R}}^4\), \({\mathbb {S}}^2\times {\mathbb {R}}^2\) or \({\mathbb {S}}^3\times {\mathbb {R}}\).
Proof
It follows from Lemma 6.1 that \(\int \text {div}^2W(\nabla f,\nabla f)e^{-f}=0\). From the proof of Theorem 1.1, we know that \((M^4,g,f)\) is rigid, i.e. it is Einstein, or a finite quotient of \({\mathbb {R}}^4\) or of the product \({\mathbb {R}}^k\times N^{4-k}\) \((k=1,2,3)\), where N is Einstein with Einstein constant \(\lambda \). We divide the arguments into three cases:
-
Case 1: \((M^4,g,f)\) is a finite quotient of \({\mathbb {R}}\times N^3\). In this case, \(N^3\) is a three-dimensional Einstein manifold with positive Einstein constant \(\lambda \). Note that the Weyl tensor vanishes identically in dimension 3, we can derive that \(N^3\) has positive constant sectional curvature \(\frac{3}{2}\lambda \). Therefore, \((M^4,g,f)\) is a finite quotient of \({\mathbb {R}}\times {\mathbb {S}}^3\).
-
Case 2: \((M^4,g,f)\) is a finite quotient of \({\mathbb {R}}^2\times N^2\). In this case, two-dimensional Einstein manifold \(N^2\) with positive Einstein constant \(\lambda \) must be \({\mathbb {S}}^2\).
-
Case 3: \((M^4,g,f)\) is a finite quotient of \({\mathbb {R}}^3\times N^1\). In this case, the scalar curvature of \(M^4\) is \(\lambda \), which contradicts Lemma 6.2.
To conclude, \((M^4,g,f)\) is either Einstein or a finite quotient of \({\mathbb {R}}^4\), \({\mathbb {S}}^2\times {\mathbb {R}}^2\) or \({\mathbb {S}}^3\times {\mathbb {R}}\). \(\square \)
References
Cao, H.D., Chen, Q.: On Bach-flat gradient shrinking Ricci solitons. Duke Math. J. 162(6), 1149–1169 (2013)
Cao, H.D., Zhou, D.: On complete gradient shrinking Ricci solitons. J. Differ. Geom. 85(2), 175–186 (2010)
Catino, G., Mastrolia, P., Monticelli, D.D.: Gradient Ricci solitons with vanishing conditions on Weyl. J. Math. Pure Appl. 108(1), 1–13 (2017)
Chen, B.L.: Strong uniqueness of the Ricci flow. J. Differ. Geom. 86(2), 362–382 (2009)
Chen, X., Wang, Y.: On four-dimensional anti-self-dual gradient Ricci solitons. J. Geom. Anal. 25, 1335–1343 (2015)
Eminenti, M., La Nave, G., Mantegazza, C.: Ricci solitons: the equation point of view. Manuscr. Math. 127, 345–367 (2008)
Fernández-López, M., García-Río, E.: Rigidity of shrinking Ricci solitons. Math. Z. 269(1), 461–466 (2011)
Fernández-López, M., García-Río, E.: On gradient Ricci solitons with constant scalar curvature. Proc. Am. Math. Soc. 144, 369–378 (2016)
Munteanu, O., Sesum, N.: On gradient Ricci solitons. J. Geom. Anal. 23, 539–561 (2013)
Naber, A.: Noncompact shrinking 4-solitons with nonnegative curvature. J. Reine Angew. Math. 645(2), 125–153 (2007)
Petersen, P., Wylie, W.: Rigidity of gradient Ricci solitons. Pac. J. Math. 241(2), 329–345 (2009)
Petersen, P., Wylie, W.: On the classification of gradient Ricci solitons. Geom. Topol. 14(4), 2277–2300 (2010)
Wu, J.Y., Wu, P., Wylie, W.: Gradient shrinking Ricci solitons of half harmonic Weyl curvature. Cal. Var. Partial Differ. 57, 141 (2018)
Yang, F., Zhang, L.: Rigidity of gradient shrinking and expanding Ricci solitons. Bull. Korean Math. Soc. 54(3), 817–824 (2017)
F. Yang, L. Zhang, Rigidity of gradient shrinking Ricci solitons. Asian J. Math. (to appear) (2019)
F. Yang, L. Zhang, On the classification of four-dimensional gradient Ricci solitons. arXiv: 1707.04846v2 [math.DG]
Zhang, Z.H.: Gadient shrinking solitons with vanishing Weyl tensor. Pac. J. Math. 242(1), 189–200 (2009)
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Young Jin Suh.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work is partially supported by the National Natural Science Foundation of China (Grant No. 71973130).
Rights and permissions
About this article
Cite this article
Yang, F., Zhang, L. & Ma, H. On Gradient Shrinking Ricci Solitons with Radial Conditions. Bull. Malays. Math. Sci. Soc. 44, 2161–2174 (2021). https://doi.org/10.1007/s40840-020-01058-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40840-020-01058-8