• 検索結果がありません。

1IntroductionandMainResults AnecessaryandsufficientconditionforsomesteadyRiccisolitonstohavepositiveasymptoticvolumeratio

N/A
N/A
Protected

Academic year: 2022

シェア "1IntroductionandMainResults AnecessaryandsufficientconditionforsomesteadyRiccisolitonstohavepositiveasymptoticvolumeratio"

Copied!
10
0
0

読み込み中.... (全文を見る)

全文

(1)

A necessary and sufficient condition for some steady Ricci solitons to have positive

asymptotic volume ratio

Xiang Gao

Abstract

In this paper, we firstly establish a useful ODE relationship between R1(c) and V1(c) on the steady Ricci soliton. Based on this, we obtain a necessary and sufficient condition for some complete noncompact steady gradient Ricci solitons to have positive asymptotic volume ratio.

1 Introduction and Main Results

Recall that a complete Riemannian manifold (Mn, g) is called a steady gra- dient Ricci soliton if there exists a smooth function f : Mn → R, called the potential function such that

Rij+∇ijf = 0. (1)

Moreover for the steady gradient Ricci soliton, we actually have

R+|∇f|2=C (2)

holds onMn, where C is a constant.

Key Words: steady Ricci soliton, asymptotic volume ratio, potential function.

2010 Mathematics Subject Classification: Primary 58C25; Secondary 35P05.

Received: March 2012.

Accepted: June 2012.

This work is supported by the Fundamental Research Funds for the Central Universities and NSFC11101267.

131

(2)

In [2], it was proved by B.-L. Chen that the complete ancient solutions to the Ricci flow, and in particular the steady Ricci soliton, must have nonnega- tive scalar curvature. As a consequence, the potential functionf satisfies the following estimate:

−√

Cr(x) +f(O)≤f(x)≤√

Cr(x) +f(O), (3)

wherer(x) denotes the distance function fromx to a fixed pointO inMn. Moreover, if we assume that the Ricci curvature is positive and the scalar curvature R approaches 0 towards spatial infinity, then by the following Lemma 1.1 proved by H. X. Guo in [5], we can derive that there is one point whereR obtains its maximum, and the point of maximum is unique.

Lemma 1.1(Guo). Let(Mn, g)be a steady gradient Ricci soliton with positive (or negative)Ricci curvature, then there is at most one critical point of R.

Thus we can denoteOthe unique point of maximum ofR, called the origin, and assumef(O) = 0 by adding a constant. Calculating the constant in (2) atO we have

R+|∇f|2=R(O) =R0. (4)

Based on these, H. X. Guo [5] also proved a more precise estimate for the potential function of a complete steady gradient Ricci soliton as follows:

Theorem 1.2 (Guo). Assume (Mn, g) is a complete steady gradient Ricci soliton with positive Ricci curvature, and the scalar curvature approaches 0 towards infinity. Then for any ε > 0, there exists rε > 0 such that when r(x)≥rε we have

pR0−ε

r(x)≤ −f(x)≤p

R0r(x), (5)

wherer(x) =d(x, O)andR0 is the maximum of R.

Then define the functions

V :R→[0,∞), R :R→[0,∞) by

V(c) = Z

{f <c}

dµ, R(c) = Z

{f <c}

R dµ.

In [1], the following ODE relating V(c) and R(c) was established for the shrink- ing Ricci soliton

0≤ n

2V(c)−R(c) =cV0(c)−R0(c). (6) In this paper, to prove our main result, we establish a similar result to (6) for the steady Ricci soliton as follows:

(3)

Theorem 1.3. Define

D1(c) ={x∈Mn|−f(x)< c}, R1(c) = Z

D1(c)

R dµ, V1(c) = Z

D1(c)

dµ,

then

R1(c) + R01(c)−R0V01(c) = 0. (7)

Recall that the asymptotic volume ratio (AVR) of a complete noncompact Riemannian manifold (Nn, h) is defined by

AVR(h) = lim

r→∞

VolhB(p, r)

ωnrn (8)

if the limit exists, whereB(p, r) denotes the geodesic ball inNn with center pand radius rand ωn is the volume of the unit Euclidean n-ball. It is easy to check that the AVR(h) is independent of the choice of p. Moreover, if (Nn, h) has nonnegative Ricci curvature, then this limit (8) exists by the Bishop-Gromov volume comparison theorem.

For the case of shrinking Ricci solitons, H.-D. Cao and D.-T. Zhou [1]

proved the following result aided by an observation of Munteanu [6].

Theorem 1.4 (Cao-Zhou). Any complete noncompact shrinking gradient linebreak Ricci soliton must have at most Euclidean volume growth, i.e.,

lim sup

r→∞

VolB(O, r)

ωnrn <∞. (9)

For the case of steady Ricci solitons, by using Theorem 1.3, we can prove the following estimate.

Theorem 1.5. For the steady gradient Ricci soliton we have V1(c)≥R1(c)

R0

+R0V1(c0)−R1(c0) R0

. (10)

In particular, more recently, observing the results in [1], [2], [4] and [8], B.

Chow, P. Lu and B. Yang [3] derived a necessary and sufficient condition for noncompact shrinking Ricci soliton to have positive AVR as follows:

(4)

Theorem 1.6 (Chow-Lu-Yang). Let (Mn, g) be a complete noncompact shrinking gradient Ricci soliton, then AVR(g) exists (and is finite). More- over, AVR(g)>0if and only if

Z

n+2

R(c)

cV(c)dc <∞. (11)

In this paper, for the case of steady Ricci solitons, we prove a similar necessary and sufficient condition for some noncompact steady solitons to have positive AVR:

Theorem 1.7. Let (Mn, g) be a complete noncompact steady gradient Ricci soliton such that the average scalar curvature

0<R (g) = lim

r→∞

R

B(O,r)Rdµ

Volg(B(O, r))<∞, (12) thenAVR(g) exists (and is finite). Moreover, AVR(g)>0if and only if

Z

c0

R1(c) Rc

0R1(s)ds−n+ 1 c

!

dc >−∞. (13)

The paper is organized as follows. In section 2, we prove Theorem 1.3 by calculating, and then obtain Theorem 1.5 applying Theorem 1.3. Based on these, in section 3, we prove our main result Theorem 1.7.

2 Proof of Theorem 1.3 and 1.5

Proof of Theorem 1.3. Firstly, by Theorem 1.2, when r(x) is greater than some constantrε, we have

pR0−ε

r(x)≤ −f(x)≤p R0r(x).

Denote by

D1(c) ={x∈Mn|−f(x)< c} and V1(c) = Z

D(c)

dV, (14) then by the Co-Area formula (cf. [7]), we have

V1(c) = Z c

0

ds Z

∂D1(s)

1

|∇(−f)|dA. (15)

(5)

Hence

V01(c) = Z

∂D1(c)

1

|∇f|dA. (16) Then taking the trace in

Rij+∇ijf = 0, we have

R+ ∆f = 0. (17)

Thus by using the Divergence Theorem and (4)

− Z

D1(c)

Rdµ= Z

D1(c)

∆f dµ

= Z

∂D1(c)

∇f ·−∇f

|∇f|dA

=− Z

∂D1(c)

|∇f|dA

= Z

∂D1(c)

R−R0

|∇f| dA

= Z

∂D1(c)

R

|∇f|dA−R0V01(c) Then by using the Co-Area formula again, we have

R1(c) = Z

D1(c)

R dµ= Z c

0

ds Z

∂D1(s)

R

|∇f|dA.

Hence

R01(c) = Z

∂D1(c)

R

|∇f|dA. (18) Therefore, we have

−R1(c) =− Z

D1(c)

Rdµ= Z

∂D1(c)

R

|∇f|dA−R0V01(c) = R01(c)−R0V01(c).

Now we turn to prove Theorem 1.5 by using Theorem 1.3.

(6)

Proof of Theorem 1.5. Integrate the identity (7) fromc0 toc we get R0(V1(c)−V1(c0)) =

Z c

c0

R0V01(s)ds

= Z c

c0

(R01(s) + R1(s))ds

= R1(c)−R1(c0) + Z c

c0

R1(s)ds

Therefore, (10) follows from the observation that R1(c) is nonnegative, because the scalar curvatureR≥0.

3 Proof of Theorem 1.7

In this section, we prove Theorem 1.7 by using Theorem 1.2 and 1.3.

Proof of Theorem 1.7. Let

P(c)=R1(c)−R0V1(c)

cn+1 and N (c) = R1(c)

R0V1(c), (19) then

N (c) N (c)−1 =

R1(c) R0V1(c) R1(c)

R0V1(c)−1 = R1(c)

R1(c)−R0V1(c) = R1(c)

cn+1P(c). (20) Note that RV1(c)

1(c) is the average scalar curvature over the set D(c), and the ODE (7) implies

P0(c) = R01(c) cn+1−(n + 1) R1(c)cn−R0V10 (c) cn+1+ (n + 1)R0V1(c) cn c2n+2

=(R01(c)−R0V01(c)) cn+1−(n + 1) cn(R1(c)−R0V1(c)) c2n+2

=−R1(c)cn+1−(n + 1) c2n+1P(c) c2n+2

=−N(c)−1N(c) c2n+2P(c)−(n + 1) c2n+1P(c) c2n+2

=−

N (c)

N (c)−1+n+ 1 c

P(c).

Then we choose c0 such that P(c0)6= 0, and integrate P0(c) =−

N (c)

N (c)−1 +n+ 1 c

P(c). (21)

(7)

from c0 to c we have

P(c) = P(c0)e

Rc

c0(N(c)−1N(c) +n+1c )dc. (22) From ODE (7) it is easy to see that

R1(c)−R0V1(c) =− Z c

0

R1(s)ds, which implies

R1(c)

cn+1P(c) = N (c)

N (c)−1 = R1(c)

R1(c)−R0V1(c) =− R1(c) Rc

0 R1(s)ds. (23) Note that (23) implies P(c)≤0. Furthermore, by the following bounds

pR0−ε

r(x)≤ −f(x)≤p R0r(x)

we have

c→∞lim P(c) = lim

c→∞

R1(c)−R0V1(c) cn+1

= lim

c→∞

R01(c)−R0V01(c) (n + 1) cn

=− lim

c→∞

R1(c) (n + 1) cn

=− 1 n + 1 lim

c→∞

R1(c) V1(c) lim

c→∞

V1(c) cn

=− 1

n + 1R (g) lim

c→∞

Vol(B O,c

R0

) cn

=− ωn

(n + 1)Rn/20

R (g) AVR(g), that is to say

− ωn

(n + 1)Rn/20

R (g) AVR(g) = lim

c→∞P(c), (24)

which exists by (21). Since the average scalar curvature R (g) = lim

r→∞

R

B(O,r)Rdµ Volg(B(O, r)) >0, we have AVR(g) exists (and is finite).

(8)

Moreover, by using (22) and (23) we have

AVR(g) =−(n + 1)Rn/20 P(c0) ωnR (g) e

R

c0(N(c)−1N(c) +n+1c )dc

=−(n + 1)Rn/20 P(c0) ωnR (g) e

R c0

R1(c) Rc

0R1(s)dsn+1c

dc

.

Note that P(c0) < 0, and by using (12) we obtain that AVR(g) >0 if and only if

Z

c0

R1(c) Rc

0R1(s)ds−n+ 1 c

!

dc >−∞.

4 Acknowledgment

I would especially like to thank the referee for meaningful suggestions that led to improvements of the article.

References

[1] H.-D. Cao, D. T. Zhou,On complete gradient shrinking Ricci solitons.J.

Diff. Geom.85, (2010), 175-185.

[2] B.-L. Chen,Strong uniqueness of the Ricci flow.J. Diff. Geom.82, (2009), 363-382.

[3] B. Chow, P. Lu, B. Yang,A necessary and sufficient condition for Ricci shrinkers to have positive AVR. Proc. Amer. Math. Soc. 140 (2012), no.

6, 2179-2181.

[4] F.-Q. Fang, J.-W. Man, Z.-L. Zhang, Complete gradient shrinking Ricci solitons have finite topological type.C. R. Math. Acad. Sci. Paris346, no.

11-12, (2008), 653-656.

[5] H. X. Guo, On the Ricci curvature of steady gradient Ricci solitons. J.

Math. Anal. Appl.363, (2010) 497-501.

[6] O. Munteanu, The volume growth of complete gradient shrinking Ricci solitons.arXiv: 0904. 0798.

(9)

[7] R. Schoen, S.-T. Yau,Lectures on Differential Geometry. Conference Pro- ceedings and Lecture Notes in Geometry and Topology, I. International Press, Cambridge, MA, (1994).

[8] S. J. Zhang,On a sharp volume estimate for gradient Ricci solitons with scalar curvature bounded below.Acta Math. Sinica27, (2011), 871-882.

Xiang Gao,

School of Mathematical Sciences, Ocean University of China,

Lane 238, Songling Road, Laoshan District, Qingdao City, Shandong Province, 266100, People’s Republic of China.

Email: [email protected]

(10)

参照

関連したドキュメント

In 2007, Lovejoy Das [5] in his paper proved that a second order symmetric parallel tensor on an α-K-contact (α ∈ R 0 ) manifold is a constant multiple of the associated metric

In the recent years, problems involving Kirchhoff type operators have been studied in many papers, we refer to [2, 9, 11, 15, 17, 19, 20] in which the authors have used

Some authors have used fixed point the- orems to show the existence of positive solutions to boundary value problems for ordinary differential equations, difference equations,

We consider a parametric Neumann problem driven by a nonlinear nonhomogeneous differential operator plus an indefinite potential term.. The reaction term is superlinear but does

Keywords: Discrete-time host-parasitoid model, Stability analysis, Steady state, Beverton-Holt equation, Nicholson Bailey model..

In [2] and [4], Berstein-Doetsch type results were proved on rationally s-convex functions, moreover, for the s-H¨ older property of s-convex functions.. Definition

We give a necessary and sufficient condition for a graph to be bipartite in terms of an eigenvector corresponding to the largest eigenvalue of the adjacency matrix of the graph..

We first apply the Cuadras identity to relax the monotonicity assumption of β(x) for a single random variable in the Chebyshev inequality, as shown in the following theorem:..