A Möbius scalar curvature rigidity on compact conformally flat hypersurfaces in \mathbb{S}^{n+1}

Limiao Lin, Tongzhu Li and Changping Wang

arXiv: 1709.01658v1 · PDF · 6 September 2017
Categories: math.DG
Licence: CC BY 4.0 The site shows the text.

Abstract

In this paper, we study conformally flat hypersurfaces of dimension n(\geq 4) in \mathbb{S}^{n+1} using the framework of Möbius geometry. First, we classify and explicitly express the conformally flat hypersurfaces of dimension n(\geq 4) with constant Möbius scalar curvature under the Möbius transformation group of \mathbb{S}^{n+1}. Second, we prove that if the conformally flat hypersurface with constant Möbius scalar curvature R is compact, then

R=(n-1)(n-2)r^2, ~~0<r<1,

and the compact conformally flat hypersurface is Möbius equivalent to the torus

\mathbb{ S}^1(\sqrt{1-r^2})\times \mathbb{S}^{n-1}(r)\hookrightarrow \mathbb{S}^{n+1}.

Items 25

Theorem

Theorem 1.1

BYRV

Let f:M^n\rightarrow \mathbb{S}^{n+1}, n\geq 4, be a conformally flat hypersurface without umbilical points. If the Möbius scalar curvature is constant, then the Möbius form is closed and f is Möbius equivalent to one of the following hypersurfaces in \mathbb{S}^{n+1},
(i) the image of \sigma of a cylinder over a curvature-spiral in \mathbb{R}^2\subset \mathbb{R}^{n+1};
(ii) the image of \sigma of a cone over a curvature-spiral in \mathbb{S}^2\subset \mathbb{R}^3\subset \mathbb{R}^{n+1};
(iii) the image of \sigma of a rotational hypersurface over a curvature-spiral in \mathbb{R}^2_+\subset \mathbb{R}^{n+1}.

Theorem

Theorem 1.2

HWWK

Let f:M^n\rightarrow \mathbb{S}^{n+1}, n\geq 4, be a compact conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant, then

R=(n-1)(n-2)r^2, ~~0<r<1,

and the compact conformally flat hypersurface is Möbius equivalent to the torus

f:\mathbb{ S}^1(\sqrt{1-r^2})\times \mathbb{S}^{n-1}(r)\to \mathbb{S}^{n+1}.

Remark

Remark 1.1

BTHS

Theorem 1.1 and Theorem 1.2 is true for n=3 provided that the 3-dimensional conformally flat hypersurface has only two distinct principal curvatures.

Theorem

Theorem 2.1

DBXE

\cite{w} Two hypersurfaces f: M^n\rightarrow \mathbb{S}^{n+1} and \bar{f}: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 3) are Möbius equivalent if and only if there exists a diffeomorphism \varphi: M^n\rightarrow M^n which preserves the Möbius metric and the Möbius second fundamental form.

Lemma

Lemma 3.1

QOWJ

The metric g=\kappa^2(s)(ds^2+I_{-\epsilon}^{n-1}) given above is of constant scalar curvature R if and only if the function \kappa(s) satisfies

-\frac{\kappa_{ss}}{\kappa^3}+\frac{(n+2)\kappa_s^2}{2\kappa^4}+\epsilon\frac{n-2}{2\kappa^2} =R,~~~~\kappa_s=\frac{d}{ds}\kappa.

Example

Example 3.1

4NN1

Let \gamma:I\rightarrow \mathbb{R}^2 be a regular curve, and s denote the arclength of \gamma(s). we define cylinder in \mathbb{R}^{n+1} over \gamma,

f(s,y)=(\gamma(s),y):I\times \mathbb{R}^{n-1}\longrightarrow \mathbb{R}^{n+1},

where y:\mathbb{R}^{n-1}\longrightarrow \mathbb{R}^{n-1} is identical maping.

Example

Example 3.2

JHT7

Let \gamma:I\rightarrow \mathbb{S}^2(1)\subset \mathbb{R}^3 be a regular curve, and s denote the arclength of \gamma(s). we define cone in R^{n+1} over \gamma,

f(s,t,y)=(t\gamma(s),y):I\times \mathbb{R}^{+}\times \mathbb{R}^{n-2}\longrightarrow \mathbb{R}^{n+1},

where y:\mathbb{R}^{n-2}\longrightarrow \mathbb{R}^{n-2} is identical mapping and \mathbb{R}^{+}=\{t|t>0\}.

Example

Example 3.3

PKVR

Let \mathbb{R}^2_+=\{(x,y)\in \mathbb{R}^2|y>0\} be the upper half-space endowed with the standard hyperbolic metric

ds^2=\frac{1}{y^2}[dx^2+dy^2]~.

Let \gamma=(x,y):I\longrightarrow \mathbb{R}^2_+ be a regular curve, and s denote the arclength of \gamma(s). we define rotational hypersurface in \mathbb{R}^{n+1} over \gamma,

f:I\times \mathbb{S}^{n-1}\longrightarrow \mathbb{R}^{n+1},~~~~~ f(x,y,\theta)=(x,y\theta),

where \theta:\mathbb{S}^{n-1}\longrightarrow \mathbb{R}^{n} is a standard immersion of a round sphere.

Lemma

Lemma 3.2

UJ8I

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a conformally flat hypersurface without umbilical points. If the Möbius scalar curvature is constant, then we can choose a local orthonormal basis \{E_1,\cdots,E_n\} with respect to the Möbius metric g such that

(B_{ij})=diag\{\frac{n-1}{n},\frac{-1}{n},\cdots,\frac{-1}{n}\},~~ (A_{ij})=diag\{a_1,a_2,\cdots,a_2\}.

Moreover, the distribution \mathbb{D}=span\{E_2,\cdots,E_n\} is integrable.

Used by Proof, Proof

Proof of Lemma 3.2

Using dB_{ij}+\sum_kB_{kj}\omega_{ki}+\sum_kB_{ik}\omega_{kj}=\sum_kB_{ij,k}\omega_k , the equation (?), we get

\begin{split} &B_{1\alpha,\alpha}=-C_1, ~~otherwise, ~~~B_{ij,k}=0;\\ &\omega_{1\alpha}=-C_1\omega_{\alpha},~~~~C_{\alpha}=0. \end{split}

Thus d\omega_1=0 and the distribution \mathbb{D}=span\{E_2,\cdots,E_n\} is integrable.

Using dC_i+\sum_kC_k\omega_{ki}=\sum_kC_{i,k}\omega_k and (Proof), we can obtain

C_{\alpha,\alpha}=-C_1^2, ~~~C_{\alpha,k}=0,\alpha\neq k.

From (Proof),

\begin{split} &d\omega_{1\alpha}=-dC_1\wedge\omega_{\alpha}-C_1d\omega_{\alpha}\\ &=-dC_1\wedge\omega_{\alpha}+C_1^2\omega_1\wedge\omega_{\alpha}-C_1\sum_{\gamma}\omega_{\gamma}\wedge\omega_{\gamma\alpha}, \end{split}

and d\omega_{1\alpha}-\sum_j\omega_{1j}\wedge\omega_{j\alpha}=-\frac{1}{2}\sum_{kl}R_{1\alpha kl}\omega_k\wedge\omega_l, we get that

R_{1\alpha 1\alpha}=C_{1,1}-C_1^2,~~~~R_{1\alpha\beta\alpha}-C_{1,\beta}=0.

Since R_{1\alpha 1\alpha}=-\frac{n-1}{n^2}+a_1+a_{\alpha}=C_{1,1}-C_1^2 and R_{1\alpha\beta\alpha}=A_{1\beta},\alpha\neq\beta, thus we have

a_2=a_3=\cdots=a_n,~~~~ A_{1\beta}=C_{1,\beta}.

Thus A|_{\mathbb{D}}=aI,~~a=a_2. Since E_1 is principal vector field, then vector E=A_{12}E_2+\cdots+A_{1n}E_n is well defined. If E=0, then A_{12}=\cdots=A_{1n}=0. If E\neq 0, we can rechoose a local orthonormal basis \{\tilde{E}_2=\frac{E}{|E|},\tilde{E}_3,\cdots,\tilde{E}_n\} of \mathbb{D} with respect to the Möbius metric g such that

(B_{ij})=diag(\frac{n-1}{n},\frac{-1}{n},\cdots,\frac{-1}{n}),~~ (A_{ij})=\left( \begin{array}{ccccc} A_{11} & A_{12} & 0 & \cdots & 0 \\ A_{21} & a_2 & 0 & \cdots & 0 \\ 0& 0 & a_2& \cdots & 0 \\ \vdots & \vdots & \vdots& \ddots & \vdots \\ 0 & 0 & 0 & \cdots &a_2 \\ \end{array} \right)

To finish the proof of the Lemma, we need to prove that A_{12}=0. Using dA_{ij}+\sum_kA_{kj}\omega_{ki}+\sum_kA_{ik}\omega_{kj}=\sum_kA_{ij,k}\omega_k , the equation (?) and (Proof), we get

\begin{split} &\sum_mA_{12,m}\omega_m=dA_{12}+(A_{11}-A_{22})\omega_{12},\\ &\sum_mA_{1\alpha,m}\omega_m=(A_{11}-a_2)\omega_{1\alpha}+A_{12}\omega_{2\alpha},~~\sum_mA_{2\alpha,m}\omega_m=A_{12}\omega_{1\alpha},~~\alpha\geq3,\\ &\sum_kA_{11,k}\omega_k=dA_{11}+2A_{12}\omega_{21}, ~~\sum_kA_{22,k}\omega_k=dA_{22}+2A_{12}\omega_{12},\\ &\sum_kA_{\alpha\alpha,k}\omega_k=dA_{\alpha\alpha},~~A_{\alpha\beta,k}=0, ~~\alpha\neq\beta, ~~\alpha,\beta\geq 3. \end{split}

From the fourth and seventh equation in (Proof), we get

E_{\alpha}(a_2)=A_{\beta\beta,\alpha}=A_{\beta\alpha,\beta}=0, ~~~\alpha\geq 3.

Since the Möbius scalar curvature is constant, tr(A)=A_{11}+(n-1)a_2 is constant. Thus

A_{1\alpha,1}=A_{11,\alpha}=E_{\alpha}(A_{11})=0,~~~\alpha\geq 3.

From the first, second and third equation in (Proof), we get

A_{12,2}=E_2(A_{12})-(A_{11}-a_2)C_1,~~~A_{1\beta,\beta}=-(A_{11}-a_2)C_1+A_{12}\omega_{2\beta}(E_{\beta}).

On the other hand, From (?), we have

\begin{split} &E_1(A_{22})=A_{22,1}=A_{12,2}+\frac{1}{n}C_1=E_2(A_{12})-(A_{11}-a_2)C_1+\frac{1}{n}C_1,\\ &E_1(A_{\alpha\alpha})=A_{\alpha\alpha,1}=A_{1\alpha,\alpha}+\frac{1}{n}C_1=-(A_{11}-a_2)C_1+A_{12}\omega_{2\beta}(E_{\beta})+\frac{1}{n}C_1,\\ \end{split}

which implies that

E_2(A_{12})=A_{12}\omega_{2\beta}(E_{\beta}).

Since A_{1\alpha,\beta}=A_{\alpha\beta,1}=A_{1\beta,1}=A_{12,\alpha}=0,~\alpha\neq\beta, from the second equation in (Proof) we can obtain

A_{12}\omega_{2\beta}(E_k)=0, ~~\beta\geq 3,~~k\neq\beta.

From the first and third equation in (Proof), we get

\begin{split} E_2(a_2)&=E_2(A_{\beta\beta})=A_{\beta\beta,2}=A_{2\beta,\beta}=-A_{12}C_1,\\ E_1(A_{12})&=A_{12,1}=A_{11,2}=E_2(A_{11})+2A_{12}C_1\\ &=E_2(-(n-1)a_2)+2A_{12}C_1=(n+1)A_{12}C_1. \end{split}

Now we assume that A_{12}\neq 0, From (Proof) and (Proof), we have

\omega_{2\alpha}=\frac{E_2(A_{12})}{A_{12}}\omega_{\alpha}:=\mu\omega_{\alpha},~~\alpha\geq 3.

Thus

\begin{split} &d\omega_{2\alpha}=d\mu\wedge\omega_{\alpha}+\mu d\omega_{\alpha}\\ &=d\mu\wedge\omega_{\alpha}-\mu C_1^2\omega_1\wedge\omega_{\alpha}+\mu^2\omega_2\wedge\omega_{\alpha} +\mu\sum_{\gamma\geq3}\omega_{\gamma}\wedge\omega_{\gamma\alpha}. \end{split}

Using d\omega_{2\alpha}-\sum_j\omega_{2j}\wedge\omega_{j\alpha}=-\frac{1}{2}\sum_{kl}R_{2\alpha kl}\omega_k\wedge\omega_l, we get that

E_(\mu)-\mu C_1=-A_{12}.

On the other hand, using (Proof) and (Proof), we have

\begin{split} E_1(\mu)&=E_1\big[\frac{E_2(A_{12})}{A_{12}}\big]=\frac{E_1E_2(A_{12})}{A_{12}}-\frac{E_2(A_{12})E_1(A_{12})}{A_{12}^2}\\ &=\frac{E_1E_2(A_{12})}{A_{12}}-(n+1)\frac{E_2(A_{12})C_1}{A_{12}}\\ &=\frac{(E_2E_1+C_1E_2)(A_{12})}{A_{12}}-(n+1)\frac{E_2(A_{12})C_1}{A_{12}}\\ &=\frac{E_2[(n+1)A_{12}C_1]}{A_{12}}-n\frac{E_2(A_{12})C_1}{A_{12}}\\ &=(n+1)C_{1,2}+\frac{E_2(A_{12})C_1}{A_{12}}, \end{split}

which implies that

(n+1)C_{1,2}=-A_{12}.

This is a contradiction by A_{12}=C_{1,2}. Therefore A_{12}=0 and we finish the proof.

Theorem

Theorem 4.1

KFE1

[lih2] Let f: M^n\to \mathbb{S}^{n+1} be a Möbius isoparametric hypersurface with two distinct principal curvatures. Then f is Möbius equivalent to an open part of one of the following Möbius isoparametric hypersurfaces in \mathbb{S}^{n+1}:
(i) the standard torus \mathbb{S}^k(r)\times \mathbb{S}^{n-k}(\sqrt{1-r^2});
(ii) the image of \sigma of the standard cylinder \mathbb{S}^k(1)\times \mathbb{R}^{n-k}\subset \mathbb{R}^{n+1};
(iii) the image of \tau of the standard \mathbb{S}^k(r)\times \mathbb{H}^{n-k}(\sqrt{1+r^2}) in \mathbb{H}^{n+1}.

Lemma

Lemma 4.1

1JLC

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant. then under the local orthonormal basis \{E_1,\cdots,E_n\} in Lemma 3.2, we have

\begin{split} &a_1=\frac{2n-1}{2n^2}-\frac{R}{2(n-1)(n-2)}+\frac{n-1}{n-2}(C_{1,1}-C_1^2),\\ &a_2=\frac{R}{2(n-1)(n-2)}-\frac{1}{2n^2}-\frac{1}{n-2}(C_{1,1}-C_1^2),\\ &A_{\alpha\alpha,1}=\frac{R}{(n-1)(n-2)}C_1-\frac{n}{n-2}(C_1C_{1,1}-C_1^3). \end{split}

Except these coefficients A_{11,1}, A_{1\alpha,\alpha} and A_{\alpha\alpha,1} the coefficients of \nabla A are equal to zero.

Used by Proof, Proof

Proof of Lemma 4.1

The first and second equation in (4.1) can derive directly from the equation tr(A)=a_1+(n-1)a_2=\frac{1}{2n}+\frac{R}{2(n-1)} and R_{1\alpha1\alpha}=-\frac{n-1}{n^2}+a_1+a_2=C_{1,1}-C_1^2 in (Proof).

From (Proof), we can get

A_{1\alpha,\alpha}=(a_2-a_1)C_1=[\frac{R}{(n-1)(n-2)}-\frac{1}{n}]C_1-\frac{n}{n-2}(C_1C_{1,1}-C_1^3).

By (?), we have A_{\alpha\alpha,1}=A_{1\alpha,\alpha}+\frac{1}{n}C_1. Combining above equation we get the third equation in (4.1).

Since tr(A)=a_1+(n-1)a_2 is constant, we have A_{11,1}=-(n-1)A_{\alpha\alpha,1}. Thus, by lemma 3.2, we know that except these coefficients A_{11,1}, A_{1\alpha,\alpha} and A_{\alpha\alpha,1} the coefficients of \nabla A are equal to zero.

Uses Proof, Lemma 3.2

Lemma

Lemma 4.2

OVDT

\begin{split} &C_{1,11}=E_1(C_{1,1})=(n+2)C_1C_{1,1}-nC_1^3-\frac{R}{n-1}C_1,\\ &C_{\alpha,\alpha1}=E_1(C_{\alpha,\alpha})=-2C_1C_{1,1},~~C_{1,\alpha\alpha}=C_{\alpha,1\alpha}=-(C_{1,1}+C_1^2)C_1. \end{split}

Except these coefficients C_{1,11}, C_{1,\alpha\alpha} and C_{\alpha,\alpha1} the coefficients of \nabla^2C are equal to zero.

Used by Proof, Proof

Proof of Lemma 4.2

Since (A_{ij})=diag\{a_1,a_2,\cdots,a_2\} under the local orthonormal basis, we have A_{\alpha\alpha,1}=E_1(A_{\alpha\alpha})=E_1(a_2)=-\frac{1}{n-2}(C_{1,11}-2C_1C_{1,1}) by Lemma 4.1, combining the first equation in (4.1), we get the first equation in (4.2).

By the equation (Proof) and the equation (Proof), (C_{i,j})=diag\big(C_{1,1}, -C_1^2,\cdots,-C_1^2\big) under the local orthonormal basis, thus we have

C_{\alpha,\alpha1}=E_1(C_{\alpha,\alpha})=-2C_1C_{1,1},~~C_{1,\alpha\alpha}=C_{\alpha,1\alpha}=-(C_{1,1}+C_1^2)C_1.

And the rest coefficients of \nabla^2C are zero.

Uses Lemma 4.1, Proof

Lemma

Lemma 4.3

2PJW

\begin{split} &S_{11,11}=-RC_{1,1}+n(n-1)C_{1,1}^2+n(n-1)^2C_1^2C_{1,1}-n^2(n-1)C_1^4-nRC_1^2,\\ &S_{11,\alpha\alpha}=\frac{(n+1)R}{n-1}C_1^2-n(n+1)[C_1^2C_{1,1}-C_1^4],~~S_{\alpha\alpha,11}=-(n-1)S_{11,11},\\ &S_{\alpha\alpha,\alpha\alpha}=3S_{\alpha\alpha,\beta\beta}=3\{\frac{-R}{n-1}C_1^2+n[C_1^2C_{1,1}-C_1^4]\},~~\alpha\neq\beta. \end{split}

Used by Proof

Proof of Lemma 4.3

Since (S_{ij})=diag(S_1,S_2,\cdots,S_2), we know that except these coefficients S_{11,1}, S_{1\alpha,\alpha} and S_{\alpha\alpha,1} the coefficients of \nabla S are equal to zero. Using the definition of the second covariant derivative of S, we can compute these equations in (4.3).

Lemma

Lemma 4.4

1LDX

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a compact conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant. then

\begin{split} &\int_{M^n}C_1^2C_{1,1}dV_g=\frac{n-1}{3}\int_{M^n}|C|^4dV_g,\\ &\int_{M^n}C_{1,1}^2dV_g=\int_{M^n}|C|^4dV_g+\frac{R}{n-1}\int_{M^n}|C|^2dV_g. \end{split}

Used by Proof, Proof, Proof

Proof of Lemma 4.4

Using the coefficients of the tensor C and S, we define two smooth vector fields

X_S=\sum_{ij}C_iS_{ij}E_j, ~~X_C=\sum_{ij}C_iE_i.

From Lemma 3.2 and the equation (4.1), we can get the divergence of X_S, X_C,

\begin{split} &div X_C=\sum_iC_{i,i}=C_{1,1}-(n-1)C_1^2,\\ &div X_S=(n-1)(C_{1,1}^2-C_1^4-\frac{R}{n-1}C_1^2)+\frac{R}{2(n-1)}div X_C. \end{split}

Since the hypersurface is compact, we have

\begin{split} &\int_{M^n}C_{1,1}dV_g=(n-1)\int_{M^n}|C|^2dV_g,\\ &\int_{M^n}C_{1,1}^2dV_g=\int_{M^n}|C|^4dV_g+\frac{R}{n-1}\int_{M^n}|C|^2dV_g, \end{split}

On the other hand, we compute \triangle |C|^2,

\begin{split} &\triangle|C|^2=\sum_i(E_iE_i-\nabla_{E_i}E_i) |C|^2=\sum_i(E_iE_i-\nabla_{E_i}E_i)C_1^2\\ &=E_1E_1(C_1^2)-\sum_i\nabla_{E_i}E_i(C_1^2)=2C_{1,1}^2+2C_1C_{1,11}-2(n-1)C_1^2C_{1,1}\\ &=2C_{1,1}^2+6C_1^2C_{1,1}-2nC_1^4-\frac{2R}{n-1}C_1^2. \end{split}

Since the hypersurface is compact, we have

\int_{M^n}C_{1,1}^2dV_g+3\int_{M^n}C_1^2C_{1,1}dV_g-n\int_{M^n}C_1^4dV_g-\frac{R}{n-1}\int_{M^n}|C|^2dV_g=0.

Combining the equation (Proof), we can derive the first equation in (4.4).

Uses Lemma 3.2, Lemma 4.1

Lemma

Lemma 4.5

XJIJ

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a compact conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant. then

\int_{M^n}\big\{C_{1,1}^3+(n+5)C_1^2C_{1,1}^2-(2n+5)C_1^4C_{1,1}+(n-1)C_1^6-\frac{2R}{3}C_1^4\big\}dV_g=0.

Proof of Lemma 4.5

Using the coefficients of the tensor C and S, we define a smooth vector field

Y_S=\sum_{ijk}C_{i,j}S_{ij,k}E_k.

Using Lemma 4.2 and the equation (4.3), we compute the divergence of Y_S,

\begin{split} &\frac{1}{n(n-1)}div Y_S=\sum_{ijk}C_{i,jk}S_{ij,k}+\sum_{ijk}C_{i,j}S_{ij,kk},\\ &=C_{1,1}^3+(n+5)C_1^2C_{1,1}^2-(2n+5)C_1^4C_{1,1}+(n-1)C_1^6+\frac{(2n-1)R}{n(n-1)}C_1^4\\ &-\frac{R}{n(n-1)}C_{1,1}^2-\frac{2(n+3)R}{n(n-1)}C_1^2C_{1,1}+\frac{R^2}{n(n-1)^2}C_1^2. \end{split}

Integrating this equation and using (4.4), we can derive the second equation in (4.5).

Uses Lemma 4.2, Lemma 4.3, Lemma 4.4

Lemma

Lemma 4.6

PWQW

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a compact conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant. then

\begin{split} &\int_{M^n}\big\{C_1^2C_{1,1}^2+C_1^4C_{1,1}-\frac{n}{3}C_1^6-\frac{R}{3(n-1)}C_1^4\big\}dV_g=0,\\ &\int_{M^n}\big\{C_{1,1}^3+(n-1)C_1^2C_{1,1}^2-(2n+1)C_1^4C_{1,1}+(n+1)C_1^6+\frac{2(n+1)R}{3(n-2)}C_1^4\big\}dV_g=0. \end{split}

Proof of Lemma 4.6

Using (4.2),

\begin{split} \frac{1}{4}\triangle |C|^4&=\frac{1}{4}\sum_i(E_iE_i-\nabla E_i{E_i})|C|^4=3C_1^2C_{1,1}^2+C_1^3C_{1,11}-(n-1)C_1^4C_{1,1}\\ &=3C_1^2C_{1,1}^2+3C_1^4C_{1,1}-nC_1^6-\frac{R}{n-1}C_1^4. \end{split}

Since the hypersurface is compact, then

\int_{M^n}\{3C_1^2C_{1,1}^2+3C_1^4C_{1,1}-nC_1^6-\frac{R}{n-1}C_1^4\}dV_g=0.

Since S_{ij} is a Codazzi tensor and tr(S) is constant, we can compute \triangle |S|^2 by (?),

\begin{split} &\frac{1}{n^2(n-1)}\triangle |S|^2=\frac{1}{2n^2(n-1)}\{\sum_{ijk}|S_{ij,k}|^2+\frac{1}{2}\sum_{ij}(S_i-S_j)^2R_{ijij}\}\\ &=C_{1,1}^3-(2n+1)C_1^4C_{1,1}+(n-1)C_1^2C_{1,1}^2-\frac{2R}{n-1}C_1^2C_{1,1}-\frac{2R}{n(n-1)}C_{1,1}^2\\ &+(n+1)C_1^6+\frac{2(n+1)R}{n(n-1)}C_1^4+\frac{(n+1)R^2}{n^2(n-1)^2}C_1^2+\frac{R^2}{n^2(n-1)^2}C_{1,1}. \end{split}

Integrating this equation and using (4.4), we can derive the second equation in (4.6).

Uses Lemma 4.2, Lemma 4.4

Lemma

Lemma 4.7

VGG9

Let f: M^n\rightarrow \mathbb{S}^{n+1} (n\geq 4) be a compact conformally flat hypersurface without umbilical points everywhere. If the Möbius scalar curvature R is constant. then

\begin{split} &\int_{M^n}\big\{C_{1,1}^3+\frac{5n+4}{2}C_1^2C_{1,1}^2-3(n+1)C_1^4C_{1,1}+\frac{n}{2}C_1^6-\frac{(7n-10)R}{3(n-2)}C_1^4\big\}dV_g=0£¬\\ &\int_{M^n}\big\{C_{1,1}^3+\frac{5n+16}{2}C_1^2C_{1,1}^2-3(n-1)C_1^4C_{1,1}-\frac{3n}{2}C_1^6-\frac{(7n+2)R}{6(n-1)}C_1^4\big\}dV_g=0. \end{split}

Proof of Lemma 4.7

Using the coefficients of the tensor C and A, we have two following smooth functions,

|C|^2_A=\sum_{ij}C_iA_{ij}C_j=C_1^2a_1, ~~|C|^2_C=\sum_{ij}C_iC_{i,j}C_j=C_1^2C_{1,1}.

Next we compute \triangle (|C|^2_A) and \triangle(|C|^2_C).

\begin{split} &\frac{n-2}{2(n-1)}\triangle (|C|^2_A)=\frac{n-2}{2(n-1)}(E_1E_1(C_1^2a_1)-(n-1)C_1E_1(C_1^2a_1))\\ &=C_{1,1}^3-3(n+1)C_1^4C_{1,1}+\frac{(5n+4)}{2}C_1^2C_{1,1}^2+\frac{n}{2}C_1^6\\ &+\frac{n-2}{2(n-1)}[\frac{2n-1}{n^2}-\frac{R}{(n-1)(n-2)}]C_{1,1}^2-\frac{n-2}{2(n-1)}[\frac{2n-1}{n}-\frac{(2n-1)R}{(n-1)(n-2)}]C_1^4\\ &\frac{n-2}{2(n-1)}[\frac{3(2n-1)}{n^2}-\frac{(7n-4)R}{(n-1)(n-2)}]C_1^2C_{1,1}\\ &-\frac{n-2}{2(n-1)}[\frac{2n-1}{n^2}-\frac{R}{(n-1)(n-2)}]\frac{R}{n-1}C_1^2. \end{split}

Integrating this equation and using (4.4), we can derive the first equation in (4.7).

\begin{split} &\frac{1}{2}\triangle (|C|^2_C)=\frac{1}{2}(E_1E_1(C_1^2C_{1,1})-(n-1)C_1E_1(C_1^2C_{1,1}))\\ &=C_{1,1}^3-3(n-1)C_1^4C_{1,1}+\frac{(5n+16)}{2}C_1^2C_{1,1}^2-\frac{3n}{2}C_1^6\\ &-\frac{7R}{2(n-1)}C_1^2C_{1,1}-\frac{3R}{2(n-1)}C_1^4. \end{split}

Integrating this equation and using (4.4), we can derive the second equation in (4.7).

Uses Lemma 4.4

Source and licence

The text on this page is from arXiv:1709.01658v1 by Limiao Lin, Tongzhu Li and Changping Wang, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.

The record of each item is in analysis/objects/1709/1709.01658.jsonl.