The entropy formula for the Ricci flow
and its geometric applications

Consider the functional $\mathcal{F}=\int_{M}{(R+|\nabla f|^{2})e^{-f}dV}$ for a riemannian metric $g_{ij}$ and a function $f$ on a closed manifold $M$. Its first variation can be expressed as follows:

where $\delta g_{ij}=v_{ij}$, $\delta f=h$, $v=g^{ij}v_{ij}$. Notice that $v/2-h$ vanishes identically iff the measure $dm=e^{-f}dV$ is kept fixed. Therefore, the symmetric tensor $-(R_{ij}+\nabla_{i}\nabla_{j}f)$ is the $L^{2}$ gradient of the functional $\mathcal{F}^{m}=\int_{M}{(R+|\nabla f|^{2})dm}$, where now $f$ denotes $\log(dV/dm)$. Thus given a measure $m$ , we may consider the gradient flow $(g_{ij})_{t}=-2(R_{ij}+\nabla_{i}\nabla_{j}f)$ for $\mathcal{F}^{m}$. For general $m$ this flow may not exist even for short time; however, when it exists, it is just the Ricci flow, modified by a diffeomorphism. The remarkable fact here is that different choices of $m$ lead to the same flow, up to a diffeomorphism; that is, the choice of $m$ is analogous to the choice of gauge.

A metric $g_{ij}(t)$ evolving by the Ricci flow is called a breather, if for some $t_{1}<t_{2}$ and $\alpha>0$ the metrics $\alpha g_{ij}(t_{1})$ and $g_{ij}(t_{2})$ differ only by a diffeomorphism; the cases $\alpha=1,\alpha<1,\alpha>1$ correspond to steady, shrinking and expanding breathers, respectively. Trivial breathers, for which the metrics $g_{ij}(t_{1})$ and $g_{ij}(t_{2})$ differ only by diffeomorphism and scaling for each pair of $t_{1}$ and $t_{2}$, are called Ricci solitons. (Thus, if one considers Ricci flow as a dynamical system on the space of riemannian metrics modulo diffeomorphism and scaling, then breathers and solitons correspond to periodic orbits and fixed points respectively). At each time the Ricci soliton metric satisfies an equation of the form $R_{ij}+cg_{ij}+\nabla_{i}b_{j}+\nabla_{j}b_{i}=0,$ where $c$ is a number and $b_{i}$ is a one-form; in particular, when $b_{i}=\frac{1}{2}\nabla_{i}a$ for some function $a$ on $M,$ we get a gradient Ricci soliton. An important example of a gradient shrinking soliton is the Gaussian soliton, for which the metric $g_{ij}$ is just the euclidean metric on $\mathbb{R}^{n}$, $c=1$ and $a=-|x|^{2}/2.$

In this and the next section we use the gradient interpretation of the Ricci flow to rule out nontrivial breathers (on closed $M$). The argument in the steady case is pretty straightforward; the expanding case is a little bit more subtle, because our functional $\mathcal{F}$ is not scale invariant. The more difficult shrinking case is discussed in section 3.

Define $\lambda(g_{ij})=\mbox{inf}\ \mathcal{F}(g_{ij},f),$ where infimum is taken over all smooth $f,$ satisfying $\int_{M}{e^{-f}dV}=1.$ Clearly, $\lambda(g_{ij})$ is just the lowest eigenvalue of the operator $-4\triangle+R.$ Then formula (1.4) implies that $\lambda(g_{ij}(t))$ is nondecreasing in $t,$ and moreover, if $\lambda(t_{1})=\lambda(t_{2}),$ then for $t\in[t_{1},t_{2}]$ we have $R_{ij}+\nabla_{i}\nabla_{j}f=0$ for $f$ which minimizes $\mathcal{F}.$ Thus a steady breather is necessarily a steady soliton.

To deal with the expanding case consider a scale invariant version $\bar{\lambda}(g_{ij})=\lambda(g_{ij})V^{2/n}(g_{ij}).$ The nontrivial expanding breathers will be ruled out once we prove the following

Claim $\bar{\lambda}$is nondecreasing along the Ricci flow whenever it is nonpositive; moreover, the monotonicity is strict unless we are on a gradient soliton.

(Indeed, on an expanding breather we would necessarily have $dV/dt>0$ for some $t{\in}[t_{1},t_{2}].$ On the other hand, for every $t$, $-\frac{d}{dt}\mbox{log}V=\frac{1}{V}\int{RdV}\geq\lambda(t),$ so $\bar{\lambda}$ can not be nonnegative everywhere on $[t_{1},t_{2}],$ and the claim applies.)

Proof of the claim.

where $f$ is the minimizer for $\mathcal{F}.$

The arguments above also show that there are no nontrivial (that is with non-constant Ricci curvature) steady or expanding Ricci solitons (on closed $M$). Indeed, the equality case in the chain of inequalities above requires that $R+\triangle f$ be constant on $M$; on the other hand, the Euler-Lagrange equation for the minimizer $f$ is $2\triangle f-|\nabla f|^{2}+R=const.$ Thus, $\triangle f-|\nabla f|^{2}=const=0$, because $\int{(\triangle f-|\nabla f|^{2})e^{-f}dV}=0.$ Therefore, $f$ is constant by the maximum principle.

In order to handle the shrinking case when $\lambda>0,$ we need to replace our functional $\mathcal{F}$ by its generalization, which contains explicit insertions of the scale parameter, to be denoted by $\tau.$ Thus consider the functional

restricted to $f$ satisfying

$\tau>0.$ Clearly ${\mathcal{W}}$ is invariant under simultaneous scaling of $\tau$ and $g_{ij}.$ The evolution equations, generalizing (1.3) are

The evolution equation for $f$ can also be written as follows: $\Box^{*}u=0,$ where $u=(4\pi\tau)^{-\frac{n}{2}}e^{-f},$ and $\Box^{*}=-\partial/\partial t-\triangle+R$ is the conjugate heat operator. Now a routine computation gives

Therefore, if we let $\mu(g_{ij},\tau)=\mbox{inf}\ {\mathcal{W}}(g_{ij},f,\tau)$ over smooth $f$ satisfying (3.2), and $\nu(g_{ij})=\mbox{inf}\ \mu(g_{ij},\tau)$ over all positive $\tau,$ then $\nu(g_{ij}(t))$ is nondecreasing along the Ricci flow. It is not hard to show that in the definition of $\mu$ there always exists a smooth minimizer $f$ (on a closed $M$). It is also clear that $\lim_{\tau\to\infty}\mu(g_{ij},\tau)=+\infty$ whenever the first eigenvalue of $-4\triangle+R$ is positive. Thus, our statement that there is no shrinking breathers other than gradient solitons, is implied by the following

Claim For an arbitrary metric $g_{ij}$ on a closed manifold M, the function $\mu(g_{ij},\tau)$ is negative for small $\tau>0$ and tends to zero as $\tau$ tends to zero.

Proof of the Claim. (sketch) Assume that $\bar{\tau}>0$ is so small that Ricci flow starting from $g_{ij}$ exists on $[0,\bar{\tau}].$ Let $u=(4\pi\tau)^{-\frac{n}{2}}e^{-f}$ be the solution of the conjugate heat equation, starting from a $\delta$-function at $t=\bar{\tau},\tau(t)=\bar{\tau}-t.$ Then ${\mathcal{W}}(g_{ij}(t),f(t),\tau(t))$ tends to zero as $t$ tends to $\bar{\tau},$ and therefore $\mu(g_{ij},\bar{\tau})\leq{\mathcal{W}}(g_{ij}(0),f(0),\tau(0))<0$ by (3.4).

Now let $\tau\to 0$ and assume that $f^{\tau}$ are the minimizers, such that

The metrics $\frac{1}{2}\tau^{-1}g_{ij}$ ”converge” to the euclidean metric, and if we could extract a converging subsequence from $f^{\tau},$ we would get a function $f$ on $\mathbb{R}^{n}$, such that $\int_{\mathbb{R}^{n}}{(2\pi)^{-\frac{n}{2}}e^{-f}dx}=1$ and

The latter inequality contradicts the Gaussian logarithmic Sobolev inequality, due to L.Gross. (To pass to its standard form, take $f=|x|^{2}/2-2\log\phi$ and integrate by parts) This argument is not hard to make rigorous; the details are left to the reader.

Let $g_{ij}(t)$ be a smooth solution to the Ricci flow $(g_{ij})_{t}=-2R_{ij}$ on $[0,T).$ We say that $g_{ij}(t)$ is locally collapsing at $T,$ if there is a sequence of times $t_{k}\to T$ and a sequence of metric balls $B_{k}=B(p_{k},r_{k})$ at times $t_{k},$ such that $r_{k}^{2}/t_{k}$ is bounded, $|Rm|(g_{ij}(t_{k}))\leq r_{k}^{-2}$ in $B_{k}$ and $r_{k}^{-n}Vol(B_{k})\to 0.$

Theorem. If $M$ is closed and $T<\infty,$ then $g_{ij}(t)$ is not locally collapsing at $T.$

Proof. Assume that there is a sequence of collapsing balls $B_{k}=B(p_{k},r_{k})$ at times $t_{k}\to T.$ Then we claim that $\mu(g_{ij}(t_{k}),r_{k}^{2})\to-\infty.$ Indeed one can take $f_{k}(x)=-\log\phi(\mbox{dist}_{t_{k}}(x,p_{k})r_{k}^{-1})+c_{k},$ where $\phi$ is a function of one variable, equal 1 on $[0,1/2],$ decreasing on $[1/2,1],$ and very close to 0 on $[1,\infty),$ and $c_{k}$ is a constant; clearly $c_{k}\to-\infty$ as $r_{k}^{-n}Vol(B_{k})\to 0.$ Therefore, applying the monotonicity formula (3.4), we get $\mu(g_{ij}(0),t_{k}+r_{k}^{2})\to-\infty.$ However this is impossible, since $t_{k}+r_{k}^{2}$ is bounded.