Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Consider a Riemannian manifold $(M,g_0)$ which is the interior of a compact manifold $(\overline{M}, \overline{g})$.

I'm interested in a kind of conformal variation of the background metric $g_0$. So for $\tau \in [0,T] \subset \mathbb{R}$ let $\omega_\tau \in C^{\infty}(\overline{M})$ and $g_{\tau} := e^{\omega(\tau)} g_0$

In addition there is a family of operators $\{H_\tau \}$ where $H_\tau: L^2(M, g_\tau) \rightarrow L^2(M, g_\tau)$. My goal is to talk about the $(L²-)$ trace of these operators with respect to the background $L^2$-space $L^2(M,g_0)$.

Is this possible? Are there suitable conditions for $\omega$ to get a statement like \begin{equation} L^2(M,g_0)= L^2(M, g_\tau) \ \forall \tau \in [0,T] \end{equation} so that the operators can be viewed as acting on the same space?

share|cite|improve this question
up vote 1 down vote accepted

If you take $\omega_\tau \in C^{\infty}(\bar{M};\mathbb{R})$, then by compactness we have that there exist a function $\Omega:[0,T]\to\mathbb{R}_+$ such that $-\Omega(\tau) < \omega(\tau,x) < \Omega(\tau)$ for all $(\tau,x)\in [0,T]\times \bar{M}$.

This means that there exists a constant $k$ depending only on the dimension of the manifold $M$ such that in a fixed coordinate charge you have $$ \sqrt{|g_0|}e^{-k\Omega(\tau)} \leq \sqrt{|g_\tau|} \leq \sqrt{|g_0|}e^{k\Omega(\tau)} $$ or, in other words, the volume forms of the two Riemannian metrics are compatible. This immediately implies that the two $L^2$ spaces are compatible.

I suspect, however, you may also be interested in the case where $\omega_\tau\in C^\infty(M,\mathbb{R})$, which would be the case if you were to consider $M$ as an asymptotically hyperbolic manifold with conformal boundary $\partial\bar{M}$. In this case $\omega_\tau$ is allowed (in fact encouraged) to blow-up at the boundary. On any open subset with compact closure $\bar{G}\Subset M$ the above argument still works: $L^2(G,g_0) = L^2(G,g_\tau)$. However degeneracies at the boundary can make big differences. If $g_0$ is induced from $\bar{g}$, then it has finite total volume; so all the constant functions are in $L^2(M,g_0)$. But if $g_\tau$ has infinite volume (as in the asymptotically hyperbolic cases usually considered), then clearly the constant functions cannot be in $L^2(M,g_\tau)$.

share|cite|improve this answer
Hello Willie Wong, you're absolutely right, I'm definitely interested in the asymptotic hyperbolic case. In fact it's always $g_0 = \frac{dx^2+d\theta^2}{x^2}$ for $d\theta^2$ the round metric on $S^1$ and $x$ a boundary defining function. In the paper where the notation comes from there is $\omega_\tau \in C^{\infty}(\overline{M})$ so that one can follow your first argumentation, right? I'm a bit confused because you handle the asymptotic hyperbolic case explicitly different? – Robin Neumann May 24 '12 at 12:40
@Robin: right, if $\omega_\tau$ is bounded above and below, then you are fine. I mentioned the AH case separately because I wasn't sure if $g_0$ is a separate metric or if it is the induced metric from $\bar{g}$. (In the latter case $L^2(M,g_0) = L^2(\bar{M},\bar{g})$ by definition.) As you indicated your $g_0$ already degenerates near $x = 0$, and hence is not the induced metric; you cannot compare $L^2(M,g_\tau)$ with $L^2(\bar{M},\bar{g})$, but I suspect that is not a problem for you. – Willie Wong May 24 '12 at 13:51

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.