I'm currently trying to understand the sectional curvature of riemannian manifolds and I don't know if I'm thinking correctly.

So, say we have a riemannian manifold $(M,g)$ with constant sectional curvature $k$. As far as i know the curvature at any given point is completely determined by the metric $g$. After doing some calculations it seems true to me, that if we equip $M$ with a new metric $$ \tilde{g} = \lambda \cdot g, $$ where $\lambda$ is just some strictly positive scalar, the curvature of the manifold $(M,\tilde{g})$ is scaled in the same way, i.e. $(M,\tilde{g})$ has constant sectional curvature $\lambda \cdot k$.

Now, I know that the sectional curvature of the sphere with radius $1$ is equal to $1$ and the curvature of the sphere with radius $2$ is equal to $\frac{1}{4}$. I'm wondering if, from a riemannian geometry viewpoint, it's somehow the same to look at the radius $2$ sphere, embedded in euclidean space OR the radius $1$ sphere, equipped with the "compressed" metric $\frac{1}{4}g$, where $g$ is the usual scalar product of $\mathbb{R}^n$.

This is probably a rather soft question, but maybe someone has something to add or can correct my reasoning :)


You probably made a mistake when you computed the curvature of the scaled manifold. In local coordinates $\mathrm{Riem} \approx \partial \Gamma + \Gamma^2$ and $\Gamma \approx g^{-1} \partial g$. So the Riemann curvature as a $(1,3)$ tensor is scale invariant. This means that the sectional curvature

$$ K(u,v) = \frac{\langle \mathrm{Riem}(u,v)v,u\rangle}{\langle u,u\rangle\langle v,v\rangle - \langle u,v\rangle^2} $$

scales like $g * g^{-2}$ or that $\tilde{K} = \lambda^{-1} K$ if $\tilde{g} = \lambda g$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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