In his book Riemannian manifolds, John Lee states the following on pages 8-9:

The most fundamental fact about geodesics is that given any point $p\in M$ and any vector $V$ tangent to $M$ at $p$, there is a unique geodesic starting at $p$ with initial tangent vector $V$.

Here is a brief recipe for computing some curvatures at a point $p\in M$:

  1. Pick a 2-dimensional subspace $\Pi$ of the tangent space to $M$ at $p$.

  2. Look at all the geodesics through $p$ whose initial tangent vectors lie in the selected plane $\Pi$. It turns out that near $p$ these sweep out a certain 2-dimensional submanifold $S_{\Pi}$ of $M$, which inherits a Riemannian metric from $M$.

  3. Compute the Gaussian curvature $S_{\Pi}$ of $p$, which the Theorema Egregium tells us can be computed from its Riemannian metric. This gives a number, denoted $K(\Pi)$, called the sectional curvature of $M$ at $p$ associated with the plane $\Pi$.

Thus the curvature of $M$ at $p$ has to be interpreted as a map \begin{equation} K : \lbrace \text{2-planes in } T_pM\rbrace \rightarrow \Bbb{R} \end{equation}

My question:

Does Lee mean select a $2$-dimensional subspace of an $n$-dimensional tangent space $T_pM$ at $p$? In other words, what is the dimensionality of rhe manifold we are working with here? Is this method designed for curvature in higher dimensions?

  • $\begingroup$ The same explanation as this appears in Ch.4 of do Carmo's Riemannian geometry book which he uses as historical motivation for Riemann curvature tensors. $\endgroup$ – user676464327 Apr 30 '20 at 22:57

The tangent space $T_pM$ has some arbitrary dimensionality $n$, but the sectional curvature is always a function of some planar subspace of the tangent space.

Of course, in the case that the tangent space is itself two-dimensional, then the scalar curvature can be expressed as a single number.

Using exterior algebra, we can say $K: \bigwedge^2 T_pM \mapsto \mathbb R$. This is a natural consequence of how the Riemann tensor can be considered as a linear map $R: \bigwedge^2 T_pM \mapsto \bigwedge^2 T_pM$. Let $B \in \bigwedge^2 T_p M$. Then the sectional curvature is just $K(B) = \langle B, R(B)\rangle/ \langle B, B \rangle$, with the inner product extended to 2-vectors as usual.

  • $\begingroup$ Does this "planar subspace" requirement apply to using the Ricci curvature tensor? $\endgroup$ – Stan Shunpike Jan 28 '15 at 4:20
  • 3
    $\begingroup$ No, the Ricci tensor would be viewed as acting on line-like subspaces--also known as common vectors. - The definition most like that of the Riemann I gave above would just be a linear operator, mapping vectors in $T_pM$ to vectors in $T_p M$. $\endgroup$ – Muphrid Jan 28 '15 at 4:21

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.