| |||||||||
In differential geometry, Ricci curvature is 2-valent tensor, obtained as a trace of the full curvature tensor.
It can be thought of as a Laplacian of the Riemannian metric tensor in the case of Riemannian manifolds.
It is usually denoted by Ric. Given an orthonormal basis <math>\{e_i\}<math> in the tangent space at p we have
where R denotes the curvature tensor, the result does not depend on the choice of orthonormal basis.
Ricci curvature can be also explained in terms of the sectional curvature in the following way: for a unit vector v, <Ric(v),v> is sum of the sectional curvatures of all the planes spanned by the vector v and a vector from an orthonormal frame containing v (there are n-1 such planes). Here <.,.> is metric scalar product.
One can think of Ricci curvature on a Riemannian manifold, as being an operator on the tangent space. If this operator is just multiplication by a constant, then we have an Einstein manifold. The Ricci curvature is proportional to the metric tensor in this case.
In dimensions 2 and 3 Ricci curvature describes completely the curvature tensor, but in higher dimensions Ricci curvature contains less information. For instance, Einstein manifolds do not have to have constant curvature in dimensions 4 and up.
The Ricci curvature can be used to define Chern classes of a manifold, which are topological invariants (so independent of the choice of metric).
Ricci curvature is also used in Ricci flow, where a metric is deformed in the direction of the Ricci curvature. On surfaces, the flow produces a metric of constant Gauss curvature and the uniformization theorem for surfaces follows. In dimension 3, it was recently used to give a complete classification of compact 3-manifolds.
Ricci curvature plays an important role in general relativity, where it is the key term in the Einstein field equations.
Here is a short list of most basic results on manifolds with positive Ricci curvature.
All results above show that positive Ricci curvature does have some geometric meaning, on the contrary negative curvature is not that restrictive, in particular as it was shown by Joachim Lohkamp, any manifold admit a metric of negative curvature.