A metric is a symmetric non-degenerate (0,2) tensor field:
- Symmetric: The indices can be interchanged since
- Non-degenerate: The determinant of a metric is non-zero; the inverse exists. So, the eigenvalues of a metric (as an nxn matrix) are non-zero.
The pattern of +ve and -ve eigenvalues is called a metric’s signature.
- (++++) - Riemannian Signature
- (-+++) - Lorentzian Signature
Although the values of the eigenvalues depend on the choice of coordinate system, Sylvester’s Law of Inertia says that the signature is coordinate-independent.
Uses of a metric:
Defines the inner product on Tₚ
Maps vectors to co-vectors
This map has an inverse too, with a (2,0) tensor called the inverse metric defined such that More generally, from tensors of type (r, s) to (r+1, s-1) and (r-1, s+1), called raising and lowering indices.
Defines a notion of length
We can define the line element as and hence the proper length and proper time are
Examples:
- TheMinkowski metric - Although this is not a tensor equation since it is not a tensor under general coordinate transformations.
- Unit Sphere with the “round” metric - Obtained by restricting the flat metric on the 3-d real space to a unit sphere, i.e.,
with 3. Poincare Disc -
This arises as a hyperbolic surface in 3dMinkowski space.
A simple coordinate system on this surface is