# nLab geodesic

### Context

#### Riemannian geometry

Riemannian geometry

## Applications

#### Differential geometry

differential geometry

synthetic differential geometry

# Contents

## Idea

On a Riemannian manifold $(X,g)$, a geodesic (or geodesic line, geodesic path) is a path $x : I \to X$, for some (possibly infinite) interval $I$, which locally minimizes distance.

## Definition

One way to define a geodesic is as a critical point of the functional of length

$(x : I \to M) \mapsto \int_I \sqrt{g(\stackrel{\cdot}{x}(t),\stackrel{\cdot}{x}(t))} dt$

on the appropriate space of curves $I \to X$, where $g$ is the metric tensor. This implies that it satisfies the corresponding Euler-Lagrange equations, which in this case means that the covariant derivative (for the Levi-Civita connection)

$\nabla_{\stackrel{\cdot}{x}} \stackrel{\cdot}{x} = 0$

vanishes.

In local coordinates, with Christoffel symbols $\Gamma^i_{jk}$ the Euler-Lagrange equations for geodesics form a system

$\frac{d^2 x^i}{dt^2} + \sum_{jk} \Gamma^i_{jk}(x) \frac{d x^j}{dt} \frac{d x^k}{dt} = 0.$

So this means that a curve is a geodesic if at every point its tangent vector is the parallel transport of the tangent vector at the start point along the curve.

## Minimizing geodesics

A geodesic may not globally minimize the distance between its end points. For instance, on a 2-dimensional sphere, geodesics are arcs? of great circle?s. Any two distinct non-antipodal points are connected by exactly two such geodesics, one shorter than the other (you can go from Los Angeles to Boston directly across North America, or the long way around the world).

A geodesic which does globally minimize distance between its end points is called a minimizing geodesic. The length of a minimizing geodesic between two points defines a distance function for any Riemannian manifold which makes it into a metric space.

Revised on December 10, 2014 16:39:22 by Urs Schreiber (163.1.81.28)