De Sitter space

From Wikipedia, the free encyclopedia
Jump to: navigation, search

In mathematics and physics, a de Sitter space is the analog in Minkowski space, or spacetime, of a sphere in ordinary, Euclidean space. The n-dimensional de Sitter space, denoted dSn, is the Lorentzian manifold analog of an n-sphere (with its canonical Riemannian metric); it is maximally symmetric, has constant positive curvature, and is simply connected for n at least 3. The de Sitter space, as well as the anti-de Sitter space is named after Willem de Sitter (1872–1934), professor of astronomy at Leiden University and director of the Leiden Observatory. Willem de Sitter and Albert Einstein worked in the 1920s in Leiden closely together on the spacetime structure of our universe.

In the language of general relativity, de Sitter space is the maximally symmetric, vacuum solution of Einstein's field equations with a positive (repulsive) cosmological constant \Lambda (corresponding to a positive vacuum energy density and negative pressure). When n = 4 (3 space dimensions plus time), it is a cosmological model for the physical universe; see de Sitter universe.

De Sitter space was discovered by Willem de Sitter, and, at the same time, independently by Tullio Levi-Civita.

More recently it has been considered as the setting for special relativity rather than using Minkowski space, since a group contraction reduces the isometry group of de Sitter space to the Poincaré group, allowing a unification of the spacetime translation subgroup and Lorentz transformation subgroup of the Poincaré group into a simple group rather than a semi-simple group. This alternate formulation of special relativity is called de Sitter relativity.


De Sitter space can be defined as a submanifold of a Minkowski space of one higher dimension. Take Minkowski space R1,n with the standard metric:

ds^2 = -dx_0^2 + \sum_{i=1}^n dx_i^2.

De Sitter space is the submanifold described by the hyperboloid of one sheet

-x_0^2 + \sum_{i=1}^n x_i^2 = \alpha^2

where \alpha is some positive constant with dimensions of length. The metric on de Sitter space is the metric induced from the ambient Minkowski metric. The induced metric is nondegenerate and has Lorentzian signature. (Note that if one replaces \alpha^2 with -\alpha^2 in the above definition, one obtains a hyperboloid of two sheets. The induced metric in this case is positive-definite, and each sheet is a copy of hyperbolic n-space.)

De Sitter space can also be defined as the quotient O(1,n)/O(1,n−1) of two indefinite orthogonal groups, which shows that it is a non-Riemannian symmetric space.

Topologically, de Sitter space is R × Sn−1 (so that if n ≥ 3 then de Sitter space is simply connected).


The isometry group of de Sitter space is the Lorentz group O(1,n). The metric therefore then has n(n+1)/2 independent Killing vectors and is maximally symmetric. Every maximally symmetric space has constant curvature. The Riemann curvature tensor of de Sitter is given by

R_{\rho\sigma\mu\nu} = {1\over \alpha^2}(g_{\rho\mu}g_{\sigma\nu} - g_{\rho\nu}g_{\sigma\mu})

De Sitter space is an Einstein manifold since the Ricci tensor is proportional to the metric:

R_{\mu\nu} = \frac{n-1}{\alpha^2}g_{\mu\nu}

This means de Sitter space is a vacuum solution of Einstein's equation with cosmological constant given by

\Lambda = \frac{(n-1)(n-2)}{2\alpha^2}.

The scalar curvature of de Sitter space is given by

R = \frac{n(n-1)}{\alpha^2} = \frac{2n}{n-2}\Lambda.

For the case n = 4, we have Λ = 3/α2 and R = 4Λ = 12/α2.

Static coordinates[edit]

We can introduce static coordinates (t, r, \ldots) for de Sitter as follows:

x_0 = \sqrt{\alpha^2-r^2}\sinh(t/\alpha)
x_1 = \sqrt{\alpha^2-r^2}\cosh(t/\alpha)
x_i = r z_i \qquad\qquad\qquad\qquad\qquad 2\le i\le n.

where z_i gives the standard embedding the (n−2)-sphere in Rn−1. In these coordinates the de Sitter metric takes the form:

ds^2 = -\left(1-\frac{r^2}{\alpha^2}\right)dt^2 + \left(1-\frac{r^2}{\alpha^2}\right)^{-1}dr^2 + r^2 d\Omega_{n-2}^2.

Note that there is a cosmological horizon at r = \alpha.

Flat slicing[edit]


x_0 = \alpha \sinh(t/\alpha) + r^2 e^{t/\alpha}/2\alpha,
x_1 = \alpha \cosh(t/\alpha) - r^2 e^{t/\alpha}/2\alpha,
x_i = e^{t/\alpha}y_i, \qquad 2 \leq i \leq n

where r^2=\sum_i y_i^2. Then in the (t,y_i) coordinates metric reads:

ds^{2} = -dt^{2} + e^{2t/\alpha} dy^{2}

where dy^2=\sum_i dy_i^2 is the flat metric on y_i's.

Open slicing[edit]


x_0 = \alpha \sinh(t/\alpha) \cosh\xi,
x_1 = \alpha \cosh(t/\alpha),
x_i = \alpha z_i \sinh(t/\alpha) \sinh\xi, \qquad 2 \leq i \leq n

where \sum_i z_i^2 = 1 forming a S^{n-2} with the standard metric \sum_i dz_i^2 = d\Omega_{n-2}^2. Then the metric of the de Sitter space reads

ds^2 = -dt^2 + \alpha^2 \sinh^2(t/\alpha) dH_{n-1}^2,


dH_{n-1}^2 = d\xi^2 + \sinh^2\xi d\Omega_{n-2}^2

is the metric of a Euclidean hyperbolic space.

Closed slicing[edit]


x_0 = \alpha \sinh(t/\alpha),
x_i = \alpha \cosh(t/\alpha) z_i, \qquad 1 \leq i \leq n

where z_is describe a S^{n-1}. Then the metric reads:

ds^2 = -dt^2 + \alpha^2 \cosh^2(t/\alpha) d\Omega_{n-1}^2.

Changing the time variable to the conformal time via \tan(\eta/2)=\tanh(t/2\alpha) we obtain a metric conformally equivalent to Einstein static universe:

ds^2 = \frac{\alpha^2}{\cos^2\eta}(-d\eta^2 + d\Omega_{n-1}^2).

This serves to find the Penrose diagram of de Sitter space.[clarification needed]

dS slicing[edit]


x_0 = \alpha \sin(\chi/\alpha) \sinh(t/\alpha) \cosh\xi,
x_1 = \alpha \cos(\chi/\alpha),
x_2 = \alpha \sin(\chi/\alpha) \cosh(t/\alpha),
x_i = \alpha z_i \sin(\chi/\alpha) \sinh(t/\alpha) \sinh\xi, \qquad 3 \leq i \leq n

where z_is describe a S^{n-3}. Then the metric reads:

ds^2 = d\chi^2 + \sin^2(\chi/\alpha) ds_{dS,\alpha,n-1}^2,


ds_{dS,\alpha,n-1}^2 = -dt^2 + \alpha^2 \sinh^2(t/\alpha) dH_{n-2}^2

is the metric of an n-1 dimensional de Sitter space with radius of curvature \alpha in open slicing coordinates. The hyperbolic metric is given by:

dH_{n-2}^2 = d\xi^2 + \sinh^2\xi d\Omega_{n-3}^2.

This is the analytic continuation of the open slicing coordinates under (t,\xi,\theta,\phi_1,\phi_2,\cdots,\phi_{n-3}) \to (i\chi,\xi,it,\theta,\phi_1,\cdots,\phi_{n-4}) and also switching x_0 and x_2 because they change their timelike/spacelike nature.

See also[edit]