The Topology of General Cosmological Models

TL;DR

This paper investigates the possible shapes and sizes of the universe using advanced differential geometry, deriving conditions that imply finiteness and classifying potential topologies under broad assumptions.

Contribution

It extends cosmological topology analysis beyond standard models by applying differential geometry techniques to general irrotational matter flows, deriving finiteness conditions and classifying possible universe topologies.

Findings

Derived a condition implying a finite universe with a bounded diameter.

Identified specific topologies compatible with a finite universe, including $S^1 imes S^2$ and $T^3$.

Ruled out many complex topologies, especially those with negative curvature.

Abstract

Is the universe finite or infinite, and what shape does it have? These fundamental questions, of which relatively little is known, are typically studied within the context of the standard model of cosmology where the universe is assumed to be homogeneous and isotropic. Here we address the above questions in highly general cosmological models, with the only assumption being that the average flow of matter is irrotational. Using techniques from differential geometry, specifically extensions of the Bonnet-Myers theorem, we derive a condition which implies a finite universe and yields a bound for its diameter. Furthermore, under a weaker condition involving the interplay between curvature and diameter, together with the assumption that the universe is finite (i.e., has closed spatial slices), we provide a concise list of possible topologies. Namely, the spatial sections then would be either the ring topologies $S^1 \times S^2$, $S^1\tilde{\times}S^2$, $S^1\times\mathbb{RP}^2$, $\mathbb{RP}^3 \# \mathbb{RP}^3$, or covered by the sphere $S^3$ or torus $T^3$. In particular, under this condition the basic construction of connected sums would be ruled out (save for one), along with the plethora of topologies associated with negative curvature. These results are obtained from consequences of the geometrization of 3-manifolds, by applying a generalization of the almost splitting theorem together with a curvature formula of Ehlers and Ellis.