Topology of event horizons and topological censorship

From MaRDI portal
Publication:4834976



Abstract: We prove that, under certain conditions, the topology of the event horizon of a four dimensional asymptotically flat black hole spacetime must be a 2-sphere. No stationarity assumption is made. However, in order for the theorem to apply, the horizon topology must be unchanging for long enough to admit a certain kind of cross section. We expect this condition is generically satisfied if the topology is unchanging for much longer than the light-crossing time of the black hole. More precisely, let M be a four dimensional asymptotically flat spacetime satisfying the averaged null energy condition, and suppose that the domain of outer communication CK to the future of a cut K of Sm is globally hyperbolic. Suppose further that a Cauchy surface Sigma for CK is a topological 3-manifold with compact boundary partialS in M, and S′ is a compact submanifold of with spherical boundary in S (and possibly other boundary components in M/S). Then we prove that the homology group H1(Sigma′,Z) must be finite. This implies that either partialS′ consists of a disjoint union of 2-spheres, or S′ is nonorientable and partialS′ contains a projective plane. Further, partialS=partialIp[K]cappartialIm[Sp], and partialSigma will be a cross section of the horizon as long as no generator of partialIp[K] becomes a generator of partialIm[Sp]. In this case, if S is orientable, the horizon cross section must consist of a disjoint union of 2-spheres.}





Cited in
(24)








This page was built for publication: Topology of event horizons and topological censorship

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4834976)