On transversally symmetric pseudo-Einstein and Fefferman-Einstein spaces (Q2385019): Difference between revisions

Let \((M,HM,J)\) be a Levi-nondegenerate CR manifold of hypersurface type, with a pseudo-Hermitian structure, i.e., a nowhere vanishing \(1\)-form \(\theta\) with \(\theta| _{HM}=0\). Denote by \(T\) the Reeb vector field, by \(g_\theta\) the Webster metric, by \(\nabla\) the Tanaka-Webster connection on \(M\), by \(\mathrm{Tor}\) its torsion and by \(\mathrm{Ric}\) the Webster-Ricci curvature. Then \(M\) is called transversally symmetric (TS) if the local flow generated by \(T\) acts by CR automorphisms, or equivalently if \(T\) is a Killing vector field for \(g_\theta\). If furthermore \(\mathrm{Ric}\) is a constant multiple of \(d\theta\), then \(M\) is called a TS-pseudo-Einstein-space. If \(M\) is TS, locally the quotient under the action of \(T\) induces a pseudo-Riemannian CR submersion \(M\rightarrow N\) onto a pseudo-Kähler manifold \((N,h)\) and the Webster-Ricci curvature of \(M\) is essentially the same as that of \(N\). It follows that if \(M\) is TS-pseudo-Einstein, then \(N\) is Káhler-Einstein. Vice versa, given a Kähler-Einstein manifold \(N\), the space \(M\) of unit vectors in the anticanonical bundle of \(N\) is in a natural way a TS-pseudo-Einstein manifold, and all TS-pseudo-Einstein manifolds are locally isomorphic to one of this form. Furthermore, the Fefferman space \(F\) built over \(M\) is never Einstein, but if \(M\) is TS-pseudo-Einstein, then \(F\) is locally conformally Einstein.