Intrinsic CR normal coordinates and the CR Yamabe problem (Q1119783)

This paper is founded on the close analogy between conformal geometry and CR geometry. The Levi form on a CR manifold, like the metric on a conformal manifold, is determined only up to multiplication by a non- vanishing function. In the conformal case, a natural normalization of the metric is provided by choosing a metric of constant scalar curvature. The fact that this can be done is the Yamabe problem, solved in dimension \(\geq 6\) by \textit{T. Aubin} [J. Math. Pures Appl., IX. Ser. 55, 269-296 (1976; Zbl 0336.53033)] and in all dimensions by \textit{R. Schoen} [J. Differ. Geom. 20, 479-495 (1984; Zbl 0576.53028)]. The CR analogue of the Yamabe problem consists of normalizing the Levi form by choosing a contact form \(\theta\) which is critical for the functional \[ \int_{M}R\theta \wedge d\theta^ n/(\int_{M}\theta \wedge d\theta^ n)^{2/p};\quad p=2+2/n \] where R is the pseudo- Hermitian scalar curvature. The authors prove (analogously to Aubin's result) that if \(n\geq 2\) (i.e. dim M\(=2n+1\geq 5)\) and M is compact and strictly pseudoconvex then the CR Yamabe problem can be solved. The proof relies on an earlier paper by the same authors [J. Differ. Geometry 25, 167-197 (1987; Zbl 0661.32026)] which proved that a solution exists if the CR Yamabe functional can be made less than a critical value \(\Lambda\). Here, by choosing special intrinsic CR normal coordinates modelled on the geometry of parabolic dilations in the Heisenberg group, the authors show (by expansion around a point where the contact form becomes ``concentrated'') that this inequality can be achieved and the problem therefore solved.