On pseudoconvex domains in \(\mathbb{P}^n\) (Q1281856): Difference between revisions

Let \(\Omega\) be a domain in \(\mathbb{C}\mathbb{P}^n\) and let \(K_\Omega\) be its Bergman kernel with respect to the Fubiny-Study metric. The authors prove first a localization principle for \(K_\Omega\). This can be stated as follows: assume that \(\Omega\) is pseudoconvex and that its complement has non-void interior. Then, given a point \(x\) in \(\overline\Omega\) and two small neighborhoods \(U\) and \(V\) of \(x\), with \(\overline V\subset U\), one has \(K_\Omega (z,z)\approx K_{\Omega\cap U}(z,z)\) for any \(z\in\Omega\cap V\). A similar theorem holds for the corresponding Bergman metric. The authors also study the notion of hyperconvexity for domains in \(\mathbb{C}\mathbb{P}^n\). They construct an example of a locally hyperconvex Stein domain in \(\mathbb{C}\mathbb{P}^5\) which is not hyperconvex.