Tautness and Fatou components in \(\mathbb P^2\) (Q1930284)

\textit{J. E. Fornæss} and \textit{N. Sibony} [Math. Ann. 301, No. 4, 813--820 (1995; Zbl 0829.58005)] proved that if \(f\) is a holomorphic self-map od \(\mathbb P^2\) of degree greater or equal than \(2\) and \(\Omega\) is a recurrent Fatou component of \(f\) then either (1) \(\Omega\) is a basin of attraction of some fixed point of \(f\) inside \(\Omega\) or (2) there exists a Riemann surface \(\Sigma\) (biholomorphic to a disc, or to a punctured disc or to an annulus) such that \(\{f^{\circ n}\}\) converges uniformly on compacta to \(\Sigma\) and \(f|_{\Sigma}\) is conjugated to an irrational rotation, or (3) \(\Omega\) is a Siegel domain. In this last case, the closure in the topology of uniform convergence on compacta \(\Gamma(f)\) of \(\{f^{\circ n}\}\) is isomorphic to \(\mathbb T^k \times F\), where \(k=1,2\) and \(F\) is a finite group. In the paper under review, the authors prove that in cases (1) and (2) \(\Omega\) is completely hyperbolic (before it was known that \(\Omega\) is taut) while in case (3) when \(\Gamma(f)\simeq \mathbb T^2\times F\), they prove that \(\Omega\) is taut. The idea of the proof in case (1) is particularly simple: every Cauchy sequence in the Kobayashi distance in \(\Omega\) can be mapped via \(f^{\circ N}\) for \(N\) sufficiently big into a compact Kobayashi ball centered at the fixed point of \(f\) in \(\Omega\), and being \(f\) proper, the sequence must be convergent. In case (2) a similar argument holds, using the fact that \(\Gamma(f)\) contains a holomorphic retraction from \(\Omega\) to \(\Sigma\). The proof of case (3) relies on a characterization of pseudoconvex domains which admit a \(2\)-torus action due to \textit{D. Barrett, E. Bedford} and \textit{J. Dadok} [Math. Z. 202, No. 1, 65--82 (1989; Zbl 0651.32007)]. The paper ends with a study of tautness for Fatou components of Hénon maps. The authors show that in case \(\Omega\) is a recurrent Fatou component for a (generalized) Hénon map and \(\Omega\) is either a basin of attraction or it has an attracting Riemann surface, then \(\Omega\) is ``taut on bounded families'', namely every bounded family of analytic discs in \(\Omega\) is normal. In case \(\Omega\) is a Siegel domain which admits a \(2\)-torus action, then \(\Omega\) is taut.