Sharp estimates of the Kobayashi metric and Gromov hyperbolicity (Q932356): Difference between revisions

Let \((M,J)\) be an almost complex manifold of real dimension 4. Let \(D\subset \subset M\) denote a strictly \(J\)-pseudoconvex domain with a smooth boundary, given by \(D=\{ \rho <0\}\), where \(\rho \) is a smooth strictly \(J\)-plurisubharmonic function in a neighborhood of \(\overline D\). The author gives effective upper and lower bounds for the Kobayashi pseudodifferential metric \(K_{D,J}\) on \(D\). Namely, let \[ M(p; \nu):= \frac{|\nu_n(p)|^2}{4 \delta_D(p)^2}+ \frac{{\mathcal L}_J \rho (\pi (p);\nu_t)}{2 \delta_D(p)} , \] where \(\delta_D (p)\) denotes the distance of a point \(p \in M\) to the boundary of \(D\) (measured with respect to some Riemannnian metric), and \(\nu\) a tangent vector to \(D\) at \(p\) of type \((1,0)\) (with respect to \(J\)), which is split into its tangential and normal components \(\nu_t\) and \(\nu_n\), respectively. This splitting is taken at the projection \(\pi (p)\) of \(p\) to \(\partial D\), which is defined as follows: If \(p\) is sufficiently close to \(\partial D\), there exists a uniquely determined point \(\pi (p) \in \partial D\) such that \(\text{dist}(p,\pi(p)) = \delta_D(p)\). By \({\mathcal L}_J \rho (x;\nu)\) we denote the \(J\)-Levi form of the function \(\rho\) at the point \(x\). With these notations the result is: Theorem A: Let \((M,J)\) and \(D\) be as before. Then, for any \(\varepsilon >0\) there exists \(0< \varepsilon_0 < \varepsilon\) and positive constants \(C\) and \(s\) such that for every \(p \in N_{\varepsilon_0}(\partial D):=\{q\in M\mid \delta_D(q)< \varepsilon_0\}\) one has \[ e^{-C\delta_D(p)^s} M(p;\nu)\leq K_{D,J}(p,\nu)\leq e^{C\delta_D(p)^s} M(p;\nu) . \] The paper concludes with two applications to Gromov hyperbolicity: Theorem B: (1) Let \(D\) be a relatively compact strictly \(J\)-pseudoconvex smooth domain in an almost complex manifold \(M\) of real dimension 4. Then the domain \(D\), endowed with the integrated Kobayashi distance is a Gromov hyperbolic space. (2) Each point in a 4-dimensional almost complex manifold admits a basis of Gromov hyperbolic neighborhoods. These results generalize in the case of real dimension 4 those of \textit{Z. Balogh} and \textit{M. Bonk} [Comment. Math. Helv. 75, No. 3, 504--533 (2000; Zbl 0986.32012)], who had studied the case of strictly pseudoconvex domains in \(\mathbb C^n\).