On the intersection of dynamical covering sets with fractals (Q2126078): Difference between revisions

Let \((X,\mathcal B, T, \mu, d)\) be a metric measure-preserving system. The well-known Poincaré recurrence theorem shows that for \(\mu\)-a.e.~\(x\in X\) we have \(\liminf_{n\to\infty}d(T^n x, x)=0\). \textit{M. D. Boshernitzan} [Invent. Math. 113, No. 3, 617--631 (1993; Zbl 0839.28008)] proved that if \(\mathcal H^\tau(X)\) is \(\sigma\)-finite for some \(\tau>0\), then \(\liminf_{n\to\infty}n^{\frac{1}{\tau}}d(T^n x, x)<\infty\) for \(\mu\)-a.e.~\(x\in X\). \textit{R. Hill} and \textit{S. L. Velani} [Invent. Math. 119, No. 1, 175--198 (1995; Zbl 0834.28009)] introduced the shrinking targets theory, which concerns for a given \(y\in X\) the size of the following set \[ S(y):=\{x\in X: d(T^n x, y)\le \ell_n~\textrm{infinitely often}\}, \] where \(\{\ell_n\}_{n\ge 1}\) is a sequence of positive real numbers tending to zero. On the other hand, \textit{A.-H. Fan} et al. [Proc. Lond. Math. Soc. (3) 107, No. 5, 1173--1219 (2013; Zbl 1347.37023)] proposed the dynamical covering set \[ E(x):=\{y\in X: d(T^n x, y)\le \ell_n~\textrm{infinitely often}\}, \] and calculated its Hausdorff dimension for the case that \(X\) is the unit interval and \(T: x\mapsto 2x\pmod 1\). In this paper the authors consider the following questions: for a given set \(G\subset X\), when is the intersection \(E(x)\cap G\ne\emptyset\) for \(x\in X\)? Furthermore, if the intersection is nonempty, then how large is it? They prove the following result. {Theorem}. Let \((X, \mathcal B, \mu, T, d)\) be an exponentially mixing metric measure-preserving system, and let \(\mu\) be an Ahlfors \(s\)-regular measure with \(s\in(0,\infty)\). Suppose \(\{\ell_n\}_{n\ge 1}\) is a sequence of positive numbers tending to zero such that \[ \alpha:=\inf\{t>0: \sum_{n=1}^\infty\ell_n^t<\infty\}<s. \] Then for any analytic set \(G\subset X\) we have \[ \left\{ \begin{array} {lll} E(x)\cap G=\emptyset&\textrm{if}& \dim_P G<s-\alpha,\\ E(x)\cap G\ne\emptyset&\textrm{if}&\dim_H G>s-\alpha, \end{array}\right. \] for \(\mu\)-a.e.~\(x\in X\). Furthermore, for \(\mu\)-a.e.~\(x\in X\) we have \[ \left\{ \begin{array} {lll} \dim_H(E(x)\cap G)\le \dim_P G+\alpha-s&\textrm{if}& \dim_P G\ge s-\alpha,\\ \dim_H(E(x)\cap G)=-\infty&\textrm{if}&\dim_P G<s-\alpha,\\ \dim_H(E(x)\cap G)\ge\dim_H G+\alpha-s &\textrm{if}&\dim_H G>s-\alpha. \end{array}\right. \] Some applications of this theorem to continued fraction dynamical systems, \(\beta\)-dynamical systems and middle-third Cantor sets are given.