Non-uniform Hyperbolicity and Non-uniform Specification

TL;DR

This paper extends the concept of specification to non-uniform hyperbolic systems, showing that certain ergodic measures satisfy a weaker form of specification, leading to improved recurrence results and periodic point approximations.

Contribution

It introduces a non-uniform version of the specification property for hyperbolic measures and derives new consequences for recurrence and periodic point approximation.

Findings

Improves recurrence estimates without full specification.

Shows measures are limits of periodic point averages.

Calculates topological pressure via periodic points.

Abstract

In this paper we deal with an invariant ergodic hyperbolic measure $\mu$ for a diffeomorphism $f,$ assuming that $f$ it is either $C^{1+\alpha}$ or $f$ is $C^1$ and the Oseledec splitting of $\mu$ is dominated. We show that this system $(f,\mu)$ satisfies a weaker and non-uniform version of specification, related with notions studied in several recent papers, including \cite{STV,Y, PS, T,Var, Oli}. Our main results have several consequences: as corollaries, we are able to improve the results about quantitative Poincar\'e recurrence, removing the assumption of the non-uniform specification property in the main Theorem of \cite{STV} that establishes an inequality between Lyapunov exponents and local recurrence properties. Another consequence is the fact that any of such measure is the weak limit of averages of Dirac measures at periodic points, as in \cite{Sigmund}. Following \cite{Y} and \cite{PS}, one can show that the topological pressure can be calculated by considering the convenient weighted sums on periodic points, whenever the dynamics is positive expansive and every measure with pressure close to the topological pressure is hyperbolic.