Statistical properties of interval maps with critical points and discontinuities

TL;DR

This paper studies the statistical behavior of interval maps with critical points and discontinuities, establishing conditions for the existence of attractors, invariant measures, and mixing rates, providing a comprehensive understanding of their dynamics.

Contribution

It introduces new conditions under which interval maps with critical points and discontinuities exhibit finitely many attractors with absolutely continuous invariant measures and detailed statistical properties.

Findings

Existence of finitely many attractors with full probability basins

Each attractor supports an absolutely continuous invariant measure

Rates of mixing are linked to derivative growth and recurrence patterns

Abstract

We consider dynamical systems given by interval maps with a finite number of turning points (including critical points, discontinuities) possibly of different critical orders from two sides. If such a map $f$ is continuous and piecewise $C^2$, satisfying negative Schwarzian derivative and some summability conditions on the growth of derivatives and recurrence along the turning orbits, then $f$ has finitely many attractors whose union of basins of attraction has total probability, and each attractor supports an absolutely continuous invariant probability measure $\mu$. Over each attractor there exists a renormalization $(f^m,\mu)$ that is exact, and the rates of mixing (decay of correlations) are strongly related to the rates of growth of the derivatives and recurrence along the turning orbits in the attractors. We also give a sufficient condition for $(f^m,\mu)$ to satisfy the Central Limit Theorem. In some sense, we give a fairly complete global picture of the dynamics of such maps. Similarly, we can get similar statistical properties for interval maps with critical points and discontinuities under some more assumptions.