Statistical properties for flows with unbounded roof function, including the Lorenz attractor

Peter Balint,Ian Melbourne

DOI: https://doi.org/10.1007/s10955-018-2093-y

2018-03-06

Abstract:For geometric Lorenz attractors (including the classical Lorenz attractor) we obtain a greatly simplified proof of the central limit theorem which applies also to the more general class of codimension two singular hyperbolic attractors. We also obtain the functional central limit theorem and moment estimates, as well as iterated versions of these results. A consequence is deterministic homogenisation (convergence to a stochastic differential equation) for fast-slow dynamical systems whenever the fast dynamics is singularly hyperbolic of codimension two.

Dynamical Systems

What problem does this paper attempt to address?

The paper primarily focuses on the study of flows with unbounded roof functions, particularly the statistical properties of the Lorenz attractor and its related systems. Specifically, the objectives of the paper can be summarized as follows:

1. **Simplified Proof**: Provide a significantly simplified proof of the Central Limit Theorem (CLT) for geometric Lorenz attractors, including the classic Lorenz attractor.

2. **Extended Results**: Prove that the CLT and related statistical properties, such as moment estimates and exponential decay of correlations, are not only applicable to geometric Lorenz attractors but also to a broader class of codimension-two singular hyperbolic attractors.

3. **Iterative Versions**: In addition to the basic results, the paper also explores iterative versions of these statistical properties, such as the Iterated Weak Invariance Principle (IWIP) and moment estimates.

4. **Deterministic Homogenization**: Building on the above results, the paper further discusses the convergence of fast-slow dynamical systems to stochastic differential equations under specific conditions.

5. **Improved Technical Tools**: To achieve these results, the paper introduces a rearrangement technique to optimize moment estimates and discusses how this technique can improve existing estimates.

In summary, the core goal of this paper is to deepen the understanding of the statistical behavior of dynamical systems with unbounded roof functions and to expand the scope and depth of research into the statistical properties of these systems through theoretical analysis and technical improvements.