Table of Contents
Fetching ...

Mean ergodic theorems in $L^r(μ)$ and $H^r(\mathbb T)$, $0<r<1$

el Houcein el Abdalaoui, Michael Lin

TL;DR

This work analyzes mean ergodic behavior of the Koopman operator T=f∘θ on L^r(μ) for 0<r<1 and on Hardy spaces H^r(𝕋). It proves that if M_n f converges in L^r, the limit is T-invariant, hence constant under ergodicity, and it shows that M_n f→0 iff f lies in the closure of (I−T)L^r, though the converse fails in general; there exist functions in this closure for which M_n f does not converge a.e. Additionally, for p<1/r there exist dense G_δ subsets of L^p with limsup behavior of T^n h scaled by n^r diverging a.e. In the Hardy setting with irrational circle rotations, H^r(𝕋) admits a decomposition into constants and the closure of (I−T)H^r(𝕋), and M_N f converges in L^r to a(f) for each f∈H^r, while 1∉cl(I−T)H^r; nonetheless, there exist f∈(I−T)H^r with no a.e. convergence. An appendix provides delicate rate-of-convergence constructions showing sharp, nonuniform rates and limsup phenomena for T^n h.

Abstract

Let $T$ be the Koopman operator of a measure preserving transformation $θ$ of a probability space $(X,Σ,μ)$. We study the convergence properties of the averages $M_nf:=\frac1n\sum_{k=0}^{n-1}T^kf$ when $f \in L^r(μ)$, $0<r<1$. We prove that if $\int |M_nf|^r dμ\to 0$, then $f \in \overline{(I-T)L^r}$, and show that the converse fails whenever $θ$ is ergodic aperiodic. When $θ$ is invertible ergodic aperiodic, we show that for $0<r<1$ there exists $f_r \in (I-T)L^r$ for which $M_nf_r$ does not converge a.e. (although $\int |M_nf|^r dμ\to 0$). We further establish that for $1 \leq p <\frac{1}{r},$ there is a dense $G_δ$ subset ${\mathcal F}\subset L^p(X,μ)$ such that $\limsup_n \frac{|T^nh|}{n^r}=\infty$ a.e. for any $h \in {\mathcal F}$.

Mean ergodic theorems in $L^r(μ)$ and $H^r(\mathbb T)$, $0<r<1$

TL;DR

This work analyzes mean ergodic behavior of the Koopman operator T=f∘θ on L^r(μ) for 0<r<1 and on Hardy spaces H^r(𝕋). It proves that if M_n f converges in L^r, the limit is T-invariant, hence constant under ergodicity, and it shows that M_n f→0 iff f lies in the closure of (I−T)L^r, though the converse fails in general; there exist functions in this closure for which M_n f does not converge a.e. Additionally, for p<1/r there exist dense G_δ subsets of L^p with limsup behavior of T^n h scaled by n^r diverging a.e. In the Hardy setting with irrational circle rotations, H^r(𝕋) admits a decomposition into constants and the closure of (I−T)H^r(𝕋), and M_N f converges in L^r to a(f) for each f∈H^r, while 1∉cl(I−T)H^r; nonetheless, there exist f∈(I−T)H^r with no a.e. convergence. An appendix provides delicate rate-of-convergence constructions showing sharp, nonuniform rates and limsup phenomena for T^n h.

Abstract

Let be the Koopman operator of a measure preserving transformation of a probability space . We study the convergence properties of the averages when , . We prove that if , then , and show that the converse fails whenever is ergodic aperiodic. When is invertible ergodic aperiodic, we show that for there exists for which does not converge a.e. (although ). We further establish that for there is a dense subset such that a.e. for any .
Paper Structure (4 sections, 24 theorems, 103 equations)

This paper contains 4 sections, 24 theorems, 103 equations.

Key Result

Lemma 2.1

Let $T$ be induced on $L^r(\mu)$ by a measure preserving transformation $\theta$, and let $f \in L^r$. If $\frac{1}{n}\sum_{k=0}^{n-1}T^k f\to g$, then $Tg=g$. Hence $g$ is constant when $\theta$ is ergodic.

Theorems & Definitions (43)

  • Lemma 2.1
  • proof
  • Corollary 2.2
  • Theorem 2.3
  • proof
  • Theorem 2.4
  • proof
  • Proposition 2.5
  • proof
  • Proposition 2.6
  • ...and 33 more