Mean and pointwise ergodicity for composition operators on rearrangement-invariant spaces
Thomas Kalmes, Dalimil Peša
TL;DR
The paper addresses mean and pointwise ergodicity of composition operators $T_\phi$ on rearrangement-invariant spaces, introducing a mild (ACR) condition on the space and its associate and a power-measure-boundedness condition on the symbol to guarantee mean ergodicity. A central innovation is the rearrangement-invariant topology $\xi$, which sits between the norm and weak topologies and yields completeness with dual identification $X'=(X,\xi)^*$ under (ACR). The authors prove mean ergodicity of $T_\phi$ on all r.i. Banach spaces with these properties, derive a pointwise ergodic theorem via a maximal-ergodic inequality, and demonstrate these results for several non-reflexive spaces (e.g., Lorentz and Orlicz classes), providing examples and necessity discussions. Collectively, the work extends ergodic theory for composition operators beyond reflexive settings and offers a robust framework for both norm- and almost-everywhere convergence in a broad class of function spaces.
Abstract
We study ergodicity of composition operators on rearrangement-invariant Banach function spaces. More precisely, we give a natural and easy-to-check condition on the symbol of the operator which entails mean ergodicity on a very large class of rearrangement-invariant Banach function spaces. Further, we present some natural additional assumptions that allow us to obtain pointwise ergodicity. The class of spaces covered by our results contains many non-reflexive spaces, such as the Lorentz spaces $L^{p, 1}$ and $L^{p,\infty}$, $p \in (1, \infty)$, Orlicz spaces $L \log^α L$ and $\exp L^α$, $α> 0$, and the spaces $L^1$ and $L^{\infty}$ over measure spaces of finite measure. The main novelty in our approach is the application of a new locally convex topology which we introduce and which lies strictly between the norm topology and the weak topology induced by the associate space. Throughout, we give several examples which illustrate the applicability of our results as well as highlight the necessity and optimality of our assumptions.
