English

Boundaries and equivariant maps for ergodic groupoids

Dynamical Systems 2026-03-04 v2 Probability

Abstract

We give a notion of boundary pair (B,B+)(\mathcal{B}_-,\mathcal{B}_+) for measured groupoids which generalizes the one introduced by Bader and Furman \cite{BF14} for locally compact groups. In the case of a semidirect groupoid G=ΓX\mathcal{G}=\Gamma \ltimes X obtained by a probability measure preserving action ΓX\Gamma \curvearrowright X of a locally compact group, we show that a boundary pair is exactly (B×X,B+×X)(B_- \times X, B_+ \times X), where (B,B+)(B_-,B_+) is a boundary pair for Γ\Gamma. For any measured groupoid (G,ν)(\mathcal{G},\nu), we prove that the Poisson boundaries associated to the Markov operators generated by a probability measure equivalent to ν\nu provide other examples of our definition. Following Bader and Furman \cite{BF:Unpub}, we define algebraic representability for an ergodic groupoid (G,ν)(\mathcal{G},\nu). In this way, given any measurable representation ρ:GH\rho:\mathcal{G} \rightarrow H into the κ\kappa-points of an algebraic κ\kappa-group H\mathbf{H}, we obtain ρ\rho-equivariant maps B±H/L±\mathcal{B}_\pm \rightarrow H/L_\pm, where L±=L±(κ)L_\pm=\mathbf{L}_\pm(\kappa) for some κ\kappa-subgroups L±<H\mathbf{L}_\pm<\mathbf{H}. In the particular case when κ=R\kappa=\mathbb{R} and ρ\rho is Zariski dense, we show that L±L_\pm must be minimal parabolic subgroups.

Keywords

Cite

@article{arxiv.2402.15355,
  title  = {Boundaries and equivariant maps for ergodic groupoids},
  author = {Filippo Sarti and Alessio Savini},
  journal= {arXiv preprint arXiv:2402.15355},
  year   = {2026}
}

Comments

39 pages, final version to appear in Glasgow Mathematical Journal

R2 v1 2026-06-28T14:58:23.651Z