Dynamic asymptotic dimension growth for group actions and groupoids
Extracted source text · arXiv 2411.19712
\documentclass[12pt, reqno]{amsart}
\usepackage{amssymb, mathrsfs}
%\usepackage[msc-links, lite]{amsrefs}
\usepackage{latexsym}
\usepackage{amsmath}
\usepackage{amsthm}
\usepackage{amsfonts}
\usepackage{dsfont, upgreek}
\usepackage{mathtools}
\usepackage{epsfig}
%\usepackage[scr]{rsfso}
%\usepackage{cite}
%\usepackage{natbib}
\usepackage{amscd}
\usepackage{graphicx}
\usepackage{mathabx}
\usepackage[colorlinks=true,linkcolor=blue,citecolor=blue]{hyperref}
%\usepackage[margin=0.5in]{geometry}
%\usepackage{array}
\usepackage{tikz-cd}
%\usetikzlibrary{external}
%\tikzexternalize[prefix=figures/]
%\usepackage{appendix}
%\usepackage{enumerate}
\usepackage{enumitem}
\usepackage{url}
\setlist[enumerate]{itemsep=0.5ex}
%\usepackage{stmaryrd}
\usepackage{color}
\usepackage{tikz}
\usetikzlibrary{patterns}
\usepackage[left=1.2in,right=1.2in,top=1.4in,bottom=1.4in]{geometry}
%\usepackage[msc-links, lite]{amsrefs}
\usetikzlibrary{decorations.markings}
\usetikzlibrary{arrows.meta}
\usepackage{extarrows}
\usepackage{adjustbox}
\usepackage{pgfplots}
\usepgfplotslibrary{colormaps}
%\usepgfplotslibrary{polar}
%\DeclareFontFamily{U}{MnSymbolC}{}
%\DeclareSymbolFont{MnSyC}{U}{MnSymbolC}{m}{n}
%\DeclareFontShape{U}{MnSymbolC}{m}{n}{
% <-6> MnSymbolC5
% <6-7> MnSymbolC6
% <7-8> MnSymbolC7
% <8-9> MnSymbolC8
% <9-10> MnSymbolC9
% <10-12> MnSymbolC10
% <12-> MnSymbolC12}{}
%\DeclareMathSymbol{\intprod}{\mathbin}{MnSyC}{'270}
\usepackage{slashed}
%----------------------------
%\usepackage[export]{adjustbox}
%\usepackage[colorlinks=true,linkcolor=blue,citecolor=blue]{hyperref}
%\usepackage[msc-links, lite]{amsrefs}
%\renewcommand{\eprint}[1]{\href{https://arxiv.org/pdf/#1.pdf}{arXiv:math/#1}}
\usepackage[all,cmtip]{xy}
\usepackage{comment}
%\usepackage{cite}
%\usepackage{ulem}
%\bibliographystyle{abbrv}
\theoremstyle{plain}% default
\newtheorem{theorem}{Theorem}[section]
\newtheorem{proposition}[theorem]{Proposition}
\newtheorem{lemma}[theorem]{Lemma}
\newtheorem{corollary}[theorem]{Corollary}
\newtheorem{question}[theorem]{Question}
\newtheorem{conjecture}[theorem]{Conjecture}
\theoremstyle{definition}
\newtheorem{definition}[theorem]{Definition}
\newtheorem{example}[theorem]{Example}
\newtheorem*{claim*}{Claim}
\newtheorem{claim}[theorem]{Claim}
\newtheorem{setup}[theorem]{Geometric Setup}
\newtheorem{preparation}[theorem]{Preparation}
\theoremstyle{remark}
\newtheorem{remark}[theorem]{Remark}
\newtheorem{note}[theorem]{Note}
%\newenvironment{prf}{{\noindent \textbf{Proof:}\ }}{\hfill $\Box$ \\ \smallskip}
\numberwithin{equation}{section}
\newcommand{\Sc}{\mathrm{Sc}}
\newcommand{\cl}{\mathbb{C}\ell}
\newcommand{\Endo}{\mathrm{End}}
\newcommand{\Bigwedge}{\mathord{\adjustbox{raise=.4ex, totalheight=.7\baselineskip}{$\bigwedge$}}}
\newcommand{\extension}{\mathbf D_B}
\newcommand{\fiber}{\mathbb F}
\newcommand{\link}{\mathbb L}
\newcommand{\hotimes}{\mathbin{\widehat{\otimes}}}
\newcommand{\coker}{\mathrm{coker}}
\newcommand{\cK}{\mathcal{K}}
\newcommand{\dist}{\mathrm{dist}}
\newcommand{\R}{\mathbb{R}}
\newcommand{\diam}{\mathrm{diam}}
\newcommand{\Z}{\mathbb{Z}}
\newcommand{\rank}{\mathrm{rank}}
\newcommand{\Ric}{\mathrm{Ric}}
\newcommand{\lf}{\mathit{lf}}
\newcommand{\macro}{\mathrm{mac}}
\newcommand{\supp}{\textnormal{supp }}
\newcommand{\In}{\textnormal{In}}
\newcommand{\Prob}{\textnormal{Prob}}
\newcommand{\ad}{\textnormal{ad}}
\newcommand{\dad}{\textnormal{dad}}
\newcommand{\asdim}{\textnormal{asdim}}
\newcommand{\dasdim}{\textnormal{dasdim}}
\DeclareMathOperator{\spt}{spt}
\DeclareMathOperator{\width}{width}
\DeclareMathOperator{\fillrad}{FillRad}
\DeclareMathOperator{\inj}{Inj}
\DeclareMathOperator{\radsphere}{Rad_{\mathbb{S}^n}}
\DeclareMathOperator{\vol}{vol}
\DeclareMathOperator{\kr}{ker}
\DeclareMathOperator{\im}{Im}
\DeclareMathOperator{\lip}{Lip}
\DeclareMathOperator{\conj}{Conj}
\DeclareMathOperator{\conv}{Conv}
\newcommand\descitem[1]{\item{\bfseries #1}}
\newcommand{\interior}[1]{%
{\kern0pt#1}^{\mathrm{\,o}}%
}
\makeatletter
\let\save@mathaccent\mathaccent
\newcommand*\if@single[3]{%
\setbox0\hbox{${\mathaccent"0362{#1}}^H$}%
\setbox2\hbox{${\mathaccent"0362{\kern0pt#1}}^H$}%
\ifdim\ht0=\ht2 #3\else #2\fi
}
%The bar will be moved to the right by a half of \macc@kerna, which is computed by amsmath:
\newcommand*\rel@kern[1]{\kern#1\dimexpr\macc@kerna}
%If there's a superscript following the bar, then no negative kern may follow the bar;
%an additional {} makes sure that the superscript is high enough in this case:
\newcommand*\wideaccent[2]{\@ifnextchar^{{\wide@accent{#1}{#2}{0}}}{\wide@accent{#1}{#2}{1}}}
%Use a separate algorithm for single symbols:
\newcommand*\wide@accent[3]{\if@single{#2}{\wide@accent@{#1}{#2}{#3}{1}}{\wide@accent@{#1}{#2}{#3}{2}}}
\newcommand*\wide@accent@[4]{%
\begingroup
\def\mathaccent##1##2{%
%Enable nesting of accents:
\let\mathaccent\save@mathaccent
%If there's more than a single symbol, use the first character instead (see below):
\if#42 \let\macc@nucleus\first@char \fi
%Determine the italic correction:
\setbox\z@\hbox{$\macc@style{\macc@nucleus}_{}$}%
\setbox\tw@\hbox{$\macc@style{\macc@nucleus}{}_{}$}%
\dimen@\wd\tw@
\advance\dimen@-\wd\z@
%Now \dimen@ is the italic correction of the symbol.
\divide\dimen@ 3
\@tempdima\wd\tw@
\advance\@tempdima-\scriptspace
%Now \@tempdima is the width of the symbol.
\divide\@tempdima 10
\advance\dimen@-\@tempdima
%Now \dimen@ = (italic correction / 3) - (Breite / 10)
\ifdim\dimen@>\z@ \dimen@0pt\fi
%The bar will be shortened in the case \dimen@<0 !
\rel@kern{0.6}\kern-\dimen@
\if#41
#1{\rel@kern{-0.6}\kern\dimen@\macc@nucleus\rel@kern{0.4}\kern\dimen@}%
\advance\dimen@0.4\dimexpr\macc@kerna
%Place the combined final kern (-\dimen@) if it is >0 or if a superscript follows:
\let\final@kern#3%
\ifdim\dimen@<\z@ \let\final@kern1\fi
\if\final@kern1 \kern-\dimen@\fi
\else
#1{\rel@kern{-0.6}\kern\dimen@#2}%
\fi
}%
\macc@depth\@ne
\let\math@bgroup\@empty \let\math@egroup\macc@set@skewchar
\mathsurround\z@ \frozen@everymath{\mathgroup\macc@group\relax}%
\macc@set@skewchar\relax
\let\mathaccentV\macc@nested@a
%The following initialises \macc@kerna and calls \mathaccent:
\if#41
\macc@nested@a\relax111{#2}%
\else
%If the argument consists of more than one symbol, and if the first token is
%a letter, use that letter for the computations:
\def\gobble@till@marker##1\endmarker{}%
\futurelet\first@char\gobble@till@marker#2\endmarker
\ifcat\noexpand\first@char A\else
\def\first@char{}%
\fi
\macc@nested@a\relax111{\first@char}%
\fi
\endgroup
}
\makeatother
\newcommand\overbar{\wideaccent\overline}
\newcommand\doubleoverline[1]{\overline{\overline{#1}}}
\newcommand\widebarbar{\wideaccent\doubleoverline}
\makeatletter
\newcommand*{\transpose}{%
{\mathpalette\@transpose{}}%
}
\newcommand*{\@transpose}[2]{%
% #1: math style
% #2: unused
\raisebox{\depth}{$\m@th#1\intercal$}%
}
\makeatother
\begin{document}
\title[\tiny Dynamic asymptotic dimension growth for group actions and groupoids]{Dynamic asymptotic dimension growth for group actions and groupoids}
\author{Hang Wang}
\address[Hang Wang]{\normalfont{Research Center for Operator Algebras, School of Mathematical Sciences, East China Normal University, Shanghai 200062, China}}
\email{wanghang@math.ecnu.edu.cn}
\thanks{Hang Wang is supported by NSFC 12271165 and in part by the Science and Technology Commission of Shanghai Municipality (No. 22DZ2229014, 23JC1401900)}
\author{Yanru Wang}
\address[Yanru Wang]{\normalfont{Research Center for Operator Algebras, School of Mathematical Sciences\\ East China Normal University, Shanghai 200062, China}}
\email{52215500014@stu.ecnu.edu.cn}
\thanks{Yanru Wang is supported by Technology Commission of Shanghai Municipality (No.23JC1401900). Yanru Wang is the corresponding author.}
\author{Jianguo Zhang}
\address[Jianguo Zhang]{\normalfont{School of Mathematics and Statistics, Shaanxi Normal University,
Xi'an 710119, China}}
\email{jgzhang@snnu.edu.cn}
\thanks{Jianguo Zhang is supported by NSFC 12271165, 12171156, and 12301154}
\author{Dapeng Zhou}
\address[Dapeng Zhou]{\normalfont{School of Statistics and Information, Shanghai University of International Business and Economics, Shanghai 201620, China}}
\email{giantroczhou@126.com}
\thanks{Dapeng Zhou is supported by NSFC 12271165}
\subjclass[2020]{54F45, 22A22, 37A55, 37B05, 46L05}
\keywords{Dynamic asymptotic dimension growth, Group actions, Groupoids, Amenability}
%\date{\today}
\begin{abstract}
We introduce the notion of dynamic asymptotic dimension growth for actions of discrete groups on compact spaces, and more generally for locally compact \'etale groupoids. Using the work of Bartels, L\"uck, and Reich, we bridge asymptotic dimension growth for countable discrete groups with our notion for their group actions, thereby providing numerous concrete examples. Moreover, we demonstrate that the asymptotic dimension growth for a discrete metric space of bounded geometry is equivalent to the dynamic asymptotic dimension growth for its associated coarse groupoid. Consequently, we deduce that the coarse groupoid with subexponential dynamic asymptotic dimension growth is amenable. More generally, we show that every $\sigma$-compact locally compact Hausdorff \'etale groupoid with compact unit space having dynamic asymptotic dimension growth at most $x^{\alpha}$ $(0<\alpha<1)$ is amenable.
\end{abstract}
\maketitle
\tableofcontents
\section{Introduction}
The notion of asymptotic dimension was primarily proposed by Gromov in \cite{Gro93} as a large-scale analogue of the Lebesgue covering dimension. This idea gained widespread attention following Yu's work, demonstrating that the Novikov conjecture holds for groups with finite asymptotic dimension (see \cite{Yu98}). This pivotal result motivated numerous mathematicians to investigate groups and spaces with finite asymptotic dimension. For instance, Gromov initially established that hyperbolic groups and metric spaces have finite asymptotic dimension (see \cite{Gro93}). Subsequently, Roe provided a more straightforward proof of this finding (see \cite{Roe05}). Building on this, Bell and Dranishnikov showed that the free product, the amalgamated free product, and the HNN extension can preserve finite asymptotic dimension (see \cite{BD01}). Four years later, Osin demonstrated that if a finitely generated group $G$ is hyperbolic relative to a collection of subgroups with finite asymptotic dimension, then $G$ itself has finite asymptotic dimension (see \cite{Osi05}). Further advancements were made by Wright, who proved that the asymptotic dimension of a finite-dimensional CAT(0) cube complex is bounded above by its dimension (see \cite{Wri12}). Additionally, Bestvina, Bromberg, and Fujiwara established that mapping class groups have finite asymptotic dimension (see \cite{BBF15}). Following this, Behrstock, Hagen, and Sisto proved that all hierarchically hyperbolic groups have finite asymptotic dimension, thereby recovering results including mapping class groups, relatively hyperbolic groups, and CAT(0)-cube complexes (see \cite{BHS17}), among many others. For a comprehensive survey of this theory, refer to \cite{BD11}; for recent developments, see \cite{FP21, Tse23, FS24}.
Moreover, finite asymptotic dimension has found applications in the domain of $L^{p}$ operator algebras as well. Analogous to the coarse Baum-Connes conjecture, the third and fourth authors established the $L^{p}$ coarse Baum-Connes conjecture for $p\in [1,\infty)$ and proved that this conjecture holds for proper metric spaces with finite asymptotic dimension (see \cite{ZZ21}). Subsequently, the current authors provided an alternative proof of this result through the framework of $C_{0}$ coarse geometry (see \cite{WWZZ23}). Recently, the authors investigated the persistence approximation property regarding quantitative $K$-theory of filtered $L^{p}$ operator algebras and demonstrated that any $L^p$ Roe algebra for a coarsely uniformly contractible discrete metric space with bounded geometry and finite asymptotic dimension satisfies the persistence approximation property (see \cite{WWZZ24}).
However, plenty of groups are known to have infinite asymptotic dimension, including the wreath product $\mathbb{Z}\wr\mathbb{Z}$ (see \cite{Dra06}), the first Grigorchuk group (see \cite{Smi07}), Thompson’s group (see \cite{DS11}),and the Celebrated Gromov's group in \cite{Gro03} containing an expander. To study the growth of dimension functions of groups with infinite asymptotic dimension, Dranishnikov \cite{Dra06} introduced the notion of asymptotic dimension growth for a metric space. This concept extends the notion of finite asymptotic dimension, since a constant asymptotic dimension function indicates finite asymptotic dimension. Dranishnikov also established that the wreath product of a finitely generated nilpotent group with a finitely generated group of finite asymptotic dimension exhibits polynomial asymptotic dimension growth. Furthermore, he proved that polynomial asymptotic dimension growth implies Yu's property A \cite{Dra06}, a weak amenability-type condition introduced by Yu \cite{Yu00}. In conjunction with Yu's theorem on the coarse Baum-Connes conjecture \cite{Yu00}, it follows that spaces with polynomial asymptotic dimension growth satisfy the coarse Baum-Connes conjecture, and consequently, the Novikov conjecture. Building on Dranishnikov's insights, Ozawa showed that a metric space with subexponential asymptotic dimension growth has property A \cite{Oza12}; For additional reference, see \cite{Opp14} regarding discrete metric spaces of bounded geometry. In this context, subexponential asymptotic dimension growth is considered weaker than finite asymptotic dimension but stronger than property A.
Many examples of groups and spaces exhibit subexponential asymptotic dimension growth. A fascinating example is a geodesic coarse median space with finite rank and at most exponential volume growth that has subexponential asymptotic dimension growth \cite{ANWZ18}. This space includes the hyperbolic space and the mapping class group. Another intriguing example is the direct product $G=\bigoplus\limits_{n\in\mathbb{N}}f(n)\mathbb{Z}$, where $f:\mathbb{R}^{+}\rightarrow\mathbb{N}$ is a strictly increasing function that grows slower than a linear function. When endowed with the direct sum metric, $G$ becomes a discrete metric space of bounded geometry with slow asymptotic dimension growth \cite{CFY08}, a special case of subexponential asymptotic dimension growth. Additionally, Bell examined how various constructions involving groups affect the asymptotic dimension function \cite{Bel05}. In this paper, we will introduce three
equivalent interlocking definitions of asymptotic dimension growth. The first equivalent definition is derived from [\cite{ANWZ18}, Lemma 2.6]; the third equivalent definition uses the language of coarse geometry and will be applied to the proof of Theorem \ref{th 1.1}.
Inspired by Gromov’s theory of asymptotic dimension, Guentner, Willett, and Yu introduced the dynamic asymptotic dimension in their seminal paper \cite{GWY17}, which is a concept of dimension for actions of discrete groups on locally compact spaces, as well as locally compact Hausdorff \'etale groupoids. Since its inception, this notion has various applications. It provides upper bounds on the nuclear dimension of the resulting groupoid $C^{*}$-algebras, as demonstrated in \cite{GWY17, DS18, ALSS21, HW24}, and even extends to twisted groupoid $C^{*}$-algebras \cite{CDGHV24, BL24}. Additionally, it establishes a sufficient condition for Matui's HK conjecture in low dimensions \cite{BDGW23}. Furthermore, the dynamic asymptotic dimension is intimately connected to other key concepts in the field. It correlates closely with the diagonal dimension of sub-$C^{*}$-algebras \cite{LLW23}, Kerr's tower dimension \cite{Ker20}, and the conditions formulated by Bartels, Lück, and Reich \cite{BLR08, Bar16}. In a recent study \cite{Bon24}, B\"onicke enriched the understanding of this dimension by establishing several permanence properties including estimates for products and unions of groupoids, as well as comparing the asymptotic dimension of the resulting coarse space with the dynamic asymptotic dimension of the underlying groupoid.
However, group actions and groupoids with infinite dynamic asymptotic dimension have not been extensively explored. The aim of this paper is to contribute to their study by examining the growth of the dynamic asymptotic dimension function for these group actions and groupoids. This study is inspired by the work of Guentner, Willett, and Yu in \cite{GWY17} regarding finite dynamic asymptotic dimension for group actions and \'etale groupoids, along with their subsequent research in \cite{GWY24} verifying the Baum-Connes conjecture for actions with finite dynamical complexity. It should be emphasized that when the dynamic asymptotic dimension is either zero or one, the finite dynamic asymptotic dimension implies finite dynamical complexity. The work of Arzhantseva, Niblo, Wright, and Zhang in \cite{ANWZ18} on characterizing asymptotic dimension growth forms a solid foundation for our study. These influential works provide valuable methods and results that facilitate our investigation.
In the case of group actions, we will introduce the concept of dynamic asymptotic dimension growth for actions by discrete groups on compact Hausdorff spaces in Definition \ref{def 3.2}. This notion can be viewed as the dynamical counterpart of asymptotic dimension growth in an appropriate sense, and it serves as a generalization of finite dynamic asymptotic dimension. The dynamic asymptotic dimension has been estimated for various group actions \cite{ALSS21, CJMST23, GWY17}. Building upon these techniques and findings, we will provide examples of group actions exhibiting dynamic asymptotic dimension growth. For this, we link our notion to the works of Bartels, L\"uck, and Reich on the Farrel-Jones conjecture \cite{BLR08, Bar16} by introducing the $h$-BLR action, where $h: \mathbb{R}^{+}\rightarrow\mathbb{N}$ is a nondecreasing function. Furthermore, when $h$ grows at a rate slower than $\sqrt{x}$, we derive an alternative equivalent definition of the $h$-BLR action, which is formally similar to the definition of asymptotic dimension growth. Additionally, we will demonstrate that an $h$-BLR action has dynamic asymptotic dimension function at most $h$. By applying this result within the domain of coarse geometry, we will bridge the asymptotic dimension growth for countable discrete groups with the dynamic asymptotic dimension growth for their actions, as elaborated in Section \ref{section: dad}. This allows us to present several representative examples of group actions with subexponential dynamic asymptotic dimension growth.
%In this paper, we will define the dynamic asymptotic dimension growth for actions of discrete groups on compact spaces. This concept can be viewed as a generalization of the finite dynamic asymptotic dimension, and serves as the dynamical version of the asymptotic dimension growth in an appropriate sense. Analogous to the finite case, we can connect our notion to the work of Bartels et al. \cite{BLR08, Bar16} on the Farrel-Jones conjecture to introduce a new notion called an $h$-BLR action, where $h: \mathbb{R}^{+}\rightarrow\mathbb{N}$ is a nondecreasing function. We will show that an $h$-BLR action has dynamic asymptotic dimension function at most $h$. Furthermore, when $h$ grows slower than $\sqrt{x}$, we derive an equivariant condition related to the above BLR condition, which is formally similar to the definition of asymptotic dimension growth.
In the case of \'etale groupoids, we will define the concept of dynamic asymptotic dimension growth for locally compact Hausdorff \'etale groupoids, thereby extending our earlier notion for actions of discrete groups on compact spaces. Initially, we define the dynamic asymptotic dimension growth of an \'etale groupoid with compact unit space in Definition \ref{def 4.6}. For a groupoid with non-compact unit space, our definition reduces to the compact case via the Alexandrov groupoid, as outlined in[\cite{KKLRU21}, Definition 7.7]. Within the framework of coarse geometry, we establish a connection between our concept and the asymptotic dimension growth, leading to the following theorem:
\begin{theorem}[see Theorem \ref{th 4.16}]\label{th 1.1}
Let $X$ be a discrete metric space of bounded geometry, and let $G(X)$ be the associated coarse groupoid. Let $f: \mathbb{R}^{+}\rightarrow\mathbb{N}$ be a non-decreasing function. Then $\ad_{X}\approx f$ if and only if $\dad_{G(X)}\approx f$.
\end{theorem}
This theorem is our principal motivation, as it is related to the coarse Baum-Connes conjecture. It establishes that the asymptotic dimension growth of $X$ is equivalent to the dynamic asymptotic dimension growth of its associated coarse groupoid $G(X)$, which generalizes [\cite{GWY17}, Theorem 6.4]. The pivotal step in our proof involves integrating the third equivalent definition of asymptotic dimension growth with the proof technique presented in [\cite{GWY17}, Theorem 6.4].
Combining this theorem with Ozawa's result that subexponential asymptotic dimension growth implies property A \cite{Oza12}, we derive the following corollary:
\begin{corollary}[see Corollary \ref{cor 4.19}]\label{cor 1.2}
Let $X$ be a discrete metric space of bounded geometry. If the associated coarse groupoid $G(X)$ has subexponential dynamic asymptotic dimension growth, then $G(X)$ is amenable.
\end{corollary}
This corollary offers an alternative sufficient condition for the amenability of the coarse groupoid. To illustrate this result, we can present a concrete example: for a geodesic uniformly locally finite coarse median space $X$ with finite rank and at most exponential volume growth, its corresponding coarse groupoid $G(X)$ has subexponential dynamic asymptotic dimension growth, thereby confirming its amenability.
Combining Theorem \ref{th 1.1}, Ozawa's result [\cite{Oza12}, Theorem 1] and the fact that an $h$-BLR action has dynamic asymptotic dimension function at most $h$, we can establish the equivalence between the asymptotic dimension growth for countable discrete groups and the dynamic asymptotic dimension growth for group actions, as elaborated in Theorem \ref{th 4.21}. Consequently, we arrive at the following theorem:
\begin{theorem}[see Theorem \ref{th 4.22}]
Let $\Gamma$ be a countable infinite discrete group with subexponential asymptotic dimension growth $h$. Then $\Gamma$ admits a free, minimal action on the Cantor set which has dynamic asymptotic dimension growth at most $h$.
\end{theorem}
This theorem is an adaption and a generalization of [\cite{GWY17}, Theorem 6.6]. Moreover, it provides a shortcut to construct more concrete examples of group actions with subexponential dynamic asymptotic dimension growth. For example, the wreath product $\mathbb{Z}\wr\mathbb{Z}$ has infinite asymptotic dimension and polynomial asymptotic dimension growth. According to this theorem, $\mathbb{Z}\wr\mathbb{Z}$ admits a free, minimal action on the Cantor set, and the dynamic asymptotic dimension of this action is infinite and grows polynomially.
The amenable groupoids has several major results in various aspects, such as the Baum-Connes conjecture \cite{Tu99}, the Novikov conjecture \cite{Hig00}, and the classification of $C^{*}$-algebras \cite{AR00}. The main implication we explore in this paper is the amenability of groupoids. To this end, we develop a key tool known as generalized partitions of unity. Using this tool, we can directly establish the following main theorem:
\begin{theorem}(see Theorem \ref{th 6.4})
Let $\mathcal{G}$ be a $\sigma$-compact locally compact Hausdorff \'etale groupoid with compact unit space, and with dynamic asymptotic dimension growth $f\preceq x^{\alpha} (0<\alpha<1)$. Then $\mathcal{G}$ is amenable.
\end{theorem}
This theorem is a generalization of Corollary \ref{cor 1.2} in the case of slow growth. Our proof is more closely modeled on the beautiful approach by Guentner, Willett, and Yu's work [\cite{GWY24}, Appendix A ] on finite dynamical complexity for \'etale groupoids. Using a similar approach, we can prove that every locally compact Hausdorff \'etale groupoid with finite dynamic asymptotic dimension is amenable, which strengthens the result of \cite{GWY17} by removing the freeness assumption for groupoids.
The outline of this paper is as follows. In Section 2, we present three equivalent definitions of the asymptotic dimension growth and illustrate them with four typical examples. In Section 3, we introduce the concept of dynamic asymptotic dimension growth for group actions, discussing its connection to the work of Bartels, L\"uck, and Reich. In Section 4, we define the dynamic asymptotic dimension growth for \'etale groupoids. Furthermore, we demonstrate that the asymptotic dimension growth for a discrete metric space of bounded geometry is equivalent to the dynamic asymptotic dimension growth for its associated coarse groupoid. This equivalence allows us to establish a connection between the asymptotic dimension growth for countable discrete groups and the dynamic asymptotic dimension growth for their group actions. In Section 5, we develop a technical tool known as a generalized partition of unity, which is crucial for proving the amenability of groupoids. Finally, in section 6, we identify a sufficient condition for amenable groupoids: every $\sigma$-compact, locally compact Hausdorff \'etale groupoid $\mathcal{G}$ with compact unit space having dynamic asymptotic dimension growth at most $x^{\alpha}$ $(0<\alpha<1)$ is amenable.
\section{Asymptotic dimension growth for metric spaces}\label{section 2}
% Additionally, he gave examples of finitely generated groups with polynomial asymptotic dimension growth and showed that all such groups have property A \cite{Dra06}. Later, Ozawa extended this result to the metric spaces with subexponential asymptotic dimension growth \cite{Oza12}. Combined with Yu's result that the coarse Baum-Connes conjecture holds for spaces with property A, we see that this conjecture holds for spaces with subexponential asymptotic dimension growth. In this sense, subexponential asymptotic dimension growth is weaker than finite asymptotic dimension and stronger than property A. This section will provide a couple of equivalent definitions of the asymptotic dimension growth. Moreover, we will apply the coarse geometric version of the equivalent definition to the proof of Theorem \ref{th 4.16}.
The volume growth of finitely generated groups can be at most exponential. To explore the intermediate dimension growth of these groups, Dranishnikov proposed the concept of asymptotic dimension growth in \cite{Dra06}, which generalizes the notion of finite asymptotic dimension. In this section, we will introduce three equivalent definitions of asymptotic dimension growth. The third equivalent definition employs the terminology of coarse geometry, which can be used in the proof of Theorem \ref{th 4.16} in Section \ref{section: dad}. To facilitate a more intuitive understanding of this concept, we provide examples of spaces exhibiting exponential growth, subexponential growth, polynomial growth, and slow growth.
\subsection{Equivalent definitions}
In this subsection, we will review the concept of asymptotic dimension growth and present two equivalent definitions. To enhance clarity, we will illustrate this concept with several concrete examples.
To clarify the definition of asymptotic dimension function, we will outline some elementary concepts. Let $(X,d)$ be a metric space and $R>0$. A family $\mathcal{W}$ of subsets of $X$ is termed $R$-disjoint if the distance $d(W, W')>R$ for every pair of distinct subsets $W\neq W'$ within $\mathcal{W}$. A family $\mathcal{V}$ is termed uniformly bounded if its diameter, defined as $Diam(\mathcal{V})=\sup\{diam(V)\mid V\in\mathcal{V}\}$, is finite.
Let $\mathcal{U}$ be a cover of $X$. The Lebesgue number of $\mathcal{U}$, denoted by $L(\mathcal{U})$, is the maximum number $R$ such that for every subset $V$ of $X$ with diameter $diam(V)\leq R$, there exists a set $U\in\mathcal{U}$ such that $V\subseteq U$. The multiplicity of $\mathcal{U}$, denoted as $m(\mathcal{U})$, is the largest number of elements of $\mathcal{U}$ with nonempty intersection. The $R$-multiplicity of $\mathcal{U}$, denoted by $m_{R}(\mathcal{U})$, is the smallest integer $n$ such that for any $x\in X$, the ball $B(x,R)$ intersects at most $n$ sets of $\mathcal{U}$.
With this in mind we can recall the definition of asymptotic dimension function.
\begin{definition}[\cite{Dra06}]
Let $(X,d)$ be a metric space. For any $R>0$, let $\ad_{X}(R)$ be the smallest number over $n\in\mathbb{N}$ for which there exists a uniformly bounded cover $\mathcal{U}=\{U_{i}\}_{i\in I}$ of $X$ with $L(\mathcal{U})>R$ and $m(\mathcal{U})\leq n+1$. We call $\textnormal{ad}_{X}: \mathbb{R}^{+}\rightarrow\mathbb{N}$ the asymptotic dimension function of $X$.
\end{definition}
\begin{remark}
It is straightforward to see that
\begin{enumerate}
\item The function $\ad_{X}$ is nondecreasing, and the asymptotic dimension of $X$ is defined as follows:
\begin{equation}\label{eq 1}
\asdim(X)=\lim\limits_{R\rightarrow\infty}\ad_{X}(R).
\end{equation}
\item $\asdim (X)\leq d$ if and only if $\ad_{X}$ is constant and $\ad_{X}\leq d$.
\item If $\asdim(X)=\infty$, then both sides of equation \ref{eq 1} are infinite.
\end{enumerate}
\end{remark}
Analogous to the volume function, the growth type of the asymptotic dimension function is an essential concept.
Recall from \cite{Bel05} that for $f,g:\mathbb{R}^{+}\rightarrow\mathbb{R}^{+}$,
\begin{itemize}
\item we call $f$ is at most $g$, denoted as $f\preceq g$, if there exists $k\in\mathbb{N}$ such that $f(x)\leq kg(kx+k)+k$ for any $x>k$;
\item we call $f$ is equivalent to $g$, denoted as $f\approx g$, if both $f\preceq g$ and $g\preceq f$. Clearly, $\approx$ is an equivalence relation;
\item we define the growth type of $f$ to be the $\approx$-equivalence class of $f$;
\item we define the asymptotic dimension growth of $X$ to be the growth type of $\textnormal{ad}_{X}$.
\end{itemize}
%We review that the metric spaces $(X_{1},d_{1})$ and $(X_{2},d_{2})$ are quasi-isometric if there exist constants $\lambda\geq 1$, $\varepsilon\geq 0$ and $C\geq 0$ and a map $f: X\rightarrow Y$ so that for all $x$ and $y$ in $X_{1}$,
%$$
%\frac{1}{\lambda}d_{1}(x,y)-\varepsilon\leq d_{2}(f(x),f(y))\leq\lambda d_{1}(x,y)+\varepsilon
%$$
%and every point of $X_{2}$ lies in the $C$-neighborhood of the image of $f$.
Although the asymptotic dimension growth is not a coarse invariant like the finite asymptotic dimension, it turns out to be a quasi-isometric invariant by the following result. Thus, it is a group invariant because any two word metrics on a group are quasi-isometric.
\begin{proposition}[\cite{Bel05, Dra06}]
Let $X$ and $Y$ be two discrete metric spaces with bounded geometry. If $X$ and $Y$ are quasi-isometric, then $\textnormal{ad}_{X}\approx \textnormal{ad}_{Y}$. In particular, the asymptotic dimension growth is well-defined for finitely generated groups.
\end{proposition}
Below, we will present three equivalent definitions of asymptotic dimension growth. The first of these, as demonstrated in [\cite{ANWZ18}, Lemma 2.6], serves as a bridge between the preceding and subsequent definitions.
\begin{lemma}[\cite{ANWZ18}]\label{lemma 2.4}
Let $(X,d)$ be a metric space. For any $R>0$, we define $\widetilde{\textnormal{ad}_{X}}(R)$ to be the smallest number over $n\in\mathbb{N}$ for which there exists a uniformly bounded cover $\mathcal{V}=\{V_{j}\}_{j\in J}$ with $m_{R}(\mathcal{V})\leq n+1$. Then $\widetilde{\textnormal{ad}_{X}}\approx\textnormal{ad}_{X}$.
\end{lemma}
The succeeding proposition offers the second equivalent definition of asymptotic dimension growth. This pivotal step is essential in establishing the third equivalent definition.
\begin{proposition}\label{prop 2.5}
Let $(X,d)$ be a separable metric space. For any $R>0$, we define $\widehat{\ad_{X}}(R)$ to be the smallest number over $n\in\mathbb{N}$ for which there exist uniformly bounded, $R$-disjoint families $\mathcal{W}_{0},\cdots,\mathcal{W}_{n}$ of subsets of $X$ such that $\bigcup\limits_{l=0}^{n}\mathcal{W}_{l}$ is a cover of $X$. Then $\widehat{\ad_{X}}\approx\widetilde{\ad_{X}}$, hence $ \widehat{\ad_{X}}\approx\ad_{X}$.
\end{proposition}
\begin{proof}
Given $R>0$, it follows that $2R>R$. Suppose there exist uniformly bounded, $2R$-disjoint families $\mathcal{W}_{0},\cdots,\mathcal{W}_{\widehat{\ad_{X}}(2R)}$ of subsets of $X$ that cover $X$. Put $\mathcal{W}=\bigcup\limits_{l}\mathcal{W}_{l}$, which is obviously a uniformly bounded cover. We claim that $m_{R}(\mathcal{W})\leq\widehat{\ad_{X}}(2R)$. By the definition of $R$-multiplicity, for any $x\in X$, there are at most $\widehat{\ad_{X}}(2R)+1$ sets in $\mathcal{W}$ having non-empty intersection with the closed ball $B(x,R)$. Otherwise, there must be two sets $W$ and $W'$ from the same family $\mathcal{W}_{i}$ for some $i$ such that both $W\cap B(x, R)\neq\varnothing$ and $W'\cap B(x,R)\neq\varnothing$, since there are only $\widehat{\ad_{X}}(2R)+1$ different families. This implies that $d(W,W')\leq 2R$, contradicting the $2R$-disjoint property of the family $\mathcal{W}_{i}$. Therefore, $m_{R}(\mathcal{W})\leq\widehat{\ad_{X}}(2R)$ as claimed, and thus yields $\widetilde{\ad_{X}}\preceq\widehat{\ad_{X}}$.
Conversely, given $R>0$, suppose $\mathcal{V}=\{V_{j}\}_{j\in J}$ is a unifromly bounded cover with $m_{R}(\mathcal{V})\leq \widetilde{\textnormal{ad}_{X}}(R)+1$. Since $X$ is separable, $X$ is second countable, implying that any cover of $X$ must have a countable subcover. Therefore, we can assume that the cover $\mathcal{V}$ is countable. We initially claim that there exist uniformly bounded, $R$-disjoint families $\mathcal{V}_{0},\cdots,\mathcal{V}_{\widetilde{\textnormal{ad}_{X}}(R)}$ of subsets of $X$ covering $X$. This claim will be demonstrated through a three-step process.
\textbf{Step 1}. In the first step, we construct an irreducible cover of $X$. Let $\mathcal{U}=\{U_{i}\}_{i\in I}$ be a subcover of $\mathcal{V}$ such that
$$
V_{j}\bigg\backslash\left(\bigcup\limits_{V'\in\mathcal{V}\backslash\{V_{j}\}}V'\right)\neq\varnothing, \text{ for all } V_{j}\in\mathcal{V},
$$
which forms an irreducible cover of $X$. As $m_{R}(\mathcal{U})\leq m_{R}(\mathcal{V})\leq\widetilde{\textnormal{ad}_{X}}(R)+1$, we have $m_{R}(\mathcal{U})\leq\widetilde{\textnormal{ad}_{X}}(R)+1$ ( where the equality can be reached here ).
%In the first step, we remove the overlapping sets from the cover $\mathcal{V}$. For this, if $V_{j}\in\mathcal{V}$ and $V_{j}\subseteq\bigcup\limits_{k=1}^{m}V_{j_{k}}$ for some integer $m\in\mathbb{N}$, then we remove it from the cover. We repeat this process until there are no such sets, then we get a subcover of $\mathcal{V}$, denoted as $\mathcal{U}=\{U_{i}\}_{i\in I}$, which is a cover of $X$. Since $m_{R}(\mathcal{U})\leq m_{R}(\mathcal{V})$ and $m_{R}(\mathcal{V})\leq\widetilde{\textnormal{ad}_{X}}(R)+1$, $m_{R}(\mathcal{U})\leq\widetilde{\textnormal{ad}_{X}}(R)+1$ ( here the equality can be obtained ).
\textbf{Step 2}. In this step, we build $\widetilde{\ad_{X}}(R)+1$ boxes. Since $m_{R}(\mathcal{U})\leq \widetilde{\ad_{X}}(R)+1$ and the equality is achievable, there exists $x\in X$ such that $B(x,R)$ intersects $\widetilde{\ad_{X}}(R)+1$ elements $U_{0},\cdots, U_{\widetilde{\ad_{X}}(R)}$ of $\mathcal{V}$. Consequently, $d(U_{i}, U_{j})\leq 2R$, for all $i,j\in\{0,\cdots,\widetilde{\ad_{X}}(R)\}$ with $i\neq j$. We then put $U_{0},\cdots,U_{\widetilde{\ad_{X}}(R)}$ into $\widetilde{\ad_{X}}(R)+1$ distinct boxes.
\textbf{Step 3}. In the final step, we allocate the remaining sets of the cover $\mathcal{U}$ into these $\widetilde{\ad_{X}}(R)+1$ boxes. Set $\mathcal{U}\big\backslash\bigcup\limits_{i=0}^{\widetilde{\ad_{X}}(R)}U_{i}=\{U_{k}\}_{k\in I}$, and let $\Box_{l}$ represent the $l$-th box. We next claim that for any $x\in U_{k}$, there exists $l\in\{0,\cdots,\widetilde{\ad_{X}}(R)\}$ such that
$$
d(x,B_{l})>R, \forall B_{l}\in \Box_{l}.
$$
Otherwise, there exists $y\in U_{k}$ such that for all $l\in\{0,\cdots,\widetilde{\textnormal{ad}_{X}}(R)\}$, there is a set $B_{l}\in\Box_{l}$ such that
$$
d(y, B_{l})\leq R,
$$
which implies that $B(y, R)$ intersects $\widetilde{\ad_{X}}(R)+2$ sets $U_{k},B_{0},\cdots, B_{\widetilde{\ad_{X}}(R)}$ of $\mathcal{U}$, which contradicts that $m_{R}(\mathcal{U})\leq \widetilde{\ad_{X}}(R)+1$. Therefore, for any $x\in U_{k}$, $x$ must belong to some box. Subsequently, we divide $U_{k}$ into several pieces and place them into boxes sequentially. Let
\begin{align*}
U^{(0)}_{k}&=\{x\in U_{k}\mid d(x, B_{0})>R, \text{ for all }B_{0}\in\Box_{0}\},\\
U^{(l)}_{k}&=\{x\in U_{k}\big\backslash\bigcup\limits_{n=0}^{l-1}U^{(n)}_{k}\mid d(x, B_{l})>R, \text{ for all }B_{l}\in\Box_{l}\}, l=1,\cdots,\widetilde{\ad_{X}}(R).
\end{align*}
Then we put $U^{(l)}_{k}$ into the box $\Box_{l}$ for $l=0,\cdots,\widetilde{\ad_{X}}(R)$ in ascending order. If $U^{(l)}_{k}$ is the empty set, we proceed to the next box. Observe that $U^{(l)}_{k}$ is a subset of $U_{k}$, and $U_{k}=\bigsqcup\limits_{l=0}^{\widetilde{\ad_{X}}(R)}U^{(l)}_{k}$. Similarly, we can divide all elements of $\mathcal{U}\big\backslash\bigcup\limits_{j=0}^{\widetilde{\ad_{X}}(R)}U_{j}$ into $\widetilde{\ad_{X}}(R)+1$ boxes. Consequently, we obtain new boxes and for all $l\in\{0,\cdots,\widetilde{\ad_{X}}(R)\}$,
$$
d(B_{l},B'_{l})>R, \text{ for every }B_{l}\neq B'_{l} \text{ in }\Box_{l}.
$$
Let $\mathcal{W}_{l}=\{B_{l}\mid B_{l}\in\Box_{l}\}$, then each $\mathcal{W}_{l}$ is a uniformly bounded, $R$-disjoint family and $X=\bigcup\limits_{l=0}^{\widetilde{\ad_{X}}(R)}\mathcal{W}_{l}$. By definition, it is evident that $\widehat{\ad_{X}}(R)\leq \widetilde{\ad_{X}}(R)$, which yields $ \widehat{\ad_{X}}\preceq \widetilde{\ad_{X}}$.
By the transitivity of the equivalence relation, we conclude that $\textnormal{ad}_{X}\approx\widehat{\textnormal{ad}_{X}}$ by Lemma \ref{lemma 2.4}.
\end{proof}
In the second part of this subsection, we present several typical examples of groups and spaces exhibiting asymptotic dimension growth. We primarily focus on four classes of growth types: exponential growth, subexponential growth, polynomial growth, and slow growth. We define a monotone increasing function $f:\mathbb{R}^{+}\rightarrow\mathbb{R}^{+}$ as having subexponential growth if for any $b>1$,
$$
\lim\limits_{t\rightarrow\infty}\frac{f(t)}{b^{t}}=0.
$$
Observe that if $f$ has polynomial growth, it inherently exhibits subexponential growth, since polynomial functions grow at a slower rate than exponential functions. We say that $X$ has polynomial (resp. subexponential, exponential) asymptotic dimension growth, if there exists a polynomial (resp. subexponential, exponential) function $f$ such that $\ad_{X}\preceq f$. Recall from \cite{Dra00} that $X$ has slow asymptotic dimension growth, if there exists a linear function $f$ such that $\ad_{X}\preceq f$.
%The wreath product of $A$ by $\Lambda$, denoted by $A\wr\Lambda$, is the semi-direct product
%$$
%A\wr\Lambda:=(\bigoplus\limits_{\lambda\in\Lambda}A)\rtimes\Lambda,
%$$
%where $\Lambda$ acts by shifting the copies of $A$ in $\bigoplus\limits_{\lambda\in\Lambda}A$. In other words, the group multiplication is given by $(f,\gamma)(f',\gamma'):=(f+f'(\gamma^{-1}\cdot),\gamma\gamma').$
Similar to volume growth, a finitely generated group has at most exponential asymptotic dimension growth.
\begin{example}[\cite{Dra06}]
For a finitely generated group $\Gamma$ with the word metric, there exists $\alpha>0$ such that $\ad_{\Gamma}(R)\preceq e^{\alpha R}$.
\end{example}
Ozawa has demonstrated that subexponential asymptotic dimension growth implies Yu's property A \cite{Oza12}, thereby identifying two classes of spaces having this property as follows.
\begin{example}[\cite{ANWZ18}]\label{example 2.7}
Let $X$ be a geodesic coarse median space with finite rank and at most exponential volume growth, then $X$ has subexponential asymptotic dimension growth.
\end{example}
\begin{example}[\cite{Dra06}]
Let $N$ be a finitely generated nilpotent group and $G$ be a finitely generated group with finite asymptotic dimension. Then the wreath product $N\wr G$ has polynomial asymptotic dimension growth. In particular, $\mathbb{Z}\wr\mathbb{Z}$ has polynomial asymptotic dimension growth.
\end{example}
The subsequent example demonstrates that given a strictly increasing function $f$, we can create a topological space with asymptotic dimension function $f$. This provides an approach to construct spaces that have slow asymptotic dimension growth when $f$ has slow growth.
\begin{example}[\cite{CFY08}]
Let $f:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a strictly increasing function. Let
$G$ be the direct product given by $G=\bigoplus\limits_{n\in\mathbb{N}}f(n)\mathbb{Z}$, and endow $G$ with the direct sum metric. Then $G$ is a discrete metric space of bounded geometry with asymptotic dimension function $f$.
\end{example}
\subsection{The case of coarse spaces}
Our objective in this subsection is to establish the third equivalent definition of asymptotic dimension growth. To achieve this, we consider a coarse space for which the coarse structure is induced from the metric. We translate into the language of coarse geometry the two conditions discussed in Proposition \ref{prop 2.5} that define asymptotic dimension growth, and we demonstrate their equivalence.
%In this subsection, we specialize in the case of coarse spaces. In the case of a coarse space for which the coarse structure is induced from the metric, we observe that $R$-disjointness and uniformly boundedness can be converted into $E$-separatedness and $E$-boundedness in coarse geometry.
%Further, we show that statements about finite asymptotic dimensions on metric spaces and coarse spaces are equivalent. Moreover, we will define the asymptotic dimension function for such coarse spaces, and prove that the function is the same as the function defined in Proposition \ref{prop 2.5}. In the rest of this paper, we will study separable coarse spaces.
We initially recollect the notion of coarse spaces, referring the reader to the monograph \cite{Roe03} for a comprehensive treatment of coarse spaces.
\begin{definition}[\cite{Roe03}]
Let $X$ be a set, and let $\mathcal{E}$ be a collection of subsets of $X\times X$. We say that $\mathcal{E}$ is a coarse structure on $X$ if it satisfies the following axioms:
\begin{itemize}
\item The diagonal $\bigtriangleup=\{(x,x)\mid x\in X\}$ is a member of $\mathcal{E}$;
\item $E\in\mathcal{E}$ and $F\subset E$ implies $F\in\mathcal{E}$;
\item $E,F\in\mathcal{E}$ implies $E\cup F\in \mathcal{E}$;
\item $E\in\mathcal{E}$ implies $E^{-1}:=\{(y,x)\mid (x,y)\in E\}\in\mathcal{E}$;
\item $E, F\in\mathcal{E}$ implies $E\circ F:=\{(x,y)\mid \exists z, s.t. (x,z)\in E \text{ and } (z,y)\in F\}\in\mathcal{E}$;
\end{itemize}
The pair $(X,\mathcal{E})$ is called a coarse space, and the elements of $\mathcal{E}$ are called controlled sets or entourages.
\end{definition}
\begin{remark}\label{rem 2.11}
Below, we examine the special case of the coarse structure derived from a metric.
Let $(X,d)$ be a discrete metric space. For $R>0$, we define a controlled tube as follows:
$$E_{R}:=\{(x,y)\in X\times X\mid d(x,y)\leq R\}.$$
Then, the controlled sets are defined as subsets $E$ of $X\times X$ contained within some controlled tube $E_{R}$ for some $R$, thereby forming a coarse structure on $X$.
\end{remark}
The following definition introduces two concepts of $E$-separatedness and $E$-boundedness within the realm of coarse geometry.
\begin{definition}\cite{Roe03}
Let $X$ be a coarse space, and let $E$ be a controlled set for $X$. A family $\mathcal{U}=\{U_{i}\}_{i\in I}$ subsets of $X$ is:
\begin{enumerate}
\item $E$-separated if $U_{i}\times U_{j}\cap E=\varnothing$ when $i\neq j$;
\item $E$-bounded if $U_{i}\times U_{i}\subset E$ for each $i\in I$.
\end{enumerate}
\end{definition}
By translating the concepts of $R$-disjointness and uniformly boundedness from Proposition \ref{prop 2.5} into the terms of $E_{R}$-separatedness and $F$-boundedness within the framework of coarse geometry, we obtain the following third equivalent definition. This definition will be used in the proof of Theorem \ref{th 4.16} in Section \ref{section: dad}.
%In the case of coarse spaces, we give an alternative definition of the asymptotic dimension function equal to the function defined in Proposition \ref{prop 2.5}. This definition plays an important role in the proof of Theorem \ref{th 4.16} in Section \ref{section: dad}.
\begin{lemma}\label{lemma 2.13}
Let $X$ be a separable metric space. For any $R>0$, we define $\widecheck{\ad_{X}}(R)$ to be the smallest number over all $m\in\mathbb{N}$ with the following property: for the controlled set $E_{R}$, there exist a controlled set $F$ and a cover $\mathcal{U}=\{U_{i}\}_{i\in I}$ of $X$ such that $\mathcal{U}$ is $F$-bounded and such that $\mathcal{U}$ admits a decomposition
$$
\mathcal{U}=\mathcal{U}_{0}\sqcup\cdots\sqcup\mathcal{U}_{m}
$$
and each $\mathcal{U}_{i}$ is $E_{R}$-separated. Then $\widecheck{\ad_{X}}= \widehat{\ad_{X}}$, hence $\widecheck{\ad_{X}}\approx{\ad_{X}}$.
\end{lemma}
\begin{proof}
Given $R>0$. Suppose that $\mathcal{U}_{0},\cdots,\mathcal{U}_{\widehat{\ad_{X}}(R)}$ are uniformly bounded, $R$-disjoint families of subsets of $X$ covering $X$. Since each family $\mathcal{U}_{i}$ is uniformly bounded, the cover $\mathcal{U}=\bigsqcup\limits_{i=0}^{\widehat{\ad_{X}}(R)}\mathcal{U}_{i}$ is also uniformly bounded. Consequently, there exists a constant $C>0$ such that for each $U\in\mathcal{U}$, $\diam(U)\leq C$, which implies that each set $U\times U$ is contained in a controlled set $E_{C}$, thus the cover $\mathcal{U}$ is $E_{C}$-bounded. Furthermore, as each family $\mathcal{U}_{i}$ is $R$-disjoint, then for every $U\neq U'$ in $\mathcal{U}_{i}$, $d(U,U')>R$. This ensures that $(U\times U')\cap E_{R}=\varnothing$, indicating that each $\mathcal{U}_{i}$ is $E_{R}$-separated. Hence, $\widecheck{\ad_{X}}(R)\leq\widehat{\ad_{X}}(R)$, which yields $\widecheck{\ad_{X}}\leq \widehat{\ad_{X}}$.
Conversely, given $R>0$, for the controlled set $E_{R}$, there exist a controlled set $F$ and a cover $\mathcal{U}=\{U_{i}\}_{i\in I}$ of $X$ such that $\mathcal{U}$ is $F$-bounded and admits a decomposition
$$
\mathcal{U}=\mathcal{U}_{0}\sqcup\cdots\sqcup\mathcal{U}_{\widecheck{\ad_{X}}(R)}
$$
such that each $\mathcal{U}_{i}$ is $E_{R}$-separated. Since $F$ is a controlled set, $F\subseteq E_{C}$ for some $C>0$. As $\mathcal{U}$ is $F$-bounded, each set $U_{i}\times U_{i}$ is contained in $ E_{C}$, implying that $diam(U_{i})\leq C$, so $\mathcal{U}$ is a uniformly bounded cover. Since each $\mathcal{U}_{i}$ is $E_{R}$-separated, for every $U\neq U'$ in $\mathcal{U}_{i}$, $(U\times U')\cap E_{R}=\varnothing$, thus $d(U, U')>R$, which implies that each $\mathcal{U}_{i}$ is $R$-disjoint.
Therefore, $\widehat{\ad_{X}}(R)\leq \widecheck{\ad_{X}}(R)$, which yields $ \widehat{\ad_{X}}\leq \widecheck{\ad_{X}}$. Hence, $\widehat{\ad_{X}}=\widecheck{\ad_{X}}$, which implies that $\ad_{X}\approx\widecheck{\ad_{X}}$ by Proposition \ref{prop 2.5}.
\end{proof}
Based on the preceding results, we can use any one of $\ad_{X}$, $\widetilde{\ad_{X}}$, $\widehat{\ad_{X}}$, or $\widecheck{\ad_{X}}$ to define the asymptotic dimension growth.
By making use of a proof approach similar to the above, we can demonstrate that the statement about finite asymptotic dimension on the metric space is equivalent to the statement on the coarse space as stated in Remark \ref{rem 2.11}.
\begin{lemma}
Let $X$ be a coarse space. Then the following two conditions are equivalent:
\begin{enumerate}
\item for any controlled set $E$, there exist a controlled set $F$ and a cover $\mathcal{U}=\{U_{i}\}_{i\in I}$ of $X$ such that $\mathcal{U}$ is $F$-bounded and such that $\mathcal{U}$ admits a decomposition
$$
\mathcal{U}=\mathcal{U}_{i}\sqcup\cdots\sqcup\mathcal{U}_{n}
$$
such that each $\mathcal{U}_{i}$ is $E$-separated.
\item for any $R>0$, there exist uniformly bounded, $R$-disjoint families $\mathcal{U}_{0},\cdots,\mathcal{U}_{n}$ of subsets of $X$ such that $\bigcup\limits_{i=0}^{n}\mathcal{U}_{i}$ covers $X$.
\end{enumerate}
\end{lemma}
%\begin{remark}
%We can complete the proof by the method analogous to the above lemma.
%\end{remark}
\section{Dynamic asymptotic dimension growth for group actions}
In this section, we will introduce the dynamic asymptotic dimension function and the dynamic asymptotic dimension growth for actions of discrete groups on compact spaces. To investigate the dynamic asymptotic dimension growth of group actions, we will define the $h$-BLR action for a non-decreasing function $h$, which draws inspiration from the conditions formulated by Bartels, Lück, and Reich \cite{BLR08, Bar16}. Additionally, we will establish an equivalent definition of the $h$-BLR action under specific assumptions. Furthermore, we demonstrate that an $h$-BLR action has dynamic asymptotic dimension function at most $h$. By applying this result within the realm of coarse geometry, we can obtain Theorems \ref{th 4.21} and \ref{th 4.22}. This allows us to present several typical examples of subexponential dynamic asymptotic dimension growth in group actions.
%Guentner, Willett, and Yu introduced the concept of a $d$-BLR action \cite{GWY17}, drawing inspiration from the works of Bartels, Lück, and Reich regarding the Farrell-Jones conjecture \cite{BLR08, Bar16}. Building on this foundation, we extend this notion to the growth case by defining the $h$-BLR action for a nondecreasing function $h$. Additionally, under certain assumptions, we establish an equivalent definition of the $h$-BLR action, which is closely tied to the asymptotic dimension growth of groups. Furthermore, we show that an $h$-BLR action has dynamic asymptotic dimension function at most $h$. By applying this result to within the realm of coarse geometry, we will bridge the asymptotic dimension growth for countable discrete groups with the dynamic asymptotic dimension growth for their actions on compact spaces, as detailed in Section \ref{section: dad}. This enables us to present several typical examples of subexponential dynamic asymptotic dimension growth in group actions.
%In this section, we will introduce the dynamic asymptotic dimension function and the dynamic asymptotic dimension growth for group actions on compact spaces. For a nondecreasing function $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$, we define the $h$-BLR action, which is inspired by the work of Bartels, Reich, and L\"uck \cite{BLR08, Bar16}. Furthermore, we will show that an $h$-BLR action has dynamic asymptotic dimension function at most $h$. Also, when the function $h$ grows slower than $\sqrt{x}$, we will obtain an equivariant condition of the BLR condition, which is similar to the definition of asymptotic dimension growth for groups. Moreover, it bridges the dynamic asymptotic dimension growth for group actions with the asymptotic dimension growth for groups in the sense of coarse geometry in Section \ref{section: dad}, which allows us to construct several exciting examples of dynamic asymptotic dimension growth for group actions.
\subsection{The case of group actions on compact spaces}
In this subsection, we will introduce the concept of dynamic asymptotic dimension growth for actions by discrete groups on compact Hausdorff spaces. This notion extends the concept of finite dynamic asymptotic dimension for such actions. By integrating examples of groups that demonstrate subexponential asymptotic dimension growth from the previous section with Theorem \ref{th 4.22} discussed in the following section, we can present several intriguing examples of group actions that exhibit subexponential dynamic asymptotic dimension growth.
To introduce our primary definition, we initially revisit the concept of length functions on discrete groups.
\begin{definition}
Let $\Gamma$ be a discrete group. A proper length function on $\Gamma$ is a function $\ell: \Gamma\rightarrow [0,\infty)$ satisfying:
\begin{enumerate}
\item $\ell(\gamma)=0$ if and only if $\gamma=e$;
\item $\ell(\gamma_{1}+\gamma_{2})\leq\ell(\gamma_{1})+\ell(\gamma_{2})$, for all $\gamma_{1},\gamma_{2}\in\Gamma$;
\item $\ell(\gamma^{-1})=\ell(\gamma)$, for all $\gamma\in\Gamma$;
\item $\{\gamma\in\Gamma\mid\ell(\gamma)\leq R\}$ is finite, for all $R\geq 0$.
\end{enumerate}
\end{definition}
By applying the concept of proper length functions on discrete groups, we can define the dynamic asymptotic dimension growth for actions of discrete groups on compact spaces, which generalizes the finite dynamic asymptotic dimension for these actions.
\begin{definition}\label{def 3.2}
Let $\Gamma$ be a discrete group, and let $X$ be a compact Hausdorff space. Let $\Gamma\curvearrowright X$ be an action by homeomorphisms, and let $\ell_{\Gamma}: \Gamma\rightarrow [0,\infty)$ be a proper length function. For any $R>0$, let $\dad_{\Gamma\curvearrowright X}(R)$ be the smallest number over $m\in\mathbb{N}$ with the following property: for the finite subset $E=\{\gamma\in\Gamma\mid\ell_{\Gamma}(\gamma)<R\}$ of $\Gamma$, there exists an open cover $\{U_{0},\cdots, U_{m}\}$ of $X$ such that for each $i\in\{0,\cdots,m\}$, the set
$$
\begin{Bmatrix}&\Big|& \text{ there exist }x\in U_{i}\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}x\in U_{i}\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}
$$
is finite. We call $\dad_{\Gamma\curvearrowright X}:\mathbb{R}^{+}\rightarrow\mathbb{N}$ the dynamic asymptotic dimension function of $\Gamma\curvearrowright X$, and call the growth type of the function $\dad_{\Gamma\curvearrowright X}$ the dynamic asymptotic dimension growth of $\Gamma\curvearrowright X$. If no such function $\dad_{\Gamma\curvearrowright X}$ exists, we say that $\Gamma\curvearrowright X$ has infinite dynamic asymptotic dimension.
\end{definition}
\begin{remark}
\begin{enumerate}
\item The function $\dad_{\Gamma\curvearrowright X}$ is non-decreasing, and the dynamic asymptotic dimension of the action $\Gamma\curvearrowright X$ is defined as follows:
\begin{equation}\label{eq 2}
\textnormal{dasdim} (\Gamma\curvearrowright X)=\lim\limits_{R\rightarrow\infty}\dad_{\Gamma\curvearrowright X}(R).
\end{equation}
\item $\dasdim(\Gamma\curvearrowright X)\leq d$ for some finite $d$ if and only if $\dad_{\Gamma\curvearrowright X}$ is a constant function and $\dad_{\Gamma\curvearrowright X}\leq d$.
\item If $\dasdim(\Gamma\curvearrowright X)=\infty$, then both sides of equation (\ref{eq 2}) are infinite.
\item $E=E^{-1}$ and $e\in E$ since $\ell$ is a proper length function.
\end{enumerate}
\end{remark}
\begin{remark}
Dynamic asymptotic dimension has been estimated for actions of discrete groups on compact spaces. For instance, the dynamic asymptotic dimension of an action by an infinite virtually cyclic
group on a compact Hausdorff space is always one, provided the action has the marker property \cite{ALSS21}. Furthermore, the dynamic asymptotic dimension of a free action of a countable discrete group on a zero-dimensional second countable compact Hausdorff space is either infinite or equal to the asymptotic dimension of the group \cite{CJMST23}. We suspect this result may be generalized to the growth case, but we do not seriously pursue this. Moreover, for an infinite countable discrete group $\Gamma$ with asymptotic dimension $d$, the dynamic asymptotic dimension of a free, minimal action of $\Gamma$ on the Cantor set is at most $d$ \cite{GWY17}. In the next section, we will extend this result to subexponential dynamic asymptotic dimension growth. Combining this generalization with examples of groups with subexponential asymptotic dimension growth, we can provide several examples of group actions with subexponential dynamic asymptotic dimension growth.
\end{remark}
Here is a classic example of group actions with polynomial dynamic asymptotic dimension growth.
\begin{example}
Let $N$ be a finitely generated nilpotent group, and let $\Gamma$ be a finitely generated group with finite asymptotic dimension. Then the wreath product $N\wr\Gamma$ admits a free, minimal action on the Cantor set which has polynomial dynamic asymptotic dimension growth. In particular, $\mathbb{Z}\wr \mathbb{Z}$ admits a free, minimal action on the Cantor set which has polynomial dynamic asymptotic dimension growth.
\end{example}
Another example is the slow dynamic asymptotic dimension growth of group actions.
\begin{example}
Let $f:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a strictly increasing function with $f\preceq x$. Let
$G$ be the direct product given by $G=\bigoplus\limits_{n\in\mathbb{N}}f(n)\mathbb{Z}$, and endow $G$ with the direct sum metric. Then $G$ admits a free, minimal action on the Cantor set which has slow dynamic asymptotic dimension growth.
\end{example}
\subsection{Bartels-L\"uck-Reich condtions}
The Bartels-L\"uck-Reich conditions, abbreviated as BLR conditions, originated from the seminal works of Bartels, L\"uck, and Reich regarding the Farrell-Jones conjecture \cite{BLR08, Bar16}. Subsequently, Guentner, Willett, and Yu in the paper \cite{GWY17} further summarized and developed these conditions to study the finite dynamic asymptotic dimension of group actions, thereby giving birth to the concept of a $d$-BLR action. In this subsection, we aim to extend this notion to the growth case by defining the $h$-BLR action for a nondecreasing function $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$. Furthermore, when the function $h$ grows at a rate slower than $\sqrt{x}$, we establish an alternative equivalent definition for this action, which is intimately connected to the asymptotic dimension growth of a group. Moreover, we demonstrate that an $h$-BLR action has dynamic asymptotic dimension function at most $h$.
To define the $h$-BLR action, we need a couple of definitions.
\begin{definition}
Let $V$ be a countable set. The space of probability measures on $V$ is
$$
\Prob(V):=\{\mu\in \ell^{1}(V)\mid \mu(v)\geq 0,\forall v\in V\text{ and }\Vert \mu\Vert_{1}=1\},
$$
equipped with the metric
$$
d\left(\sum\limits_{v\in V}b_{v}v, \sum\limits_{v\in V}c_{v}v\right):=\sum\limits_{v\in V}\vert b_{v}-c_{v} \vert.
$$
coming from $\ell^{1}$-norm.
\end{definition}
\begin{remark}
Elements of $\Prob(V)$ can be written as $\mu=\sum\limits_{v\in V}t_{v}v$, where $\mu(v)=t_{v}$ in $[0,1]$ and $\sum\limits_{v\in V}t_{v}=1$.
Write $\Prob_{n}(V)$ for the subset of $\Prob(V)$ consisting of measures supported on at most $n+1$ points. Say that a simplicial complex is a subspace $C$ of some $\bigcup\limits_{n=0}^{\infty}\Prob_{n}(V)$ such that $V\subseteq C$ and
$$
\left\{\sum\limits_{v\in V}t_{v}v\in\bigcup\limits_{n=0}^{\infty}\Prob_{n}(V) \bigg| t_{v}\neq 0\text{ for } v\in\supp(\mu)\right\}\in C.
$$
Clearly, $V$ is the vertex set of $C$.
\end{remark}
Below we mainly focus on a simplicial complex equipped with a simplicial action, and let us first review this notion. Let $\Gamma$ be a discrete group, and let $C$ be a simplicial complex with the vertex set $V$. A simplicial action $g$ of $\Gamma$ on $C$ is an action that is induced from some action of $\Gamma$ on $V$ via the formula
$$
g\left(\sum\limits_{v\in V}t_{v}v\right)=\sum\limits_{v\in V}t_{v}(gv).
$$
A simplicial action is isometric.
The concept of an almost equivariant map from a topological space to a metric space is defined below.
\begin{definition}[\cite{GWY17}]
Let $X$ be a topological space, and let $Y$ be a metric space. Assume that $\Gamma$ is a discrete group acting on $X$ by homeomorphisms, and on $Y$ by isometries. Let $E$ be a subset of $\Gamma$ and $\varepsilon>0$. Say that a map $f:X\rightarrow Y$ is $(E,\varepsilon)$-equivariant if
$$
\sup\limits_{x\in X}d_{Y}(f(gx),gf(x))<\varepsilon, \forall g\in E
$$
\end{definition}
The next technical lemma will be used in the proof of Proposition \ref{prop 3.12} and Theorems \ref{th 3.15} and \ref{th 4.22}.
\begin{lemma}[\cite{GWY17}]\label{lemma 3.10}
Let $f:X\rightarrow C$ be a continuous $(E,\varepsilon)$-equivariant map as in the above definition, where $X$ is compact and $E$ is finite. Then there exists a finite subset $S$ of the vertex set of $C$ and an $(E,\varepsilon)$-equivariant map $f':X\rightarrow C$ such that $f'(X)\subseteq \Prob(S)\cap C$.
\end{lemma}
The underlying concept serves as the fundamental element of the Bartels-Lück-Reich conditions, which can be applied to families of finite subgroups or virtually cyclic subgroups.
\begin{definition}[\cite{GWY17}]
Let $C$ be a simplicial complex equipped with a simplicial action of a discrete group $\Gamma$. Let $\mathcal{F}$ be a family of subgroups of $\Gamma$ satisfying the following conditions:
\begin{enumerate}
\item $\mathcal{F}$ is closed under conjugation;
\item $\mathcal{F}$ is closed under taking subgroups;
\item $\mathcal{F}$ is closed under taking finite index supergroups.
\end{enumerate}
Then $C$ is called a $(\Gamma,\mathcal{F})$-complex if the stabilizer of every vertex in $C$ is an element of $\mathcal{F}$.
\end{definition}
We now define the $h$-BLR action, which is close to the amenability of group actions.
\begin{definition}
Let $\Gamma\curvearrowright X$ be an action with $X$ compact, and let $\mathcal{F}$ be a family of subgroups of $\Gamma$. Let $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function. We say that the action is $h$-BLR for $\mathcal{F}$ if for any $R>0$, for the finite subset $E=\{\gamma\in\Gamma\mid\ell_{\Gamma}(\gamma)<R\}$ of $\Gamma$ and any $\varepsilon>0$, there exist a $(\Gamma,\mathcal{F})$-complex $C$ of dimension at most $h(R)$ and a continuous $(E,\varepsilon)$-equivariant map $f:X\rightarrow C$.
\end{definition}
If $h\preceq x^{\alpha}(0<\alpha<\frac{1}{2})$, we can obtain an equivalent definition of the $h$-BLR action, which is similar to the definition of asymptotic dimension growth for a group.
\begin{proposition}\label{prop 3.12}
Let $\Gamma\curvearrowright X$ be an action with $X$ compact, and let $\mathcal{F}$ be a family of subgroups of $\Gamma$. Let $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function with $h\preceq x^{\alpha}(0<\alpha<\frac{1}{2})$. Then the following are equivalent.
\begin{enumerate}
\item The action is $h$-$BLR$ for $\mathcal{F}$.
\item For any $R>0$, there exists a nondecreasing function $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ such that for the finite subset $E=\{\gamma\in\Gamma\mid\ell_{\Gamma}(\gamma)<R\}$ of $\Gamma$, there exists a $\Gamma$-equivaraint open cover $\mathcal{U}$ of $X\times\Gamma$ such that the following hold.
\begin{enumerate}
\item For each $U\in\mathcal{U}$ and $\gamma\in\Gamma$, $\gamma U=U$ or $\gamma U\cap U=\varnothing$.
\item For each $U\in\mathcal{U}$, $\{\gamma\in\Gamma\mid \gamma U=U\}\in\mathcal{F}$.
\item $m(\mathcal{U})\leq h(R)+1$.
\item $\mathcal{U}/\Gamma$ is finite.
\item For each $\gamma\in\Gamma$ and $x\in X$, there exists $U\in\mathcal{U}$ such that $\{x\}\times\gamma E\subset U$.
\end{enumerate}
\end{enumerate}
\end{proposition}
\begin{remark}
This proposition can be proved in a similar way as shown in [\cite{GWY17}, Proposition 4.5]. In the proof of $(1)$ implying $(2)$, there are no extra restrictions on the growth rate of $h$, while in the case of $(2)$ implying $(1)$, the function $h$ is required to grow at a rate slower than $\sqrt{x}$.
\end{remark}
The following theorem shows that an $h$-BLR action has dynamic asymptotic dimension function at most $h$, which plays an important role in the proof Theorem \ref{th 4.21}.
\begin{theorem}\label{th 3.15}
Let $\Gamma\curvearrowright X$ be an action with $X$ compact, and let $\mathcal{F}$ be the family of finite subgroups of $\Gamma$. Let $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function. If $\Gamma\curvearrowright X$ is an $h$-BLR action for the family $\mathcal{F}$, then it has dynamic asymptotic dimension function at most $h$.
\end{theorem}
\begin{proof}
We divide the proof into three steps. First, we construct a desired open cover of $X$. Assume that $\Gamma\curvearrowright X$ is an $h$-$BLR$ action for $\mathcal{F}$. Given $R>0$, there exists a nondecreasing function $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ such that for the finite subset $E=\{\gamma\in\Gamma\mid\ell(\gamma)<R\}$ of $\Gamma$ and $\varepsilon=\frac{1}{3}10^{-h(R)}$, there exist a $(\Gamma,\mathcal{F})$-complex $C$ of dimension at most $h(R)$ and a continuous $(E,\varepsilon)$-equivariant map $f: X\rightarrow C$. By Lemma \ref{lemma 3.10}, we may assume that there exists a finite subset $S$ of the vertex set of $C$ such that $f(X)\subseteq C\cap \Prob(S)$. Define
$$
F:=\{g\in\Gamma\mid gS\cap S\neq\varnothing\}.
$$
Since $S$ is finite and the stabilizer of each vertex in $S$ is finite, $F$ is a finite subset of $\Gamma$. Observe that $C$ is a $(\Gamma,\mathcal{F})$-complex $C$ of dimension at most $h(R)$, there exists an open cover $\{V_{0},\cdots, V_{h(R)}\}$ of $C$ as in [\cite{GWY17}, Lemma 4.1]. For each $i\in\{0,\cdots, h(R)\}$, and each $i$-simplex $\Delta$, we define $U_{i}:=f^{-1}(V_{i})$, $U_{i\Delta}:=f^{-1}(V_{i\Delta})$. Therefore, $\{U_{0},\cdots, U_{h(R)}\}$ is an open cover of $X$, and $U_{i}=\bigsqcup\limits_{\Delta\text{ an i-simplex }}U_{i\Delta}$. Indeed, $f(X)\subseteq C\cap \Prob(S)\subseteq\bigcup\limits_{i=0}^{h(R)}V_{i}$, which implies that $X\subseteq f^{-1}(\bigcup\limits_{i=0}^{h(R)}V_{i})=\bigcup\limits_{i=0}^{h(R)}f^{-1}(V_{i})=\bigcup\limits_{i=0}^{h(R)}U_{i}$. By [\cite{GWY17}, Lemma 4.1], we know that $V_{i\Delta}$ and $V_{i\Delta'}$ are distinct for different $i$-simplices, which implies that $f^{-1}(V_{i\Delta})$ and $f^{-1}(V_{i\Delta'})$ are distinct, hence $U_{i\Delta}$ and $U_{i\Delta'}$ are distinct.
Next, we claim that the following statement holds for each $i$ and $i$-simplex $\Delta$:
$$
x\in U_{i\Delta}, gx\in U_{i}, g\in E\Rightarrow gx\in U_{i g\Delta}.
$$
Indeed, we see that $gx\in U_{i g\Delta'}$ for some $i$-simplex $\Delta'$. Since $f$ is $(E,\varepsilon)$-equivariant, we have
$$
d(f(gx),gf(x))<\varepsilon.
$$
Since $x\in U_{i\Delta}$, we have $f(x)\in V_{i \Delta}$, which implies that $gf(x)\in V_{i g\Delta}$. Clearly, $gx\in f^{-1}(V_{i\Delta'})$ implies $f(gx)\in V_{i\Delta'}$. Observe that $d(V_{i\Delta'},V_{i g\Delta})\geq\frac{1}{3}10^{-h(R)}$ when $\Delta'\neq g\Delta$, but $d(f(gx),gf(x))<\frac{1}{3}10^{-h(R)}$, which forces $\Delta'=g\Delta$. Hence, $gx\in U_{i g\Delta}$ as claimed.
Finally, we show that the action $\Gamma\curvearrowright X$ has dynamic asymptotic dimension function at most $h$. Suppose that $x\in U_{i}$ for some $i$ and $g=g_{n}\cdots g_{1}$ with $g_{k}\in E$ and $g_{k}\cdots g_{1}x\in U_{i}$, $\forall k\in\{1,\cdots,n\}$. If $x\in U_{i\Delta}$ for some $i$-simplex $\Delta$, then $gx\in U_{i g\Delta}$ by using the claim repeatedly, thus $f(gx)\in V_{i g\Delta}\cap f(X)\subseteq C\cap \Prob(S)$. Observe that both $f(gx)$ and $f(x)$ are supported on vertices on $S$, which forces $gS\cap S\neq\varnothing$. Therefore, $g$ belongs to the finite set $F$. According to Definition \ref{def 3.2}, it follows that $\dad_{\Gamma\curvearrowright X}(R)\leq h(R)$. Hence, we conclude that $\dad_{\Gamma\curvearrowright X}\leq h$.
\end{proof}
\section{Dynamic asymptotic dimension growth for groupoids} \label{section: dad}
In this section, we formulate the definition of dynamic asymptotic dimension growth for \'etale groupoids, which generalizes our notion for actions of discrete groups on compact spaces. Moreover, we will examine the relation between our notion with the asymptotic dimension growth. In coarse geometry, we show that the asymptotic dimension growth for a discrete metric space of bounded geometry $X$ is equivalent to the dynamic asymptotic dimension growth for its corresponding coarse groupoid $G(X)$, thus implying that the groupoid $G(X)$ with subexponential dynamic asymptotic dimension growth is amenable. Using this result, we establish the equivalence between the asymptotic dimension growth for a countable discrete group and the dynamic asymptotic dimension growth for its action on some compact Hausdorff space. Consequently, we obtain that a countable discrete group with subexponential asymptotic dimension growth admits a free, amenable, minimal action on the Cantor set which has at most subexponential dynamic asymptotic dimension growth. This allows us to provide many concrete examples for our notion.
Recall from \cite{Ren80} that a groupoid consists of a set $\mathcal{G}$ of morphisms, a subset $\mathcal{G}^{(0)}\subseteq\mathcal{G}$ of obejects (called the unit space), together with source and range maps $s,r:\mathcal{G}\rightarrow\mathcal{G}^{(0)}$, a composition map
$$
\mathcal{G}^{(2)}:=\{(g_{1},g_{2})\in\mathcal{G}\times\mathcal{G}:s(g_{1})=r(g_{2})\}\ni (g_{1},g_{2})\mapsto g_{1}g_{2}\in\mathcal{G},
$$
and an inverse map $g\mapsto g^{-1}$ satisfying a couple of axioms. A locally compact Hausdorrf groupoid is equipped with a locally compact and Hausdorff topology such that structure maps are continuous. It is called \'etale if the range map $r$ (hence the source map $s$) is a local homeomorphism, i.e., for any $g\in\mathcal{G}$, there exists a neighborhood $U$ of $g$ such that $r(U)$ is open and $r|_{U}$ is a homeomorphism. From now on, we assume that groupoids are always locally compact and Hausdorff.
\subsection{The case of \'etale groupoids}
We now focus on \'etale groupoids, from which it is easier to find sufficient conditions for the amenability of groupoids. In this subsection, we introduce the dynamic asymptotic dimension growth for \'etale groupoids. We first define the dynamic asymptotic dimension growth for \'etale groupoids with compact unit spaces by using length functions on these groupoids. Further, borrowing the notion of Alexandrov one-point compactification for groupoids defined in \cite{KKLRU21}, we can define the dynamic asymptotic dimension growth for \'etale groupoids with non-compact unit spaces. Finally, we show that this notion is a generalization of the dynamic asymptotic dimension growth for actions of discrete groups on compact spaces.
The notion of length functions on \'etale groupoids was first appeared in \cite{OY19}, and later generalized by Ma and Wu in their paper \cite{MW20}.
\begin{definition}[\cite{OY19, MW20}]
Let $\mathcal{G}$ be an \'etale groupoid. A length function on $\mathcal{G}$ is a function $\ell:\mathcal{G}\rightarrow [0,\infty)$ satisfying
\begin{enumerate}
\item $\ell(x)=0$ if and only if $x\in\mathcal{G}^{(0)}$;
\item $\ell(x)=\ell(x^{-1})$, for all $x\in\mathcal{G}$;
\item $\ell(xy)\leq\ell(x)+\ell(y)$, if $x$ and $y$ are composable in $\mathcal{G}$.
\end{enumerate}
\end{definition}
The length functions on \'etale groupoids enjoy the following properties.
\begin{definition}[\cite{MW20}]
Let $\mathcal{G}$ be an \'etale groupoid. Let $\ell:\mathcal{G}\rightarrow [0,\infty)$ be a length function on $\mathcal{G}$. For any subset $K\subseteq\mathcal{G}$, we write
$$
\bar{\ell}(K)=\sup\limits_{x\in K}\ell(x).
$$
We say $\ell$ is
\begin{enumerate}
\item proper if for any $K\subseteq\mathcal{G}\backslash\mathcal{G}^{(0)}$, $\bar{\ell}(K)<\infty$ implies $K$ is precompact;
\item controlled if for any $K\subseteq\mathcal{G}$, $K$ is precompact implies $\bar{\ell}(K)<\infty$;
\item coarse if it is both proper and controlled;
\item continuous if it is a continuous function with regard to the topology of $\mathcal{G}$.
\end{enumerate}
\end{definition}
\begin{remark}
When the unit space $\mathcal{G}^{(0)}$ is compact, the item (1) can be rewritten as:
$\ell$ is proper if for any $K\subseteq\mathcal{G}$, $\bar{\ell}(K)<\infty$ implies $K$ is precompact. It is straightforward to see that a continuous length function is controlled, hence the item (4) implies the item (2). Recall from \cite{MW20} that two length functions $\ell_{1}, \ell_{2}$ are said to be coarsely equivalent if for any $r>0$, we have
$$
\sup\{\ell_{1}(x): \ell_{2}(x)\leq r\}<\infty \text{ and }\sup\{\ell_{2}(x): \ell_{1}(x)\leq r\}<\infty.
$$
\end{remark}
To obtain the amenability of groupoids, we are mainly concerned about coarse continuous length functions. The existence of coarse continuous length functions was examined by Ma and Wu in \cite{MW20} in the following case.
\begin{theorem}[\cite{MW20}]
Up to coarse equivalence, any $\sigma$-compact locally compact Hausdorff \'etale groupoid has a unique coarse continuous length function.
\end{theorem}
The following transformation groupoid is a typical example of the above theorem, which plays an important role in the proof of Lemma \ref{lemma 4.12}.
\begin{example}[\cite{MW20}]\label{ex 4.5}
Let $\Gamma$ be a countable discrete group, and let $X$ be a $\sigma$-compact, locally compact Hausdorff space. The action $\Gamma\curvearrowright X$ by homeomorphisms induces an associated transformation groupoid $X\rtimes\Gamma$ as follows:
\begin{itemize}
\item $X\rtimes\Gamma:=\{(\gamma x,\gamma,x)\mid x\in X,\gamma\in\Gamma\}$;
\item the unit space ${(X\rtimes\Gamma)}^{(0)}$ is $\{(x,e,x)\mid x\in X\}$, which is identified with $X$;
\item source and range maps: $s(\gamma x,\gamma,x)=x$, $r(\gamma x,\gamma,x)=\gamma x$;
\item composition: $(\gamma\gamma'x,\gamma,\gamma'x)(\gamma'x,\gamma',x)=(\gamma\gamma'x,\gamma\gamma',x)$;
\item inverse: $(\gamma x,\gamma,x)^{-1}=(x,\gamma^{-1},\gamma x)$.
\end{itemize}
The topology coming from $X\times\Gamma$, then $X\rtimes\Gamma$ is a $\sigma$-compact, locally compact \'etale groupoid. Let $\ell_{\Gamma}:\Gamma\rightarrow[0,\infty)$ be a proper length function on $\Gamma$, and let $g: X\rightarrow [0,\infty)$ be a continuous proper function on $X$. We define
$$
\ell_{X\rtimes\Gamma}:X\rtimes\Gamma\rightarrow [0,\infty),
$$
$$
\quad\qquad\qquad\qquad\qquad\qquad (\gamma x,\gamma,x)\mapsto \ell_{\Gamma}(\gamma)(1+\max\{g(\gamma x),g(x)\}).
$$
Clearly, it is a coarse continuous length function on $X\rtimes\Gamma$. If $X$ is compact, we may simply choose $g=0$, and thus
$$
\ell_{X\rtimes\Gamma}(\gamma x,\gamma,x)=\ell_{\Gamma}(\gamma), \text{ for any }(\gamma x,\gamma,x)\in X\rtimes\Gamma.
$$
\end{example}
By using coarse continuous length functions on \'etale groupoids with compact unit spaces, we can define the dynamic asymptotic dimension growth for these groupoids.
\begin{definition}\label{def 4.6}
Let $\mathcal{G}$ be an \'etale groupoid with compact unit space $\mathcal{G}^{(0)}$, and let $\ell:\mathcal{G}\rightarrow [0,\infty)$ be a coarse continuous length function. For any $R>0$, let $\dad_{\mathcal{G}}(R)$ be the smallest number over all $m\in\mathbb{N}$ with the following property: for the open precompact subset $K={\{z\in\mathcal{G}\mid\ell(z)< R\}}$ of $\mathcal{G}$, there exists an open cover $\{U_{0},\cdots, U_{m}\}$ of $\mathcal{G}^{(0)}$ such that for each $i$, the set $\{g\in K\mid s(g), r(g)\in U_{i}\}$ is contained in a precompact subgroupoid of $\mathcal{G}$. We call $\dad_{\mathcal{G}}:\mathbb{R}^{+}\rightarrow\mathbb{N}$ the dynamic asymptotic dimension function of $\mathcal{G}$, and call the growth type of the function $\dad_{\mathcal{G}}$ the dynamic asymptotic dimension growth of $\mathcal{G}$. If no such function $\dad_{\mathcal{G}}$ exists, we say that $\mathcal{G}$ has infinite dynamic asymptotic dimension.
\end{definition}
If $\mathcal{G}$ is an \'etale groupoid with non-compact unit space, the following definition provides a suitable way to reduce the unit space to the compact case.
\begin{definition}[\cite{KKLRU21}]
Let $\mathcal{G}$ be an \'etale groupoid with locally compact Hausdorff space of units. Then the Alexandrov groupoid $\mathcal{G}^{+}$ is the set $\mathcal{G}\cup(\mathcal{G}^{(0)})^{+}$ with the topology determined by specifying that the inclusions $\mathcal{G}$, $(\mathcal{G}^{(0)})^{+}\subseteq\mathcal{G}^{+}$ are open, and the groupoid structure extending the groupoid structure of $\mathcal{G}$ and making $\infty$ a unit.
\end{definition}
Observe that $\mathcal{G}^{+}$ is an \'etale groupoid whose unit space is the compact Hausdorff space $(\mathcal{G}^{(0)})^{+}$ by taking Alexandrov one-point compactification of $\mathcal{G}^{(0)}$. The next proposition demonstrated that the original groupoid $\mathcal{G}$ shares the same dynamic asymptotic dimension with its Alexandrov groupoid $\mathcal{G}^{+}$.
\begin{proposition}[\cite{CDGHV24}]
Let $\mathcal{G}$ be a locally compact, Hausdorff and \'etale groupoid. Then $\dad(\mathcal{G})=\dad(\mathcal{G}^{+})$.
\end{proposition}
For groupoids with non-compact unit spaces, we apply the above proposition to Definition \ref{def 4.6}, and let $\dad_{\mathcal{G}}(R):=\dad_{\mathcal{G}^{+}}(R)$, for any $R>0$. In this way, we can define the dynamic asymptotic dimension growth for groupoids with non-compact unit spaces.
\begin{definition}\label{def 4.9}
Let $\mathcal{G}$ be an \'etale groupoid with non-compact unit space $\mathcal{G}^{(0)}$, and let $\mathcal{G}^{+}$ be the Alexandrov groupoid of $\mathcal{G}$. We define the dynamic asymptotic dimension function of $\mathcal{G}$ as $\dad_\mathcal{G}:=\dad_{\mathcal{G}^{+}}$, and call the growth type of the function $\dad_{\mathcal{G}}$ the dynamic asymptotic dimension growth of $\mathcal{G}$.
\end{definition}
%We will reduce the case of noncompact unit space to the compact case, thus we only focus on the compact case.
%\begin{definition}
%Let $\mathcal{G}$ be an \'etale groupoid, and let $\ell:\mathcal{G}\rightarrow [0,\infty)$ be a proper continuous length function. For any $R>0$, let $dad_{\mathcal{G}}(R)$ be the smallest number over all $m\in\mathbb{N}$ with the following property: for every open precompact subset $K={\{z\in\mathcal{G}\mid\ell(z)< R\}}$ of $\mathcal{G}$, there exist open subsets $U_{0},\cdots, U_{m}$ of $\mathcal{G}^{(0)}$ that cover $s(K)\cup r(K)$ such that for each $i$, the set $\{g\in K\mid s(g), r(g)\in U_{i}\}$ is contained in a precompact subgroupoid of $\mathcal{G}$. We call $dad_{\mathcal{G}}:\mathbb{R}^{+}\rightarrow\mathbb{N}$ the dynamic asymptotic dimension growth of $\mathcal{G}$.
%\end{definition}
\begin{remark}
Clearly, the function $\dad_{\mathcal{G}}$ is non-decreasing, and the dynamic asymptotic dimension of $\mathcal{G}$ is defined as follows:
\begin{equation}\label{eq 3}
\dasdim(\mathcal{G})=\lim\limits_{R\rightarrow\infty}\dad_{\mathcal{G}}(R).
\end{equation}
\begin{enumerate}
\item $\dasdim(\mathcal{G})\leq d$ for some finite $d$ if and only if $\dad_{\mathcal{G}}$ is a constant function and $\dad_{\mathcal{G}}\leq d$.
\item If $\dasdim(\mathcal{G})=\infty$, then both sides of equation (\ref{eq 3}) are infinite.
\item For each $i$, the subgroupoid $\mathcal{G}_{i}=\langle K\cap\mathcal{G}|_{U_{i}}\rangle$ generated by the set $$\{g\in K\mid s(g), r(g)\in U_{i}\}$$ is open precompact by the next lemma.
\end{enumerate}
\end{remark}
The following lemma will be used in the proof of Lemma \ref{lemma 5.6}.
\begin{lemma}[\cite{GWY17}]\label{lemma 4.11}
Let $\mathcal{G}$ be an \'etale groupoid.
\begin{enumerate}
\item If $K, H$ are open (resp. compact, precompact) subsets of $\mathcal{G}$, then
$$K\cdot H:=\{kh\in\mathcal{G}\mid k\in K, h\in H, s(k)=r(h)\}$$
is open (resp. compact, precompact).
\item If $K\subseteq\mathcal{G}$ is open, then the subgroupoid of $\mathcal{G}$ generated by $K$ is also open, and is itself an \'etale groupoid.
\end{enumerate}
\end{lemma}
Using the length function on the transformation groupoid constructed in Example \ref{ex 4.5}, we prove that the group action $\Gamma\curvearrowright X$ shares the same dynamic asymptotic dimension function with its corresponding transformation groupoid $X\rtimes\Gamma$. This means that Definition \ref{def 4.6} generalizes the dynamic asymptotic dimension growth for group actions in the sense of Definition \ref{def 3.2}.
\begin{lemma}\label{lemma 4.12}
Let $X$ be a compact Hausdorff space, and let $f:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function. An action $\Gamma\curvearrowright X$ has dynamic asymptotic dimension function $f$ if and only if the corresponding transformation groupoid $\mathcal{G}=X\rtimes\Gamma$ has dynamic asymptotic dimension function $f$.
\end{lemma}
%\begin{lemma}\label{lemma 4.8}
%Let $X$ be a compact Hausdorff space, and let $f:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function. An action $\Gamma\curvearrowright X$ has dynamic asymptotic dimension function $f$ if and only if the corresponding transformation groupoid $\mathcal{G}=X\rtimes\Gamma$ has dynamic asymptotic dimesnion function $f$.
%Then the dynamic asymptotic dimension function of an action $\Gamma\curvearrowright X$ is the same as that of the corresponding transformation groupoid $\mathcal{G}=X\rtimes\Gamma$.
%\end{lemma}
\begin{proof}
We may assume that an action $\Gamma\curvearrowright X$ has dynamic asymptotic dimension function $\dad_{\Gamma\curvearrowright X}$, and its associated transformation groupoid has dynamic asymptotic dimension function $\dad_{\mathcal{G}}$. It thus suffices to show that $\dad_{\mathcal{G}}=\dad_{\Gamma\curvearrowright X}$. First, we prove that $\dad_{\mathcal{G}}\leq \dad_{\Gamma\curvearrowright X}$. For any $R>0$, for the open precompact subset $K={\{(\gamma x,\gamma,x)\in \mathcal{G}\mid\ell(\gamma x,\gamma,x)< R\}}$ of $\mathcal{G}$. Then $\ell_{\Gamma}(\gamma)=\ell(\gamma x,\gamma,x)<R$, thus $K$ has the form
$$
\{(\gamma x,\gamma,x)\in \mathcal{G}\mid x\in X, \gamma\in E\}
$$
for a finite subset $E=\{\gamma\in\Gamma\mid\ell_{\Gamma}(\gamma)<R\}$ of $\Gamma$. Since the action has dynamic asymptotic dimension function $\dad_{\Gamma\curvearrowright X}$, there is an open cover $U_{0},\cdots,U_{\dad_{\Gamma\curvearrowright X}(R)}$ of $\mathcal{G}^{(0)}=X$ such that the set
$$
F_{i}:=\begin{Bmatrix}&\Big|& \text{ there exist }x\in U_{i}\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}x\in U_{i}\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}
$$
is finite for each $i\in\{0,\cdots,\dad_{\Gamma\curvearrowright X}(R)\}$.
Therefore, the subgroupoid $\mathcal{G}_{i}$ generated by the set
$$
\{(\gamma x,\gamma,x)\in K\mid x,\gamma x\in U_{i}\}
$$
is contained in $\{(\gamma x,\gamma,x)\in\mathcal{G}\mid x\in X, \gamma\in F_{i}\}.$
Since $F_{i}$ is finite and $X$ is compact, the set $\{(\gamma x,\gamma,x)\in\mathcal{G}\mid x\in X, \gamma\in F_{i}\}$ is precompact, and thus $\mathcal{G}_{i}$ is precompact. Hence, $\dad_{\mathcal{G}}(R)\leq \dad_{\Gamma\curvearrowright X}(R)$, which implies that $\dad_{\mathcal{G}}\leq \dad_{\Gamma\curvearrowright X}$.
On the other hand, we show that $\dad_{\Gamma\curvearrowright X}\leq\dad_{\mathcal{G}}$. For any $R>0$, for the finite subset $E=\{\gamma\in\Gamma\mid\ell_{\Gamma}(\gamma)<R\}$ of $\Gamma$, there is an open precompact subset $K=\{(\gamma x,\gamma,x)\in\mathcal{G}\mid x\in X, \gamma\in E\}$ of $\mathcal{G}$. Since $\mathcal{G}$ has the dynamic asymptotic dimension function $\dad_{\mathcal{G}}$, there is an open cover $\{U_{0},\cdots, U_{\dad_{\mathcal{G}}(R)}\}$ of $X$ such that for each $i\in\{0,\cdots,\dad_{\mathcal{G}}(R)\}$, the set
$$\{(\gamma x,\gamma,x)\in K\mid x,\gamma x\in U_{i}\}$$
is contained in a precompact subgroupoid of $\mathcal{G}$. Then the subgroupoid $\mathcal{G}_{i}$ generated by the set
$$
\{(\gamma x,\gamma,x)\in\mathcal{G}\mid\gamma\in E, x, \gamma x\in U_{i}\},
$$
is precompact. For each $i$, let
$$
F_{i}:=\begin{Bmatrix}&\Big|& \text{ there exist }x\in U_{i}\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}x\in U_{i}\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}.
$$
Therefore, the subset $$
\{(\gamma x,\gamma,x)\in\mathcal{G}\mid\gamma\in F_{i}, x\in U_{i}\}
$$
is contained in $\mathcal{G}_{i}$, and is thus precompact, which implies that $F_{i}$ is finite. Clearly, $\dad_{\Gamma\curvearrowright X}(R)\leq \dad_{\mathcal{G}}(R)$, which forces $\dad_{\Gamma\curvearrowright X}\leq\dad_{\mathcal{G}}$. Hence, $\dad_{\Gamma\curvearrowright X}=\dad_{\mathcal{G}}$, namely, $\dad_{\Gamma\curvearrowright X}=f$ if and only if $\dad_{\mathcal{G}}=f$.
\end{proof}
\begin{remark}
Using the above notation, let $\mathcal{G}^{+}$ be the Alexandrov groupoid of $\mathcal{G}$, and let $X_{+}$ be the one-point compactification of $X$. It is easy to see that the action of $\Gamma$ on $X_{+}$ given by
$$
\gamma\cdot x=\begin{cases}\gamma\cdot x& \text{ if }x\in X\\
\infty&\text{ if }x=\infty
\end{cases}
$$
is continuous. Then we have
$$
\mathcal{G}^{+}=\{(\gamma,x):\gamma\in\Gamma, x\in X\}\cup\{(e,\infty)\}\text{ and }X_{+}\rtimes\Gamma=\{(\gamma,x): \gamma\in\Gamma, x\in X_{+}\}.
$$
It is easy to verify that the Alexandrov groupoid $\mathcal{G}^{+}$ is not identical to the transformation groupoid $X_{+}\rtimes\Gamma$; it is merely its wide and open subgroupoid [\cite{CDGHV24}, Example 3.5]. Therefore, using the Alexandrov construction, the above lemma does not extend to the non-compact Hausdorff space $X$.
\end{remark}
\subsection{Example: coarse spaces with asymptotic dimension growth}
In this subsection, we study a geometric class of groupoids. Given a discrete metric space of bounded geometry $X$, Skandalis, Tu, and Yu introduced an \'etale groupoid $G(X)$ in \cite{STY02}, called the coarse groupoid since it encodes many coarse geometric properties of the underlying space $X$. Moreover, we show that the asymptotic dimension growth for $X$ is equivalent to the dynamic asymptotic dimension growth for its associated coarse groupoid $G(X)$. This example served as our primary motivation and also inspired the terminology `dynamic asymptotic dimension growth'.
\begin{definition}[\cite{STY02}]
Let $(X,d)$ be a discrete metric space of bounded geometry. The coarse groupoid $G(X)$ associated to $(X,d)$ is defined as follows:
\begin{enumerate}
\item for any $R>0$, we define $E_{R}:=\{(x,y)\in X\times X\mid d(x,y)\leq R\}$;
\item as a topological space, we have $G(X):=\bigcup\limits_{R\geq 0}\overline{E_{R}}$ in $\beta(X\times X)$, the Stone-\v{C}ech compactification of $X\times X$;
\item equip $G(X)$ with the weak topology coming from this union, this is a locally compact topology for which each $\overline{E_{R}}$ is a compact open subset of $G(X)$;
\item $G(X)^{(0)}=\overline{E_{0}}\cong\beta X$;
\item $r(x,y)=x,s(x,y)=y$, for all $(x,y)\in G(X)$;
\item $E_{R}\circ E_{S}\subseteq E_{R+S}$, for all $R, S\geq 0$.
\end{enumerate}
\end{definition}
\begin{remark}
The coarse groupoid $G(X)$ is $\sigma$-compact, locally compact Hausdorff, principal and \'etale \cite{STY02}. Here a discrete metric space of bounded geometry is also called a uniformly locally finite metric space. Recall from \cite{MW20} that there is a canonical length function $\ell$ on $G(X)$ defined by extending the metric $d: X\times X\rightarrow [0,\infty]$ to $\beta(X\times X)$ and observing that it takes finite values on $G(X)$. The canonical length function $\ell$ on $G(X)$ has the following properties:
\begin{enumerate}
\item $\ell$ is continuous since it comes from the metric $d$.
\item $\ell$ is coarse. Since $X\times X$ is dense in $\beta(X\times X)$, for any $R>0$, the open set $\ell^{-1}([0,R))$ is contained in $\overline{\ell^{-1}([0,R))\cap X\times X}$, which is contained in the compact set $\overline{E_{R}}$. Hence, $\ell^{-1}([0,R))$ is precompact, which yields that $\ell$ is proper. Since $\ell$ is continuous, it is controlled.
%Observe that for any precompact subset $K$ of $G(X)$, $K$ must contained in a controlled set $E_{R}$ for some $R$ by the definition of topology of $G(X)$. Thus $\ell$ is controlled.
\item $\ell^{-1}([0,R])=\overline{E_{R}}$ and $\ell^{-1}(\{0\})=\overline{E}_{0}\cong\beta X$.
\end{enumerate}
\end{remark}
The following theorem bridges the dynamic asymptotic dimension growth for the coarse groupoid $G(X)$ with the asymptotic dimension growth for the underlying space $X$, which generalizes the finite case of [\cite{GWY17}, Theorem 6.4].
\begin{theorem}\label{th 4.16}
Let $X$ be a discrete metric space of bounded geometry, and let $G(X)$ be the associated coarse groupoid. Let $f: \mathbb{R}^{+}\rightarrow\mathbb{N}$ be a non-decreasing function. Then $\textnormal{ad}_{X}\approx f$ if and only if $\dad_{G(X)}\approx f$.
\end{theorem}
\begin{proof}
Let us assume that $G(X)$ has dynamic asymptotic dimension function $\dad_{G(X)}$, and $X$ has asymptotic dimension function $\widecheck{\ad}_{X}$ as defined in Lemma \ref{lemma 2.13}. Below, we show that $\widecheck{\ad}_{X}\approx\dad_{G(X)}$. First, we need to verify that $\widecheck{\ad}_{X}\preceq\dad_{G(X)}$. For any $R>0$, for the open precompact subset $K=\{z\in G(X)\mid\ell(z)< R\}$ of $G(X)$, there exists a controlled set $E_{R-\frac{1}{n}}$ for some positive integer $n$ such that $\overline{E_{R-\frac{1}{n}}}\subseteq K$. Observe that $\overline{E_{R-\frac{1}{n}}}$ is a compact open subset of $G(X)$ containing the unit space $G(X)^{(0)}=\beta X$. Let $U_{0},\cdots,U_{\dad_{G(X)}(R)}$ be an open cover of $G(X)^{(0)}=\beta X$ such that for each $i\in\{0,\cdots,\dad_{G(X)}(R)\}$, the set
$$
\{g\in K\mid s(g),r(g)\in U_{i}\}
$$
is contained in a precompact subgroupoid of $G(X)$. Therefore, the subgroupoid $\mathcal{G}_{i}$ generated by the set
$$
\{g\in \overline{E_{R-\frac{1}{n}}}\mid s(g),r(g)\in U_{i}\}
$$
is precompact. By the definition of the topology on $G(X)$, there exists a controlled set $F\subseteq X\times X$ such that each $\mathcal{G}_{i}$ is contained in $\overline{F}\subseteq G(X)$. Fix $i$, and let $\sim$ denote the equivalence relation on $U_{i}$ induced by $\mathcal{G}_{i}$, so $x\sim y$ if there exists $g\in \mathcal{G}_{i}$ such that $s(g)=x, r(g)=y$. We denote by $\{U^{j}_{i}\mid j\in J_{i}\}$ the equivalence classes for this relation. Define
$$
\mathcal{U}_{i}:=\{U^{j}_{i}\cap X\mid j\in J_{i}\}.
$$
Let $\mathcal{U}=\mathcal{U}_{0}\sqcup\cdots\sqcup\mathcal{U}_{\textnormal{dad}_{G(X)}(R)}$, and it is a cover of $X$ since $\{U_{0},\cdots, U_{\textnormal{dad}_{G(X)}(R)}\}$ is a cover of $\beta X$.
We next claim that the cover $\mathcal{U}$ is $F$-bounded and each $\mathcal{U}_{i}$ is $E_{R-\frac{1}{n}}$-separated. Indeed, the equivalence relation on each $U^{j}_{i}\cap X$ induced by $\mathcal{G}_{i}$ is entirely contained in $F$ as $\mathcal{G}_{i}\subseteq\overline{F}$, so $\mathcal{U}$ is $F$-bounded. For each $\mathcal{U}_{i}$, if $(U^{j}_{i}\times U^{k}_{i})\cap E_{R-\frac{1}{n}}\neq\varnothing$ for $j\neq k$, then there exists $(x,y)\in (U^{j}_{i}\times U^{k}_{i})\cap E_{R-\frac{1}{n}}$. Observe that $s(x,y)=y, r(x,y)=x$, so $(x,y)\in \mathcal{G}_{i}$ and $x\sim y$. However, when $j\neq k$, $U^{j}_{i}\cap U^{k}_{i}=\varnothing$, contradiction. Thus, for $j\neq k$, $(U^{j}_{i}\times U^{k}_{i})\cap E_{R}=\varnothing$, i.e. $\mathcal{U}_{i}$ is $E_{R-\frac{1}{n}}$-separated. Hence, $\widecheck{\ad}_{X}(R-\frac{1}{n})\leq \textnormal{dad}_{G(X)}(R)$, which implies that $\widecheck{\ad}_{X}\preceq\textnormal{dad}_{G(X)}$.
The next step is to show that $\textnormal{dad}_{G(X)}\preceq \widecheck{\ad}_{X}$. For any $R>0$, for the open precompact subset $K=\{z\in G(X)\mid\ell(z)< R\}$ of $G(X)$. The definition of the topology on $G(X)$ implies that $K\subseteq\overline{E_{R}}$ for the controlled set $E_{R}$. Since $X$ has asymptotic dimension function $\widecheck{\ad}_{X}$, there exist a controlled set $F$ and a cover $\mathcal{U}$ of $X$ such that $\mathcal{U}$ is $F$-bounded and such that
$$
\mathcal{U}=\mathcal{U}_{0}\sqcup\cdots\sqcup\mathcal{U}_{\widecheck{\ad}_{X}(R)}
$$
and each $\mathcal{U}_{i}$ is $E_{R}$-separated.
For each $i$, set
$$
U_{i}:=\overline{\bigsqcup\limits_{U\in\mathcal{U}_{i}}U},
$$
which is a compact open subset of $\beta X$. Since $\mathcal{U}$ is a cover of $X$, $\{U_{0},\cdots,U_{\widecheck{\ad}_{X}(R)}\}$ is a cover of $\beta X$. Furthermore, for each $i$,
$$
\mathcal{G}_{i}:=\bigsqcup\limits_{U\in\mathcal{U}_{i}}U\times U
$$
is a subgroupoid of the pair groupoid that is contained in $F$, and by continuity of the groupoid operations, the subgroupoid of $G(X)$ generated by
$$
\{g\in E_{R}\mid r(g), s(g)\in U_{i}\}
$$
is contained in $\overline{\mathcal{G}_{i}}$, a compact subgroupoid of $G(X)$ contained in the compact subset $\overline{F}$. Thus the subset $\{g\in K\mid s(g), r(g)\in U_{i}\}$ is contained in a compact open subgroupoid of $G(X)$. Therefore, $\textnormal{dad}_{G(X)}(R)\leq \widecheck{\ad}_{X}(R)$, which yields that $\textnormal{dad}_{G(X)}\preceq \widecheck{\ad}_{X}$. Hence, we obtain $\textnormal{dad}_{G(X)}\approx \widecheck{\ad}_{X}$ as desired.
Finally, by Lemma \ref{lemma 2.13}, we see that $\ad_{X}\approx\widecheck{\ad}_{X}$, which implies that $\ad_{X}\approx\dad_{G(X)}$. Therefore, $\ad_{X}\approx f$ if and only if $\textnormal{dad}_{G(X)}\approx f$.
\end{proof}
\begin{remark}
Based on the second step of the above proof, we see that if $\ad_{X}\preceq f$, then $\dad_{G(X)}\preceq f$ in the following slightly stronger form: For any $R>0$, for the open precompact subset $K=\{z\in G(X)\mid\ell(z)<R\}$ of $G(X)$, there exists an open cover $\{U_{0},\cdots, U_{\widecheck{\ad}_{X}(R)}\}$ of $G(X)^{(0)}$ such that the subset $\{g\in K\mid s(g), r(g)\in U_{i}\}$ is contained in a compact open subgroupoid of $G(X)$.
\end{remark}
By a result of Ozawa and Oppenheim, the subexponential asymptotic dimension growth implies property A.
\begin{theorem}\cite{Oza12, Opp14}\label{th 4.18}
Let $X$ be a discrete metric space of bounded geometry. If $X$ has subexponential asymptotic dimension growth, then $X$ has property A.
\end{theorem}
Since property A for $X$ is equivalent to the amenability of the corresponding coarse groupoid $G(X)$ [\cite{STY02}, Theorem 5.3]. Combining Theorems \ref{th 4.16} and \ref{th 4.18}, we have the following result:
\begin{corollary}\label{cor 4.19}
Let $X$ be a discrete metric space of bounded geometry. If the associated coarse groupoid $G(X)$ has subexponential dynamic asymptotic dimension growth, then $G(X)$ is amenable.
\end{corollary}
The following example follows from Example \ref{example 2.7}, Theorem \ref{th 4.16}, and Corollary \ref{cor 4.19}.
\begin{example}
Let $X$ be a geodesic uniformly locally finite coarse median space with finite rank and at most exponential volume growth. Then its corresponding coarse groupoid $G(X)$ has subexponential dynamic asymptotic dimension growth, and it is amenable.
\end{example}
\subsection{Equivalence with asymptotic dimension growth}
In this subsection, we will establish the equivalence between the asymptotic dimension growth for countable discrete groups and the dynamic asymptotic dimension growth for their group actions. As a result, we demonstrate that countable discrete groups with subexponential asymptotic dimension growth all admit free, amenable, minimal actions on the Cantor set which have at most subexponential dynamic asymptotic dimension growth.
Since property A for a countable discrete group $\Gamma$ is equivalent to the (topological) amenability of the natural action of $\Gamma$ on its Stone-\v{C}ech compactification $\beta\Gamma$ \cite{HR00, Wil09}. A combination of Theorems \ref{th 3.15}, \ref{th 4.16}, \ref{th 4.18} and Lemma \ref{lemma 4.12} yields the following equivalences.
\begin{theorem}\label{th 4.21}
Let $\Gamma$ be a countable discrete group, and let $h:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function with subexponential growth. Then the following are equivalent:
\begin{enumerate}
\item $\Gamma$ admits an action $\Gamma\curvearrowright X$ on a compact space $X$ that is $\ad_{X}$-BLR with respect to the family of finite subgroups and $\ad_{X}\preceq h$;
\item $\Gamma$ admits an action $\Gamma\curvearrowright X$ on a compact space $X$ with dynamic asymptotic dimension function $\dad_{\Gamma\curvearrowright X}\preceq h$.
\item The canonical action of $\Gamma$ on $\beta\Gamma$ has dynamic asymptotic dimension function $\dad_{\Gamma\curvearrowright\beta\Gamma}\preceq h$.
\item $\Gamma$ equipped with the canonical left-invariant coarse structure for which the controlled sets are
$$
\mathcal{E}=\{E\subseteq\Gamma\times\Gamma\mid\{s^{-1}t\in\Gamma\mid(s,t)\in E\}\text{ is finite}\},
$$
has asymptotic dimension function $\ad_{X}\preceq h$.
\item The canonical action of $\Gamma$ on $\beta\Gamma$ is $\ad_{X}$-BLR with respect to the family of finite subgroups and $\ad_{X}\preceq h$.
\end{enumerate}
\end{theorem}
\begin{proof}
The fact that $(1)$ implies $(2)$ follows from Theorem \ref{th 3.15}.
$(2)\Rightarrow (3)$: assume that $\Gamma\curvearrowright X$ is an action on a compact space $X$ with dynamic asymptotic dimension function $\dad_{\Gamma\curvearrowright X}$ and $\dad_{\Gamma\curvearrowright X}\preceq h$. For any $x\in X$, let $f: \Gamma\rightarrow X, \gamma\mapsto\gamma\cdot x$ be any orbit map, and it is $\Gamma$-equivariant. Since $\Gamma$ acts by homeomorphisms on $X$, $f$ is continuous. By universal property of $\beta\Gamma$, $f$ extends uniquely to a continuous map $\phi:\beta\Gamma\rightarrow X$ such that $\phi\circ i=f$, where the inclusion map $i:\Gamma\hookrightarrow\beta\Gamma$ is continuous. Moreover, $\phi$ is $\Gamma$-equivariant as $f$ is $\Gamma$-equivariant and $\Gamma$ is dense in $\beta\Gamma$. By definition, for any $R>0$, for the finite subset $E=\{\gamma\in\Gamma\mid\ell(\gamma)<R\}$ of $\Gamma$, there exists an open cover $\{U_{0},\cdots, U_{\dad_{\Gamma\curvearrowright X}(R)}\}$ of $X$ such that for each $i\in\{0,\cdots,\dad_{\Gamma\curvearrowright X}(R)\}$, the set
\begin{align}\label{eq 4.1}
\begin{Bmatrix}&\Big|& \text{ there exist }x\in U_{i}\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}x\in U_{i}\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}
\end{align}
is finite. Note that $\phi(\beta\Gamma)$ is a compact subset of $X$ as $\phi$ is continuous and $\beta\Gamma$ is compact, thus $\{U_{0},\cdots, U_{\textnormal{dad}_{\Gamma\curvearrowright X}(R)}\}$ is also an open cover of $\phi(\beta\Gamma)$. Moreover, $\{\phi^{-1}(U_{0}),\cdots,\phi^{-1}(U_{\textnormal{dad}_{\Gamma\curvearrowright X}(R)})\}$ is an open cover of $\beta\Gamma$ since $\beta\Gamma\subseteq\phi^{-1}(\phi(\beta\Gamma))$ and $\phi$ is continuous. Furthermore,
if $y\in\phi^{-1}(U_{i})$ for some $i$ and $\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}$ with $\gamma_{k}\in E$ and $\gamma_{k}\cdots\gamma_{1}y\in \phi^{-1}(U_{i})$ for each $k\in\{1,\cdots,n\}$, then $\phi(y)\in U_{i}$ and $\gamma_{k}\cdots\gamma_{1}\phi(y)=\phi(\gamma_{k}\cdots\gamma_{1}y)\in U_{i}$ as $\phi$ is $\Gamma$-equivariant.
Hence, for each $i\in\{0,\cdots,\textnormal{dad}_{\Gamma\curvearrowright X}(R)\}$, the set
\begin{align}\label{eq 4.2}
\begin{Bmatrix}&\Big|& \text{ there exist }y\in \phi^{-1}(U_{i})\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}y\in\phi^{-1}(U_{i})\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}
\end{align}
is contained in the set in line (\ref{eq 4.1}), and is thus finite. This shows that $\textnormal{dad}_{\Gamma\curvearrowright\beta\Gamma}(R)\leq\textnormal{dad}_{\Gamma\curvearrowright X}(R)$, and thus
$\textnormal{dad}_{\Gamma\curvearrowright\beta\Gamma}\leq\textnormal{dad}_{\Gamma\curvearrowright X}\preceq h$.
$(3)\Rightarrow (4)$: the transformation groupoid $\beta\Gamma\rtimes\Gamma$ has dynamic asymptotic dimension function $\textnormal{dad}_{\beta\Gamma\rtimes\Gamma}\preceq h$ by Lemma \ref{lemma 4.12}. On the other hand, $\beta\Gamma\rtimes\Gamma=G(\Gamma)$ by [\cite{STY02}, Proposition 3.4], thus $G(\Gamma)$ has dynamic asymptotic dimension function $\textnormal{dad}_{G(X)}\preceq h$. By Theorem \ref{th 4.16}, $\Gamma$ has asymptotic dimension function $\textnormal{ad}_{X}\preceq h$.
$(4)\Rightarrow (5)$: Combining Theorem \ref{th 4.18} with the assumption that $\Gamma$ has at most subexponential asymptotic dimension growth, we have that $\Gamma$ has property A. Then the canonical action of $\Gamma$ on $\beta\Gamma$ is amenable by [\cite{HR00}, Theorem 3.3]. Further, given $R>0$, the finite subset $E=\{\gamma\in\Gamma\mid\ell(\gamma)<R\}$ and $\varepsilon>0$, there exists a continuous $f:\beta\Gamma\rightarrow \Prob_{\ad_{X}(R)}(\Gamma)$ such that
$$
\sup\limits_{\gamma'\in\beta\Gamma}\Vert f(\gamma\gamma')-\gamma f(\gamma')\Vert_{1}<\varepsilon, \forall \gamma\in E,
$$
which implies that $f$ is a $(E, \varepsilon)$-equivariant map.
Therefore, the action of $\Gamma$ on $\beta\Gamma$ is $\ad_{X}$-BLR with respect to the family of finite subgroups and $\ad_{X}\preceq h$.
$(5)\Rightarrow (1)$: trivial.
\end{proof}
Combining Theorem \ref{th 4.21} with the results of R\o rdam and Sierakowski in [\cite{RS12}, Section 6], we obtain the following theorem.
\begin{theorem}\label{th 4.22}
Let $\Gamma$ be a countable infinite discrete group with subexponential asymptotic dimension growth $h$. Then $\Gamma$ admits a free, minimal action on the Cantor set which has dynamic asymptotic dimension growth at most $h$.
\end{theorem}
\begin{proof}
\textbf{Step 1}. In the first step, we will construct countable sets $\{V_{n}\}_{n\in\mathbb{N}}$.
From the fact that the statement $(4)$ implies the statement $(5)$ of Theorem \ref{th 4.21}, we see that the action of $\Gamma$ on $\beta\Gamma$ is amenable and is $\textnormal{ad}_{X}$-BLR for the family $\mathcal{F}$ of finite subgroups of $\Gamma$ with $\textnormal{ad}_{X}\approx h$. For each $n\in\mathbb{N}$, let $E_{n}=\{\gamma\in\Gamma\mid\ell(\gamma)<n\}$, which is a finite subset of $\Gamma$. By the properties of the length function $\ell$, there is a nested sequence
$$
E_{1}\subseteq E_{2}\subseteq E_{3}\subseteq\cdots
$$
such that $\Gamma=\bigcup\limits_{n}E_{n}$. Then for each $n\in\mathbb{N}$, there exist a $(\Gamma,\mathcal{F})$-complex $C_{n}$ of dimension at most $\textnormal{ad}_{X}(n)$ and a continuous $(E_{n},\frac{1}{n})$-equivariant map $f_{n}:\beta\Gamma\rightarrow C_{n}$. Since $\beta\Gamma$ is compact, we assume by Lemma \ref{lemma 3.10} that $f_{n}(\beta\Gamma)\subseteq\Prob(S_{n})\cap C_{n}$, where $S_{n}$ is a finite subset of $C_{n}$ with cardinality $\vert S_{n}\vert\leq \textnormal{ad}_{X}(n)+1$. Further, we replace $C_{n}$ by $C_{n}\cap \Prob(\Gamma\cdot S_{n})$, we may assume that there are only finitely many $\Gamma$-orbits of $S_{n}$, which implies that the vertex set $V_{n}$ of $C_{n}$ is countable.
\textbf{Step 2}. In the next step, we show that $\Gamma$ admits a free, $\textnormal{ad}_{X}$-BLR action on a totally disconnected, metrizable compact space $\hat{A}$. Let $A$ be a unital $\Gamma$-invariant $C^{*}$-subalgebra of $\ell^{\infty}(\Gamma)$ , and let $\hat{A}$ denote the spectrum of $A$. By Gelfand duality, $\hat{A}$ identifies with a quotient space of $\beta\Gamma$ such that the quotient map $q: \beta\Gamma\rightarrow\hat{A}$ is equivariant.
For each $n$, write
$$
f_{n}(x)=\sum\limits_{v\in V_{n}}t_{v}(x)v,
$$
where each $t_{v}:\beta\Gamma\rightarrow [0,1]$ is a continuous map. Define $\Omega_{1}:=\{t_{v}\mid v\in V_{n},n\in\mathbb{N}\}$, which is a countable subset of $\ell^{\infty}(\Gamma)$ as $\ell^{\infty}(\Gamma)$ identifies with $C(\beta\Gamma)$. If $A$ contains $\Omega_{1}$, then the action of $\Gamma$ on $\hat{A}$ is $\textnormal{ad}_{X}$-BLR. Indeed, the maps $t_{v}:\beta\Gamma\rightarrow [0,1]$ descend to the maps $\hat{A}\rightarrow [0,1]$, whence the maps $f_{n}: \beta\Gamma\rightarrow C_{n}$ descend to the maps $f_{n}:\hat{A}\rightarrow C_{n}$ by [\cite{Hig00}, Lemma 3.5]. On the other hand, [\cite{RS12}, Lemma 6.4] shows that there exists a countable subset $\Omega_{2}$ of $\ell^{\infty}(\Gamma)$ such that if $A$ contains $\Omega_{2}$, then the action of $\Gamma$ on $\hat{A}$ is free.
Let $\Omega=\Omega_{1}\cup\Omega_{2}$, a countable subset of $\ell^{\infty}(\Gamma)$. According to [\cite{RS12}, Lemma 6.7], there exists a countable $\Gamma$-invariant collection of projections $\mathcal{P}$ such that the $C^{*}$-algebra $C^{*}(\mathcal{P})$ generated by $\mathcal{P}$ contains $\Omega$. Let $A=C^{*}(\mathcal{P}, 1_{\ell^{\infty}(\Gamma)})$, then $\hat{A}$ is a $\Gamma$-space, and the action of $\Gamma$ is free and $\textnormal{ad}_{X}$-BLR. Since $A$ is unital and generated by countably many projections, the space $\hat{A}$ is compact, metrizible and totally disconnected.
\textbf{Step 3}. In this step, we prove that $\Gamma$ admits a free, minimal, $\textnormal{ad}_{X}$-BLR action on a totally disconnected, metrizable compact space $X$. Let $Y:=\hat{A}$, then $\Gamma$ admits a free, $\textnormal{ad}_{X}$-BLR action on a totally disconnected, metrizable compact space $Y$. Observe that if $Z$ is any non-empty closed $\Gamma$-invariant subset of $Y$, then $Z$ and the induced $\Gamma$ action on $Z$ have the above properties. Indeed, there is a nested sequence of closed non-empty $\Gamma$-invariant subsets of $Y$:
$$
\cdots\subseteq Y_{3}\subseteq Y_{2}\subseteq Y_{1}=Z\subseteq Y.
$$
Compactness of $Y$ implies that any descending chain for this order has a non-empty intersection. Zorn's lemma yields that there is a minimal element $X$. Thus the induced $\Gamma$ action on $X$ is minimal. Hence, $\Gamma$ admits a free, minimal, $\ad_{X}$-BLR action on a totally disconnected, metrizable compact space $X$. By Theorem \ref{th 3.15}, this action has dynamic asymptotic dimension function $\textnormal{dad}_{\Gamma\curvearrowright X}\leq\textnormal{ad}_{X}\approx h$, and thus has dynamic asymptotic dimension growth at most $h$.
\textbf{Step 4}. In the final step, we finish the proof by showing that this $X$ is a copy of the Cantor set, and it thus suffices to show that it has no isolated points. Observe that if $x\in X$ is isolated, then the orbit of $\Gamma\cdot x$ is open, and is dense in $X$ by minimality, thus $X\backslash\Gamma\cdot x=\varnothing$. Since the action is free and $\Gamma$ is infinite, the open cover $\{\{\gamma x\}\mid\gamma\in\Gamma\}$ of $X=\Gamma\cdot x$ has no finite subcover, which contradicts compactness. It follows that there is no isolated point.
\end{proof}
\begin{remark}
Suppose a countable discrete group $\Gamma$ has infinite asymptotic dimension and subexponential asymptotic dimension growth. In that case, it admits a free, minimal action on the Cantor set, and the dynamic asymptotic dimension of this action is infinite by [\cite{Bon24}, Theorem B].
\end{remark}
%\begin{remark}
%Combining Theorem \ref{th 4.17} and examples of groups with subexponential asymptotic dimension growth, we can construct several examples of group actions with subexponential dynamic asymptotic dimension growth.
%\end{remark}
%\begin{theorem}\label{th 6.6}\cite{HR00, Wil09}
%A countable discrete group $\Gamma$ has property A if and only if it admits an amenable action on some compact Hausdorff space.
%\end{theorem}
%Combining Theorem \ref{th 6.6} and [\cite{Oza12}, Theorem 1], we obtain the following result:
%\begin{corollary}Let $\Gamma$ be a countable discrete group with subexponential asymptotic dimension growth. Then $\Gamma$ admits an amenable action $\Gamma\curvearrowright X$ on a compact space $X$ which has at most subexponential dynamic asymptotic dimension growth. Moreover, the corresponding transformation groupoid $\mathcal{G}=X\rtimes\Gamma$ has at most subexponential dynamic asymptotic dimension growth, and it is amenable.\end{corollary}
%\begin{proof}Since $\Gamma$ has subexponential asymptotic dimension growth, $\Gamma$ has property $A$. Then $\Gamma$ admits an amenable action on some compact Hausdorff space $X$, which implies that this action has at most subexponential dynamic asymptotic dimension growth by Theorem \ref{th 4.15}. Moreover, the corresponding transformation groupoid $\mathcal{G}=X\rtimes\Gamma$ has at most subexponential dynamic asymptotic dimension growth by Lemma \ref{lemma 4.8}, and it is amenable as the action is.
%\end{proof}
\section{Partitions of unity}\label{section 5}
The goal of this section is to develop a critical tool called a generalized partition of unity. This tool serves not only as a partition of unity for groupoids with slow dynamic asymptotic dimension growth, but also provides almost invariant properties and beneficial estimates. Furthermore, this tool is essential for demonstrating the amenability of groupoids in the following section.
%Throughout this section, $\mathcal{G}$ is an \'etale groupoid, and $\ell$ is a coarse continuous length function on $\mathcal{G}$.
The key techniques needed in the proof of the amenability of groupoids in the next section are summarized in the following proposition. We are now ready to present our key technical proposition, which generalizes [\cite{GWY17}, Proposition 7.1].
% which is an optimized version of [\cite{GWY17}, Proposition 7.1].
\begin{proposition}\label{prop 5.1}
Let $\mathcal{G}$ be an \'etale groupoid with compact unit space, and with dynamic asymptotic dimension function $f\preceq x^{\alpha} (0<\alpha<1)$. Then for any $R>0$, for the open precompact subset $K={\{z\in\mathcal{G}\mid\ell(z)< R\}}$ of $\mathcal{G}$ and any $\varepsilon>0$, there exist a constant $c$ depending on parameters $\alpha, R, \varepsilon$, a positive integer $p$ depending only on $\alpha$, and an open cover $\{U_{0},\cdots, U_{f([(R+1)^{c}]R+1)}\}$ of $\mathcal{G}^{(0)}$ with the following properties:
\begin{enumerate}
\item For each $i$, the set $$\{g\in K\mid s(g), r(g)\in U_{i}\}$$ is contained in an ( open and ) precompact subgroupoid of $\mathcal{G}$.
\item For all $x\in\mathcal{G}^{(0)}$, the `partial orbit' $s(K\cap r^{-1}(x))$ is completely contained in some $U_{i}$.
\item There exists a collection of continuous functions $\{\phi_{i}:\mathcal{G}^{(0)}\rightarrow [0,1]\}_{i=0}^{f([(R+1)^{c}]R+1)}$ such that $\supp(\phi_{i})\subseteq U_{i}$ for each $i$, and $\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left(\phi_{i}(x)\right)^{p}=1$ for all $x\in\mathcal{G}^{(0)}$, and such that for any $g\in K$,
$$
p\cdot\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left|\phi_{i}(r(g))-\phi_{i}(s(g))\right|<\varepsilon.
$$
\end{enumerate}
\end{proposition}
To understand the above proposition more intuitively, we look at the case of group actions on compact spaces.
\begin{corollary}
Let $\Gamma\curvearrowright X$ be an action with $X$ compact, and with dynamic asymptotic dimension function $f\preceq x^{\alpha} (0<\alpha<1)$. Then for any $R>0$, for the finite subset $E=\{\gamma\in\Gamma\mid \ell(\gamma)<R\}$ of $\Gamma$ and any $\varepsilon>0$, there exist a constant $c$ depending on parameters $\alpha, R, \varepsilon$, a positive integer $p$ depending only on $\alpha$, and an open cover $\{U_{0},\cdots, U_{f([(R+1)^{c}]R+1)}\}$ of $X$ with the following properties:
\begin{enumerate}
\item For each $i$, the set $$\begin{Bmatrix}&\Big|& \text{ there exist }x\in U_{i}\text{ and }\gamma_{n},\cdots,\gamma_{1}\in E\text{ such that }\\\gamma\in\Gamma&\Big|&\gamma=\gamma_{n}\cdots\gamma_{2}\gamma_{1}\text{ and for all }k\in\{1,\cdots,n\},\qquad\\&\Big|&\gamma_{k}\cdots\gamma_{1}x\in U_{i}\qquad\qquad\qquad\qquad\qquad\qquad\qquad\end{Bmatrix}$$ is finite.
\item For all $x\in X$, the collection $E\cdot x:=\{\gamma^{-1}x\mid\gamma\in E\}$ is completely contained in some $U_{i}$.
\item There exists a collection of continuous functions $\{\phi_{i}:\mathcal{G}^{(0)}\rightarrow [0,1]\}_{i=0}^{f([(R+1)^{c}]R+1)}$ such that $\supp(\phi_{i})\subseteq U_{i}$ for each $i$, and $\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left(\phi_{i}(x)\right)^{p}=1$ for all $x\in X$, and such that for any $\gamma\in E$,
$$
p\cdot\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left|\phi_{i}(\gamma x)-\phi_{i}(x)\right|<\varepsilon.
$$
\end{enumerate}
\end{corollary}
To obtain the main proposition of this section, we need a couple of lemmas. The first step is to show the following lemma, which plays an important auxiliary role in the proof of subsequent lemmas and theorems.
\begin{lemma}\label{lemma 5.3}
Let $\mathcal{G}$ be an \'etale groupoid, and let $K$ be a symmetric subset of $\mathcal{G}$ containing $\mathcal{G}^{(0)}$. Let $\mathcal{U}=\{U_{j}\}_{j\in J}$ be an open cover of $\mathcal{G}^{(0)}$. Then for any $g\in K$, $s(g)\in U_{j}$ if and only if $r(g)\in U_{j}$ for some $j\in J$.
\end{lemma}
\begin{proof}
For any $x\in U_{j}\subseteq\mathcal{G}^{(0)}\subseteq K$, $s(x)=r(x)=x$, thus $U_{j}\subseteq K\cap r^{-1}(U_{j})$, which implies that $U_{j}=s(U_{j})\subseteq s(K\cap r^{-1}(U_{j}))$. If $g\in K$ and $s(g)\in U_{j}$ for some $j$, then $r(g)$ must be in $U_{j}$. Otherwise, $g\notin K\cap r^{-1}(U_{j})$, and so $s(g)\notin s(K\cap r^{-1}(U_{j}))\supseteq U_{j}$, hence $s(g)\notin U_{j}$, contradiction. Conversely, assume that $g\in K$ and $r(g)\in U_{j}$ for some $j$. Due to symmetry, $g^{-1}\in K$, $s(g^{-1})=r(g)\in U_{j}$. From the previous steps, we see that $s(g)=r(g^{-1})\in U_{j}$.
\end{proof}
Applying the above conclusion, we have the following lemma, which is an adaption of [\cite{GWY17}, Lemma 7.3].
\begin{lemma}\label{lemma 5.4}
Let $\mathcal{G}$ be an \'etale groupoid with compact unit space, and with dynamic asymptotic dimension growth $f$. Then for any $R>0$, for the open precompact subset $K={\{z\in\mathcal{G}\mid\ell(z)< R\}}$ of $\mathcal{G}$, there is a cover $\{U_{0},\cdots, U_{f(R)}\}$ of $\mathcal{G}^{(0)}$ by open precompact subsets of $\mathcal{G}^{(0)}$ such that for each $i$, the set
$$
\{g\in K\mid s(g), r(g)\in U_{i}\}
$$
generates an open precompact subgroupoid of $\mathcal{G}$, and moreover so that for each $x\in\mathcal{G}^{(0)}$, the set $s(K\cap r^{-1}(x))$ is completely contained in some $U_{i}$.
\end{lemma}
\begin{proof}
Since $\mathcal{G}$ has dynamic asymptotic dimension growth $f$ and $\mathcal{G}^{(0)}$ is compact, there is an open cover $\{U_{0},\cdots, U_{f(R)}\}$ of $\mathcal{G}^{(0)}$ by precompact subsets of $\mathcal{G}^{(0)}$ such that for each $i$, the set
$$
\{g\in K\mid s(g), r(g)\in U_{i}\}
$$
is contained in a precompact subgroupoid of $\mathcal{G}$, and hence generates an open precompact subgroupoid of $\mathcal{G}$. Observe that $K$ is symmetric as $\ell(x)=\ell(x^{-1})$ for all $x\in\mathcal{G}$. Moreover, for all $x\in\mathcal{G}^{(0)}\subseteq K$, we know that $r(x)=x$, so $x\in K\cap r^{-1}(x)$, thus $K\cap r^{-1}(x)\neq\varnothing$. For any $g\in K\cap r^{-1}(x)$, we have $g\in K$ and $r(g)=x\in U_{i}$ for some $i$. By Lemma \ref{lemma 5.3}, $s(g)\in U_{i}$. Hence, $s(K\cap r^{-1}(x))\subseteq U_{i}$ for some $U_{i}$.
\end{proof}
If the unit space of an \'etale groupoid is compact and Hausdorff, then it is normal, leading to an excellent result as follows.
\begin{lemma}\label{lemma 5.5}
Let $\mathcal{G}$ be an \'etale groupoid with compact unit space, and let $V_{0},\cdots, V_{n}$ be an open cover of $\mathcal{G}^{(0)}$. Then there exist open subsets $\mathcal{U}_{0},\cdots,\mathcal{U}_{n}$ of $\mathcal{G}^{(0)}$ such that $\bigcup\limits_{i=0}^{n}\mathcal{U}_{i}$ covers $\mathcal{G}^{(0)}$ and $\overline{\mathcal{U}_{i}}\subseteq V_{i}$ for each $i$.
\end{lemma}
\begin{proof}
We divide our proof into four steps. First, we show that for any open subset $V$ of $\mathcal{G}^{(0)}$ and any $x\in V$, there exists an open neighborhood $U_{x}$ of $x$ such that $\overline{U_{x}}\subseteq V$. Note that $\mathcal{G}^{(0)}$ is Hausdorff, thus the singleton set $\{x\}$ is closed. Additionally, since $\mathcal{G}^{(0)}$ is compact, it is also normal. Observe that the boundary $\partial V$ is closed, and $\{x\}\cap\partial V\subseteq V\cap\partial V=\varnothing$ as $V$ is open. Hence by normality of $\mathcal{G}^{(0)}$, there exist an open neighborhood $U_{x}$ of $x$ and open set $V'$ containing $\partial V$ such that
$$U_{x}\cap V'=\varnothing.$$
We may assume that $U_{x}\subseteq V$, otherwise we can replace $U_{x}$ with $U_{x}\cap V$. It suffices to show that $\overline{U_{x}}\subseteq V$. If this does not happen, we may find a point $y\in\partial U_{x}\subseteq\overline{V}$ such that $y\notin V$, which implies $y\in\partial V\subseteq V'$. It follows that $V'$ is an open neighborhood of $y$, and that $U_{x}\cap V'\neq\varnothing$ as $y\in\partial U_{x}$, contradiction. Hence, we conclude that $\overline{U_{x}}\subseteq V$, as required.
In the second step, we build subsets $V_{0}'$ and $\mathcal{U}_{0}$ of $\mathcal{G}^{(0)}$ such that $$V_{0}'\subseteq\mathcal{U}_{0}\subseteq\overline{\mathcal{U}_{0}}\subseteq V_{0}.$$ If $y\in\partial V_{0}$, namely $y\notin V_{0}$, then $y$ must be in $V_{i}$ for some $i\in\{1,\cdots,n\}$. Thus there exists an open neighborhood $U_{y}'$ of $y$ such that $U_{y}'\subseteq V_{i}$ as $V_{i}$ is open. Observe that for any $z\in\partial V_{i}$, $z$ and $y$ are distinct points. Since $\mathcal{G}^{(0)}$ is Hausdorff, there exist open neighborhoods $\widetilde{U_{z}}$ of $z$ and $\widetilde{U_{y}}$ of $y$ such that $\widetilde{U_{z}}\cap\widetilde{U_{y}}=\varnothing$. Set $U_{y}=\widetilde{U_{y}}\cap U_{y}'$, then $U_{y}\cap\widetilde{U_{z}}=\varnothing$ and $U_{y}\subseteq V_{i}$. Let $V_{0}'=\overline{V_{0}}\backslash\bigcup\limits_{y\in\partial V_{0}}U_{y}$, it is closed and hence compact. Note that $V_{0}'\subseteq V_{0}$, then by the first step, for any $x\in V_{0}'$, there exists an open neighborhood $U_{x}$ of $x$ such that $\overline{U_{x}}\subseteq V_{0}$ and $\bigcup\limits_{x\in V_{0}'}U_{x}$ is a cover of $V_{0}'$. In view of the compactness of $V_{0}'$, we can find finitely many points $x_{0}^{0},\cdots,x_{0}^{m}$ in $V_{0}'$ such that $V_{0}'\subseteq\bigcup\limits_{k=0}^{m}U_{x_{0}^{k}}$ with $\overline{U_{x_{0}^{k}}}\subseteq V_{0}$. Let $\mathcal{U}_{0}=\bigcup\limits_{k=0}^{m}U_{x_{0}^{k}}$, then $V_{0}'\subseteq\mathcal{U}_{0}$ and $\overline{\mathcal{U}_{0}}=\overline{\bigcup\limits_{k=0}^{m}U_{x_{0}^{k}}}=\bigcup\limits_{k=0}^{m}\overline{U_{x_{0}^{k}}}\subseteq V_{0}$.
Another step is to construct subsets $V_{i}'$ and $\mathcal{U}_{i}$ of $\mathcal{G}^{(0)}$ such that $U_{y}\subseteq V_{i}'$ when $y\in\partial V_{0}\cap V_{i}$ for some $i\in\{1,\cdots,n\}$, and such that $V_{i}'\subseteq\mathcal{U}_{i}\subseteq\overline{\mathcal{U}_{i}}\subseteq V_{i}$. From the second step, we see that $U_{y}\subseteq V_{i}$ and $U_{y}\cap\widetilde{U_{z}}=\varnothing$, $\forall z\in\partial V_{i}$. For fixed $i$ and $z\in\partial V_{i}$, $z\notin V_{i}$ implies that there is at least one $l\in\{0,\cdots,i-1,i+1,\cdots,n\}$ such that $z\in V_{l}$. Therefore, there exists an open neighborhood $U_{z}'$ of $z$ such that $U_{z}'\subseteq V_{l}$ as $V_{l}$ is open. Similar to the second step, for any $h\in\partial V_{l}$, there exist open neighborhoods $\widetilde{U_{h}}$ of $h$ and $U_{z}''$ of $z$ such that $\widetilde{U_{h}}\cap U_{z}''=\varnothing$. Replacing $U_{z}'$ by $U_{z}'\cap U_{z}''$, we have $\widetilde{U_{h}}\cap U_{z}'=\varnothing$ and $U_{z}'\subseteq V_{l}$. Set $U_{z}=\widetilde{U_{z}}\cap U_{z}'$, then $U_{z}\subseteq V_{l}$, $U_{z}\cap U_{y}=\varnothing$ and $U_{z}\cap\widetilde{U_{h}}=\varnothing$. Let $V_{i}'=\overline{V_{i}}\backslash\bigcup\limits_{z\in\partial V_{i}}U_{z}$. Since $U_{z}\cap U_{y}=\varnothing$, we get the desired inclusion $U_{y}\subseteq V_{i}'$. Repeating the second step, we may obtain an open subset $\mathcal{U}_{i}$ such that $V_{i}'\subseteq\mathcal{U}_{i}\subseteq\overline{\mathcal{U}_{i}}\subseteq V_{i}$.
Finally, we show that there are open subsets $\mathcal{U}_{0},\cdots, \mathcal{U}_{n}$ of $\mathcal{G}^{(0)}$ covering $\mathcal{G}^{(0)}$ such that $\overline{\mathcal{U}_{i}}\subset V_{i}$ for each $i$. By the third step, for any $y\in\partial V_{0}$, there exist $i\in\{1,\cdots,n\}$ and a closed subset $V_{i}'$ such that $U_{y}\subseteq V_{i}'$. Therefore, $\bigcup\limits_{y\in\partial V_{0}}U_{y}\subseteq\bigcup\limits_{i=1}^{n}V_{i}'\subseteq\bigcup\limits_{i=1}^{n}\mathcal{U}_{i}$. It follows that $\overline{V_{0}}=V_{0}'\cup\bigcup\limits_{y\in\partial V_{0}}U_{y}\subseteq\mathcal{U}_{0}\cup\bigcup\limits_{i=1}^{n}\mathcal{U}_{i}=\bigcup\limits_{i=0}^{n}\mathcal{U}_{i}$. Repeating the previous steps, we obtain
$$
\overline{V_{j}}\subseteq\bigcup\limits_{i=0}^{n}\mathcal{U}_{i}, \text{ for each } j\in\{0,\cdots,n\}.
$$
Hence, $\mathcal{G}^{(0)}\subseteq\bigcup\limits_{j=0}^{n}\overline{V_{j}}\subseteq\bigcup\limits_{i=0}^{n}\mathcal{U}_{i}$ with $\overline{\mathcal{U}_{i}}\subseteq V_{i}$ for each $i$.
\end{proof}
%With the help of the preceding two lemmas, we can now acquire the following key lemma, which is an adaption of [\cite{GWY17}, Lemma 7.4].
The following lemma constructs a nested family of open precompact subsets of $\mathcal{G}^{(0)}$ that satisfy several favorable properties. This is the main technical ingredient needed to prove the main proposition.
\begin{lemma}\label{lemma 5.6}
Let $\mathcal{G}$ be an \'etale groupoid with compact unit space, and with dynamic asymptotic dimension function $f$. Let $N:\mathbb{R}^{+}\rightarrow\mathbb{N}$ be a nondecreasing function. For any $R>0$, for the open precompact subset $K={\{z\in\mathcal{G}\mid\ell(z)< R\}}$ of $\mathcal{G}$, and each $i\in\{0,\cdots,f(N(R)R+1)\}$, there is a nested family
$$
U^{(0)}_{i}\subseteq U^{(1)}_{i}\subseteq\cdots\subseteq U_{i}^{(N(R))}
$$
of open precompact subsets of $\mathcal{G}^{(0)}$ with the following properties.
\begin{enumerate}
\item The family $\{U_{0}^{(0)},\cdots, U_{f(N(R)R+1)}^{(0)}\}$ covers $\mathcal{G}^{(0)}$;
\item For all $i,n$, $\overline{U^{(n)}_{i}}\subseteq U_{i}^{(n+1)}$;
\item For all $i,n$,
$$
s(K\cap r^{-1}(U^{(n)}_{i}))\subseteq U_{i}^{(n+1)};
$$
\item For all $i$, the set
$$
\{g\in K\mid s(g),r(g)\in U_{i}^{(N(R))}\}
$$
generates an open precompact subgroupoid of $\mathcal{G}$.
\end{enumerate}
\end{lemma}
\begin{proof}
We will prove this lemma in two steps. First, we will build a nested sequence of open subsets of $\mathcal{G}^{(0)}$. Given $R>0$ and a nondecreasing function $N:\mathbb{R}^{+}\rightarrow\mathbb{N}$. Observe that $K$ is symmetric and contains $\mathcal{G}^{(0)}$. For each $n\in\mathbb{N}$, we define
$$
\overline{K}^{n}=\{g_{n}\cdots g_{1}\mid g_{k}\in\overline{K} \text{ and } s(g_{k+1})=r(g_{k}),\forall k\in\{1,\cdots,n-1\}\}.
$$
Since $\overline{K}$ is compact, it follows from Lemma \ref{lemma 4.11} that $\overline{K}^{n}$ is also compact. Furthermore, we have $\overline{K}^{n}\subseteq\overline{K}^{n+1}$ for each $n\in\mathbb{N}$, provided we set $g_{n+1}=r(g_{n})\in \overline{K}$. For any $z\in\overline{K}^{N(R)}$, we can observe that $\ell(z)\leq N(R)R<N(R)R+1$, thus $\overline{K}^{N(R)}$ is contained in $K'=\{z\in\mathcal{G}\mid\ell(z)<N(R)R+1\}$, which is an open precompact symmetric subset of $\mathcal{G}$ containing $\mathcal{G}^{(0)}$. According to Lemma \ref{lemma 5.4}, there exist open precompact subsets $V_{0},\cdots, V_{f(N(R)R+1)}$ of $\mathcal{G}^{(0)}$ that cover $\mathcal{G}^{(0)}$ such that for each $i$, the set
$\{g\in K'\mid s(g),r(g)\in V_{i}\}$ generates an open precompact subgroupoid of $\mathcal{G}$.
In addition, for each $i\in\{0,\cdots,f(N(R)R+1)\}$ and each $n\in\{0,\cdots,N(R)\}$, we define
$$
V_{i}^{(n)}:=\{x\in V_{i}\mid s(\overline{K}^{N(R)-n}\cap r^{-1}(x))\subseteq V_{i}\}.
$$
The sequence $\mathcal{G}^{(0)}=\overline{K}^{0}\subseteq\overline{K}\subseteq\cdots\subseteq\overline{K}^{N(R)}$ induces a nested sequence of subsets
$$
V^{(0)}_{i}\subseteq V_{i}^{(1)}\subseteq\cdots\subseteq V_{i}^{(N(R))}=V_{i}.
$$
By Lemma \ref{lemma 5.4}, for any $x\in\mathcal{G}^{(0)}$, $s(K'\cap r^{-1}(x))\subseteq V_{i}$ for some $i$, namely, if $g\in K'$ with $r(g)=x$, then $s(g)\in V_{i}$. By Lemma \ref{lemma 5.3}, $x=r(g)\in V_{i}$. Further, $s(\overline{K}^{N(R)}\cap r^{-1}(x))\subseteq s(K'\cap r^{-1}(x))\subseteq V_{i}$, then $x\in V_{i}^{(0)}$. Therefore, the family $\{V_{0}^{(0)},\cdots, V^{(0)}_{f(N(R)R+1)}\}$ covers $\mathcal{G}^{(0)}$.
We claim that $V_{i}^{(n)}$ is open. Say for contradiction that $V_{i}^{(n)}$ is not open, then we may find a net $(x_{\lambda})_{\lambda\in\Lambda}$ in $\mathcal{G}^{(0)}$ with $x_{\lambda}\rightarrow x$ for some $x\in V_{i}^{(n)}$ such that for each $\lambda\in\Lambda$, there exists $g_{\lambda}\in \overline{K}^{N(R)-n}\cap r^{-1}(x_{\lambda})$ such that $s(g_{\lambda})\notin V_{i}$. Since $\overline{K}^{N(R)-n}$ is compact, the net $(g_{\lambda})_{\lambda\in\Lambda}$ has a convergent subnet $(g_{\lambda_{j}})_{j\in J}$, thus there exists $g\in\overline{K}^{N(R)-n}$ such that $\lim\limits_{j}g_{\lambda_{j}}=g$. As $r$ is continuous and $r(g_{\lambda})=x_{\lambda}$, $x=\lim\limits_{\lambda}x_{\lambda}=\lim\limits_{j}x_{\lambda_{j}}=\lim\limits_{j}r(g_{\lambda_{j}})=r(g)$, then $g\in\overline{K}^{N(R)-n}\cap r^{-1}(x)$. Because $V_{i}$ is open and $s(g_{\lambda})\notin V_{i}$, $s(g)\notin V_{i}$, which contradicts that $x\in V_{i}^{(n)}$. Hence, $V_{i}^{(n)}$ is open as claimed.
In the second step, we will construct the desired nested sequence $U_{i}^{(n)}$. Since $\mathcal{G}^{(0)}$ is compact, by Lemma \ref{lemma 5.5}, there exist open subsets $U_{0}^{(0)},\cdots, U_{f(N(R)R+1)}^{(0)}$ of $\mathcal{G}^{(0)}$ covering $\mathcal{G}^{(0)}$ and
$\overline{U_{i}^{(0)}}\subseteq V_{i}^{(0)}$ for each $i$. For each $i$,
$$
\overline{s(K\cap r^{-1}(U_{i}^{(0)}))}\subseteq s(\overline{K}\cap r^{-1}(\overline{U_{i}^{(0)}}))\subseteq s(\overline{K}\cap r^{-1}(V_{i}^{(0)}))\subseteq V_{i}^{(1)}.
$$
By normality of the compact set $\overline{V_{i}}$, there exists an open set $U_{i}^{(1)}$ such that
$$
\overline{s(K\cap r^{-1}(U_{i}^{(0)}))}\subseteq U_{i}^{(1)}\subseteq\overline{U_{i}^{(1)}}\subseteq V_{i}^{(1)}.
$$
We also see that
$$
\overline{s(K\cap r^{-1}(U_{i}^{(1)}))}\subseteq s(\overline{K}\cap r^{-1}(V_{i}^{(1)}))\subseteq V_{i}^{(2)},
$$
there exists an open set $U_{i}^{(2)}$ such that
$$
\overline{s(K\cap r^{-1}(U_{i}^{(1)}))}\subseteq U_{i}^{(2)}\subseteq\overline{U_{i}^{(2)}}\subseteq V_{i}^{(2)}.
$$
Repeating this process, we get a nested sequence of open subsets
$$
U_{i}^{(0)}\subseteq U_{i}^{(1)}\subseteq\cdots\subseteq U_{i}^{(N(R))},
$$
such that for each $i,n$,
$$
\overline{s(K\cap r^{-1}(U_{i}^{(n)}))}\subseteq U_{i}^{(n+1)}\text{ and }\overline{U_{i}^{(n)}}\subseteq V_{i}^{(n)}.
$$
Due to $U_{i}^{(n)}\subseteq V_{i}^{(n)}\subseteq V_{i}\subseteq\mathcal{G}^{(0)}$, $U_{i}^{(n)}$ is precompact. Moreover, $\overline{U_{i}^{(n)}}\subseteq\overline{s(K\cap r^{-1}(U_{i}^{(n)}))}\subseteq U_{i}^{(n+1)}$. Note that for all $i$, the set
$$
\{g\in K\mid s(g),r(g)\in U_{i}^{(N(R))}\}\subseteq\{g\in K'\mid s(g),r(g)\in V_{i}\},
$$
which implies that the set $\{g\in K\mid s(g),r(g)\in U_{i}^{(N(R))}\}$ generates an open precompact subgroupoid of $\mathcal{G}$.
\end{proof}
We have established the necessary lemmas for Proposition \ref{prop 5.1}, enabling us to commence its proof.
\begin{proof}[Proof of Proposition \ref{prop 5.1}]
\textbf{Step 1}. In the first step, we build a desired open cover of $\mathcal{G}^{(0)}$. Given $R>0$, the open precompact subset $K=\{z\in\mathcal{G}\mid \ell(z)<R\}$ of $\mathcal{G}$, and $\varepsilon>0$. Clearly, $K=K^{-1}$, $s(K)\cup r(K)=\mathcal{G}^{(0)}\subseteq K$. Let $p=[\frac{\alpha}{1-\alpha}]+1$, and let $N(R)=[(R+1)^{c}]$ for some constant $c$ in Lemma \ref{lemma 5.6}, such that
\begin{equation}\label{eq 5.1}
\frac{2p\left(f(N(R)R+1)+1\right)+2p\left(f(N(R)R+1)+1\right)^{\frac{1}{p}+1}}{N(R)}<\varepsilon,
\end{equation}
such constant $c$ exists and depends on given parameters $\alpha, R,\varepsilon$. Let $\{U_{i}^{(n)}\mid i\in\{0,\cdots,f(N(R)R+1)\}, n\in\{0,\cdots, N(R)\}\}$ have the properties in Lemma \ref{lemma 5.6}. For each $i$, let $U_{i}=U_{i}^{(N(R))}$. By Lemma \ref{lemma 5.6} (1)(2), $\{U_{0}^{(0)},\cdots, U_{f(N(R)R+1)}^{(0)}\}$ covers $\mathcal{G}^{(0)}$ and $U_{i}^{(0)}\subseteq U_{i}^{(N(R))}=U_{i}$, thus the family $\{U_{0},\cdots, U_{f(N(R)R+1)}\}$ covers $\mathcal{G}^{(0)}$. By Lemma \ref{lemma 5.6} (3), we have that $s(K\cap r^{-1}(U_{i}^{(0)}))\subseteq U_{i}^{(1)}\subseteq U_{i}$ for some $i$, which implies that for any $x\in\mathcal{G}^{(0)}$, $s(K\cap r^{-1}(x))\subseteq U_{i}$ for some $U_{i}$. By Lemma \ref{lemma 5.6} (4), we see that
$\{g\in K\mid s(g), r(g)\in U_{i}\} $ generates an open precompact subgroupoid of $\mathcal{G}$.
\textbf{Step 2}. The second step is to construct a partition of unity. By Lemma \ref{lemma 5.6} (2), for each $i\in\{0,\cdots,f(N(R)R+1)\}$, and each $n\in\{1,\cdots, N(R)\}$, $\overline{U_{i}^{(n-1)}}\subseteq U_{i}^{(n)}$. According to Urysohn's lemma, there exist continuous functions
$$
\psi_{i}^{(n)}:\mathcal{G}^{(0)}\rightarrow [0,1]
$$
such that $\psi_{i}^{(n)}(x)=1$, for all $x\in U_{i}^{(n-1)}$; $\psi^{(n)}_{i}(x)=0$, for all $ x\in\mathcal{G}^{(0)}\backslash U_{i}^{(n)}$.
For $i\in\{0,\cdots,f(N(R)R+1)\}$, put
$$
\psi_{i}=\frac{1}{N(R)}\sum\limits_{n=1}^{N(R)}\psi_{i}^{(n)},
$$
then $\supp(\psi_{i})\subseteq U_{i}$ and $\psi_{i}(x)\leq 1, \text{ for all }x\in\mathcal{G}^{(0)}$.
Define
$$
\phi_{i}=\frac{\psi_{i}}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}\right)^{\frac{1}{p}}},
$$
which are continuous functions and $\supp(\phi_{i})\subseteq U_{i}$.
By Lemma \ref{lemma 5.6} (1), for any $x\in\mathcal{G}^{(0)}$, there exists at least one $j\in\{0,\cdots,f(N(R)R+1)\}$ such that $x\in U_{j}^{(0)}$, thus $\psi_{j}(x)=1$, so
$\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(x)\right)^{\frac{1}{p}}\geq 1$, therefore $$0\leq\phi_{i}(x)\leq 1 \text{ and }
\sum\limits_{i=0}^{f(N(R)R+1)}\left(\phi_{i}(x)\right)^{p}=1. $$
\textbf{Step 3}. The final step is to show that for any $g\in K$,
$$p\cdot\sum\limits_{i=0}^{f(N(R)R+1)}\left|\phi_{i}(s(g))-\phi_{i}(r(g))\right|<\varepsilon. $$
For each $j\in\{0,\cdots,f(N(R)R+1)\}$, set $U_{j}^{(n)}=\mathcal{G}^{(0)}$ for $n\geq N(R)+1$.
Define $M=M_{j}:=\min\{n\mid r(g)\in U_{j}^{(n)}\}$. Thus $r(g)\in U_{j}^{(M)}\backslash U_{j}^{(M-1)}$, and
$$
\psi_{j}^{(n)}(r(g))=\begin{cases} 1,\quad & n\in [M+1, N(R)]\\ 0,\quad & n\in [1, M-1]\\ [0,1],\quad & n=M.\end{cases}
$$
Hence
\begin{equation}\label{eq 5.2}
\frac{N(R)-(M+1)}{N(R)}\leq\psi_{j}(r(g))\leq\frac{N(R)-M}{N(R)}.
\end{equation}
Since $g\in K$ and $r(g)\in U_{j}^{(M)}$, $g\in K\cap r^{-1}(U_{j}^{(M)})$. By Lemma \ref{lemma 5.6} (2), we see that $s(g)\in s(K\cap r^{-1}(U_{j}^{(M)}))\subseteq U_{j}^{(M+1)}$. Note that $r(g)\notin U_{j}^{(M-1)}\supseteq s(K\cap r^{-1}(U_{j}^{(M-2)}))$, thus $s(g^{-1})=r(g)\notin s(K\cap r^{-1}(U_{j}^{(M-2)}))$. We see that $g^{-1}\in K$ as $K=K^{-1}$, thus $g^{-1}\notin r^{-1}(U_{j}^{(M-2)})$, which implies that $s(g)=r(g^{-1})\notin U_{j}^{(M-2)}$. Therefore $s(g)\in U_{j}^{(M+1)}\backslash U_{j}^{(M-2)}$, and
$$
\psi_{j}^{(n)}(s(g))=\begin{cases} 1,\quad & n\in [M+2, N(R)]\\ 0,\quad & n\in [1, M-2]\\ [0,1],\quad & n=M-1, M, M+1.\end{cases}
$$
Hence
\begin{equation}\label{eq 5.3}
\frac{N(R)-(M+2)}{N(R)}\leq\psi_{j}(s(g))\leq\frac{N(R)-(M-1)}{N(R)}.
\end{equation}
Combining equation (\ref{eq 5.2}) and equation (\ref{eq 5.3}), we have that
\begin{equation}\label{eq 5.4}
\left|\psi_{j}(s(g))-\psi_{j}(r(g))\right|\leq\frac{2}{N(R)}.
\end{equation}
By Lemma \ref{lemma 5.3}, there exists at least one $j\in\{0,\cdots,f(N(R)R+1)\}$ such that $s(g), r(g)$ in $U_{j}^{(0)}$, hence $\psi_{j}(r(g))=\psi_{j}(s(g))=1$, which implies that
\begin{equation}\label{eq 5.5}
\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\geq 1 \text{ and } \sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\geq 1.
\end{equation}
For fixed $i\in\{0,\cdots,f(N(R)R+1)\}$ and $g\in K$,
\begin{align}\label{eq 5.6}
\left|\phi_{i}(r(g))-\phi_{i}(s(g))\right|&=\left|\frac{\psi_{i}(r(g))}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}}-\frac{\psi_{i}(s(g))}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\right)^{\frac{1}{p}}}\right|\nonumber\\
&\leq\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}}\left|\psi_{i}(r(g))-\psi_{i}(s(g))\right|\nonumber
\\&+\left|\psi_{i}(s(g))\right|\left|\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}}-\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\right)^{\frac{1}{p}}}\right|.
\end{align}
Combining equation (\ref{eq 5.4}) and equation (\ref{eq 5.5}), we have
\begin{equation}\label{eq 5.7}
\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}}\left|\psi_{i}(r(g))-\psi_{i}(s(g))\right|\leq\frac{2}{N(R)}.
\end{equation}
On the other hand, using equations (\ref{eq 5.4}) and (\ref{eq 5.5}) again, we obtain
\begin{align}\label{eq 5.8}
&\left|\psi_{i}(s(g))\right|\left|\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}}-\frac{1}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\right)^{\frac{1}{p}}}\right|\nonumber\\
&\leq \frac{\left|\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\right)^\frac{1}{p}-\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}\right|}{\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(r(g))\right)^{\frac{1}{p}}\left(\sum\limits_{j=0}^{f(N(R)R+1)}\psi_{j}^{p}(s(g))\right)^{\frac{1}{p}}}
\nonumber\\
&\leq\left(\sum\limits_{j=0}^{f(N(R)R+1)}\left|\psi_{j}(s(g))-\psi_{j}(r(g))\right|^{p}\right)^{\frac{1}{p}}\nonumber\\
&\leq\frac{2\left(f(N(R)R+1)+1\right)^{\frac{1}{p}}}{N(R)},
\end{align}
where the second inequality comes from the Minkowski inequality $$\left|\left(\sum\limits_{j=0}^{d}a_{j}^{p}\right)^\frac{1}{p}-\left(\sum\limits_{j=0}^{d}b_{j}^{p}\right)^{\frac{1}{p}}\right|\leq\left(\sum\limits_{j=0}^{d}|a_{j}-b_{j}|^{p}\right)^\frac{1}{p}, \text{ for }a_{j}, b_{j}\geq 0.$$
According to equations (\ref{eq 5.6}), (\ref{eq 5.7}), and (\ref{eq 5.8}), we have
\begin{align*}
\left|\phi_{i}(r(g))-\phi_{i}(s(g))\right|&\leq\frac{2}{N(R)}+\frac{2\left(f(N(R)R+1)+1\right)^{\frac{1}{p}}}{N(R)}\\
&=\frac{2+2\left(f(N(R)R+1)+1\right)^{\frac{1}{p}}}{N(R)}.
\end{align*}
Finally, from the equation (\ref{eq 5.1}), we obtain
\begin{align*}
&p\cdot\sum\limits_{i=0}^{f(N(R)R+1)}\left|\phi_{i}(r(g))-\phi_{i}(s(g))\right|\\
&\leq\frac{2p\left(f(N(R)R+1)+1\right)+2p\left(f(N(R)R+1)+1\right)^{\frac{1}{p}+1}}{N(R)}
<\varepsilon.
\end{align*}
\end{proof}
\begin{remark}
Let $0<\alpha<1$, $R>0$ and $\varepsilon>0$. We consider the function $f(x)=[x^{\alpha}]$, $p=[\frac{\alpha}{1-\alpha}]+1$ to examine the existence of the constant $c$. Through precise calculations, we found that if the constant $c$ satisfies the following inequality:
$$
c>\max\left\{\frac{\alpha(p+1)\In 2-p\In \frac{\varepsilon}{4p}}{\left(p-\alpha(p+1)\right)\In (R+1)},\frac{\alpha\In 2-\In\frac{\varepsilon}{8p}}{(1-\alpha)\In (R+1)}\right\},
$$
we have
$$\frac{2p\left(f([(R+1)^{c}]R+1)+1\right)+2p\left(f([(R+1)^{c}]R+1)+1\right)^{\frac{1}{p}+1}}{[(R+1)^{c}]}<\varepsilon.$$
\end{remark}
After thoroughly examining the proof of Proposition 7.1 in \cite{GWY17}, we conclude that finite dynamic asymptotic dimension leads to an almost invariant partition of unity without the restriction of the compactness of the unit space. Consequently, we can present a stronger result as follows:
\begin{proposition}[\cite{GWY17}]\label{prop 5.8}
Let $\mathcal{G}$ be an \'etale groupoid with dynamic asymptotic dimension $d$. Then for any open precompact subset $K$ of $\mathcal{G}$ and $\varepsilon>0$, there exists an open cover $\{U_{0},\cdots, U_{d}\}$ of $r(K)\cup s(K)$ with the following properties.
\begin{enumerate}
\item For each $i$, the set $$\{g\in K\mid s(g), r(g)\in U_{i}\}$$ is contained in an ( open and ) precompact subgroupoid of $\mathcal{G}$.
\item For each $x\in r(K)\cup s(K)$, the `partial orbit' $s(r^{-1}(x)\cap K)$ is completely contained in some $U_{i}$.
\item There exists a collection of continuous functions $\{\phi_{i}:\mathcal{G}^{(0)}\rightarrow [0,1]\}_{i=0}^{d}$ such that $\supp(\phi_{i})\subseteq U_{i}$ for each $i$, $\sum\limits_{i=0}^{d}\phi_{i}(x)\leq 1$ for all $x\in\mathcal{G}^{(0)}$, and $\sum\limits_{i=0}^{d}\phi_{i}(x)=1$ for all $x\in r(K)\cup s(K)$, and such that for any $g\in K$ and each $i$,
$$
\vert\phi_{i}(r(g))-\phi_{i}(s(g))\vert<\varepsilon.
$$
\end{enumerate}
\end{proposition}
\section{Amenability}
%The amenability of groupoids has several major results in various aspects, such as the Baum-Connes conjecture \cite{Tu99}, the Novikov conjecture \cite{Hig00}, exact $C^{*}$-algebras, and Eillott's conjecture \cite{}.
Our motivations for exploring the relationship between the dynamic asymptotic dimension growth of groupoids and amenability are threefold. Firstly, Guentner, Willett, and Yu showed that the finite dynamic asymptotic dimension of a free, locally compact Hausdorff \'etale groupoid implies amenability \cite{GWY17}. Secondly, they demonstrated that the finite dynamical complexity of a locally compact Hausdorff \'etale groupoid implies amenability \cite{GWY24}. It is worth noting that the finite dynamic asymptotic dimension leads to finite dynamical complexity when the dynamic asymptotic dimension is either zero or one. Lastly, according to Corollary \ref{cor 4.19} in the previous section, we see that for a discrete metric space with bounded geometry $X$, subexponential dynamic asymptotic dimension growth of the coarse groupoid $G(X)$ implies amenability.
Based on the above methods and results, we can also find sufficient conditions for the amenability of groupoids. By combining the ideas in \cite{GWY24} with the optimized Proposition \ref{prop 5.8}, we can establish the theorem that every locally compact Hausdorff \'etale groupoid $\mathcal{G}$ with finite dynamic asymptotic dimension is amenable, strengthening the result of \cite{GWY17}. Furthermore, using a similar approach and Proposition \ref{prop 5.1}, we can directly demonstrate that every $\sigma$-compact, locally compact Hausdorff \'etale groupoid $\mathcal{G}$ with compact unit space, and with dynamic asymptotic dimension growth at most $x^{\alpha}$ $(0<\alpha<1)$, is also amenable, thus generalizing Corollary \ref{cor 4.19} in the case of slow growth.
%Our motivations for discussing the relationship between the dynamic asymptotic dimension growth of a groupoid and the amenability are fourfold. First, in the work of \cite{Tu99}, Tu showed that the Baum-Connes conjecture holds for foliations whose holonomy groupoid is Hausdorff and amenable, particularly for amenable groupoids. Second, for a countable discrete group $\Gamma$ with property A, Higson showed that the Novikov conjecture holds \cite{Hig00}. If such group $\Gamma$ has bounded geometry, then property A is equivalent to the amenability of its associated coarse groupoid $G(\Gamma)$ \cite{STY02}. Therefore, if a countable discrete group with bounded geometry has the amenable coarse groupoid, then the Novikov conjecture holds. Third, according to Corollary \ref{cor 4.19}, we see that the subexponential dynamic asymptotic dimension growth of the coarse groupoid $G(X)$ implies amenability. Finally, analogous to the fact that finite dynamical complexity of a groupoid implies the amenability \cite{GWY24}, we prove that every locally compact Hausdorff \'etale groupoid $\mathcal{G}$ with finite dynamic asymptotic dimension is amenable by using the optimized Proposition \ref{prop 5.8}. More generally, applying Proposition \ref{prop 5.1} in the previous section, we directly prove that every $\sigma$-compact, locally compact Hausdorff \'etale groupoid $\mathcal{G}$ with compact unit space having dynamic asymptotic dimension growth at most $x^{\alpha}$ $(0<\alpha<1)$ is amenable.
A reference \cite{AR00} provides a standard description of an amenable groupoid, while another reference \cite{BO08} addresses the étale case. The following equivariant definition is a slight variation of Definition 5.6.13 from \cite{BO08} and is used in the proofs of two theorems in this section.
\begin{definition}[\cite{GWY24}]
Let $\mathcal{G}$ be a locally compact, Hausdorff, \'etale groupoid. Say $\mathcal{G}$ is amenable if for all compact subset $K\subseteq\mathcal{G}$ and all $\varepsilon>0$, there exists a continuous, compactly supported function $\mu:\mathcal{G}\rightarrow [0,1]$ such that:
\begin{enumerate}
\item for all $x\in\mathcal{G}^{(0)}$, we have $\sum\limits_{g\in\mathcal{G}_{x}}\mu(g)\leq 1$;
\item for all $k\in K$, we have $\vert 1-\sum\limits_{g\in\mathcal{G}_{r(k)}}\mu(g)\vert<\varepsilon$;
\item for all $k\in K$, we have $\sum\limits_{g\in\mathcal{G}_{r(k)}}\vert\mu(g)-\mu(gk)\vert<\varepsilon$.
\end{enumerate}
\end{definition}
The following lemma demonstrates that an open precompact subgroupoid of an \'etale groupoid is amenable. This fact is crucial in the proofs of two theorems in this section.
\begin{lemma}[\cite{GWY24}]\label{lemma 6.2}
Let $\mathcal{G}$ be an \'etale groupoid, and let $H$ be an open subgroupoid of $\mathcal{G}$ with compact closure. Then $H$ is amenable.
\end{lemma}
The following theorem is inspired in part by the work of Guentner, Willett, and Yu showing that finite dynamical complexity of a groupoid implies amenability \cite{GWY24}, and their earlier research on finite dynamic asymptotic dimension of \'etale groupoids \cite{GWY17}, as well as the fact that finite dynamic asymptotic dimension leads to finite dynamical complexity when the dynamic asymptotic dimension is either zero or one. They demonstrated that a free, locally compact, Hausdorff \'etale groupoid with finite dynamic asymptotic dimension is amenable by using the nuclear dimension for the reduced groupoid $C^{*}$-algebra (see \cite{GWY17}, Corollary 8.25). In contrast to their proof method in \cite{GWY17}, we adopt a direct approach similar to that in [\cite{GWY24}, Theorem A.9]. This allows us to strengthen their result [\cite{GWY17}, Corollary 8.25] by removing the freeness assumption for groupoids.
\begin{theorem}\label{th 6.3}
Let $\mathcal{G}$ be a locally compact, Hausdorff, \'etale groupoid with finite dynamic asymptotic dimension. Then $\mathcal{G}$ is amenable.
\end{theorem}
\begin{proof}
Let us assume that $\mathcal{G}$ has dynamic asymptotic dimension $d$. For any compact subset $K$ of $\mathcal{G}$ and any $\varepsilon>0$. Using local compactness, we expand $K$ slightly and write it as $K'$, which is open and precompact. Replacing $K'$ with $K'\cup K'^{-1}\cup r(K')\cup s(K')$, we may assume that $K'=K'\cup K'^{-1}\cup r(K')\cup s(K')$, which remains open and precompact. Next, let $\{U_{0},\cdots,U_{d}\}$ be an open cover of $r(K')\cup s(K')$, satisfying the properties outlined in Proposition \ref{prop 5.8} for the open precompact subset $K'$ and the error estimate $\frac{\varepsilon}{d+2}$. By taking the intersection of each $U_{i}$ with $s(K')\cup r(K')$, we may assume that each $U_{i}$ is contained in $s(K')\cup r(K')\subseteq K'$. Let $\mathcal{G}_{i}$ represent the subgroupoid generated by the set $K_{i}=\{k\in K'\mid s(k),r(k)\in U_{i}\}$. According to Proposition \ref{prop 5.8}, $\mathcal{G}_{i}$ is both open and precompact.
We claim that $K'=\bigcup\limits_{i}K_{i}$, where each $K_{i}$ is open, symmetric, precompact, and has compact closure in $\mathcal{G}_{i}$. We first demonstrate that $K'=\bigcup\limits_{i}K_{i}$. According to items (ii) and (iii) of [\cite{GWY17}, Lemma 7.4], we have $s(K'\cap r^{-1}(U_{i}^{(0)}))\subseteq U_{i}^{(1)}$ and $U_{i}^{(0)}\subseteq U_{i}^{(1)}\subseteq U_{i}$. Thus, it follows that $K'\cap r^{-1}(U_{i}^{(0)})\subseteq K_{i}$. Furthermore, by the item (i) of [\cite{GWY17}, Lemma 7.4], the collection $\{U_{0}^{(0)},\cdots, U_{d}^{(0)}\}$ covers $\overline{r(K')\cup s(K')}$. Therefore,
\begin{align*}
\bigcup\limits_{i=0}^{d}K_{i}\supseteq \bigcup\limits_{i=0}^{d}K'\cap r^{-1}(U_{i}^{(0)})&=K'\cap r^{-1}(\bigcup\limits_{i=0}^{d}U_{i}^{(0)})\\&\supseteq K'\cap r^{-1}(\overline{r(K')\cup s(K')})\supseteq K'.
\end{align*}
Clearly, $K_{i}$ is open, precompact, and has compact closure in $\mathcal{G}_{i}$. It remains to show that $K_{i}=K_{i}^{-1}$. Indeed, for any $k\in K_{i}$, specifically, $k\in K'$, $s(g),r(g)\in U_{i}$. Consequently, $k^{-1}\in K'$ since $K'=K'^{-1}$. Furthermore, since $r(k^{-1})=s(k)$ and $s(k^{-1})=r(k)$, we have $r(k^{-1}), s(k^{-1})\in U_{i}$, which implies that $k^{-1}\in K_{i}$. This completes the proof of the claim.
Since $\mathcal{G}_{i}$ is an open precompact subgroupoid of $\mathcal{G}$, it follows from Lemma \ref{lemma 6.2} that $\mathcal{G}_{i}$ is amenable. For each $i$, let $\mu_{i}:\mathcal{G}_{i}\rightarrow [0,1]$ be a continuous, compactly supported function as defined in the context of amenability concerning the compact subset $\overline{K_{i}}$ and the error estimate $\frac{\varepsilon}{d+2}$. We extend $\mu_{i}$ to the entire $\mathcal{G}$ by defining
$\mu_{i}(x)=0$ when $x\notin\mathcal{G}_{i}$.
We then define a function
$\mu:\mathcal{G}\rightarrow [0,1]$,
$g\mapsto\sum\limits_{i=0}^{d}\phi_{i}(s(g))\mu_{i}(g)$. Clearly, $\mu$ is a well-defined function that is continuous and has compact support in $\mathcal{G}$.
For any $x\in\mathcal{G}^{(0)}$,
\begin{align*}
\sum\limits_{g\in\mathcal{G}_{x}}\mu(g)&=\sum\limits_{g\in\mathcal{G}_{x}}\sum\limits_{i=0}^{d}\phi_{i}(s(g))\mu_{i}(g)\\
&=\sum\limits_{i=0}^{d}\phi_{i}(x)\sum\limits_{g\in(\mathcal{G}_{i})_{x}}\mu_{i}(g)\\
&\leq\sum_{i=0}^{d}\phi_{i}(x)\leq 1,
\end{align*}
where the third inequality comes from the amenability of $\mathcal{G}_{i}$ and the fact that if $x\in U_{i}$, then $x\in K'$ and $s(x)=r(x)=x\in U_{i}$, thus $x\in K_{i}\subseteq\mathcal{G}_{i}$, and if $x\notin U_{i}$, then $\phi_{i}(x)=0$ as $\supp(\phi_{i})\subseteq U_{i}$. The final equality is due to the item (3) of Proposition \ref{prop 5.8}.
For any $k\in K\subseteq K'$,
\begin{align*}
\left|1-\sum\limits_{g\in\mathcal{G}_{r(k)}}\mu(g)\right|&=\left|1-\sum\limits_{g\in\mathcal{G}_{r(k)}}\sum\limits_{i=0}^{d}\phi_{i}(s(g))\mu_{i}(g)\right|\\
&=\left|\sum\limits_{i=0}^{d}\phi_{i}(r(k))(1-\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(g))\right|\\
&\leq\sum\limits_{i=0}^{d}\phi_{i}(r(k))\left|1-\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(g)\right|\\&<\sum\limits_{i=0}^{d}\phi_{i}(r(k))\varepsilon=\varepsilon,
\end{align*}
where the fourth inequality follows from the amenability of $\mathcal{G}_{i}$ and the fact that if $r(k)\in U_{i}$, then $r(k)\in K_{i}$ and $r(r(k))=r(k)$, if $r(k)\notin U_{i}$, then $\phi_{i}(r(k))=0$. The fifth equality is due to the partition of unity as stated in Proposition \ref{prop 5.8}.
For any $k\in K\subseteq K'$,
\begin{align*}
\sum\limits_{g\in\mathcal{G}_{r(k)}}\left|\mu(g)-\mu(gk)\right|&=\sum\limits_{g\in\mathcal{G}_{r(k)}}\left|\sum\limits_{i=0}^{d}\left(\phi_{i}(s(g))\mu_{i}(g)-\phi_{i}(s(gk))\mu_{i}(gk)\right)\right|\\
&\leq\sum\limits_{i=0}^{d}\sum\limits_{g\in\mathcal{G}_{r(k)}}\phi_{i}(r(k))\left|\mu_{i}(g))-\mu_{i}(gk)\right|+\mu_{i}(gk)\left|\phi_{i}(r(k))-\phi_{i}(s(k))\right|\\
&=\sum\limits_{i=0}^{d}\phi_{i}(r(k))\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\left|\mu_{i}(g)-\mu_{i}(gk)\right|\\&+\sum\limits_{i=0}^{d}\left|\phi_{i}(r(k))-\phi_{i}(s(k))\right|\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(gk)\\&<\sum\limits_{i=0}^{d}\phi_{i}(r(k))\cdot\frac{\varepsilon}{d+2}+(d+1)\cdot\frac{\varepsilon}{d+2}\cdot 1=\varepsilon,
\end{align*}
where the fourth inequality follows from the amenability of $\mathcal{G}_{i}$ and Proposition \ref{prop 5.8}.
\end{proof}
The following main theorem is inspired by Corollary \ref{cor 4.19}, which shows that for a discrete metric space of bounded geometry $X$, the associated coarse groupoid $G(X)$ with subexponential dynamic asymptotic dimension growth is amenable. It follows that slow dynamic asymptotic dimension growth of the groupoid $G(X)$ guarantees its amenability as this condition indicates subexponential growth. By adopting an approach analogous to that of Theorem \ref{th 6.3} and using the key technical Proposition \ref{prop 5.1}, we can generalize the result of Corollary \ref{cor 4.19} for the case of slow growth.
\begin{theorem}\label{th 6.4}
Let $\mathcal{G}$ be a $\sigma$-compact \'etale groupoid with compact unit space, and with dynamic asymptotic dimension growth $f\preceq x^{\alpha} (0<\alpha<1)$. Then $\mathcal{G}$ is amenable.
\end{theorem}
\begin{proof}
Without loss of generality, we may assume that $\mathcal{G}$ has dynamic asymptotic dimension function $f\preceq x^{\alpha}$ for $\alpha\in (0,1)$. For any compact subset $K$ of $\mathcal{G}$ and any $\varepsilon>0$, there exists $R>0$ such that $K\subseteq K'=\{z\in\mathcal{G}\mid\ell(z)<R\}$ as the length function $\ell$ is controlled. Note that $K'$ is open and precompact, $K'=K'^{-1}$, and $s(K')\cup r(K')=\mathcal{G}^{(0)}\subseteq K'$. By Proposition \ref{prop 5.1}, there exists a constant $c$ depending on parameters $\alpha, R, \varepsilon$, a positive integer $p$ depending only on $\alpha$, and an open cover $\{U_{0},\cdots, U_{f([(R+1)^{c}]R+1)}\}$ of $\mathcal{G}^{(0)}$ with the properties for the open precompact subset $K'$ and the error estimate $\frac{\varepsilon}{2}$.
Let $\mathcal{G}_{i}$ be the subgroupoid generated by the subset $K_{i}=\{k\in K'\mid s(k), r(k)\in U_{i}\}$. Similar to the claim in the proof of Theorem \ref{th 6.3}, we also have the following statement: $K'=\bigcup\limits_{i}K_{i}$ and $K_{i}$ is open, symmetric, precompact, and has compact closure in $\mathcal{G}_{i}$. According to Proposition \ref{prop 5.1}, $\mathcal{G}_{i}$ is an open precompact subgroupoid of $\mathcal{G}$. Therefore, by Lemma \ref{lemma 6.2}, $\mathcal{G}_{i}$ is amenable. For each $i$, let $\mu_{i}:\mathcal{G}_{i}\rightarrow [0,1]$ be a continuous, compactly supported function as in the definition of amenability concerning the compact subset $\overline{K_{i}}$ and the error estimate $\frac{\varepsilon}{2}$. We extend $\mu_{i}$ to the whole $\mathcal{G}$ by defining
$\mu_{i}(x)=0$, if $x\notin\mathcal{G}_{i}$. We then define
$\mu:\mathcal{G}\rightarrow [0,1], g\mapsto\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(s(g))\mu_{i}(g)$, which is a well-defined continuous, compactly supported function on $\mathcal{G}$.
For any $x\in\mathcal{G}^{(0)}$,
\begin{align*}
\sum\limits_{g\in\mathcal{G}_{x}}\mu(g)&=\sum\limits_{g\in\mathcal{G}_{x}}\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(s(g))\mu_{i}(g)\\
&=\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(x)\sum\limits_{g\in(\mathcal{G}_{i})_{x}}\mu_{i}(g)\\
&\leq\sum_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(x)=1,
\end{align*}
where the third inequality comes from the amenability of $\mathcal{G}_{i}$, and the fourth one is due to the partition of unity in Proposition \ref{prop 5.1}.
For any $k\in K\subseteq K'$,
\begin{align*}
\left|1-\sum\limits_{g\in\mathcal{G}_{r(k)}}\mu(g)\right|&=\left|1-\sum\limits_{g\in\mathcal{G}_{r(k)}}\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(s(g))\mu_{i}(g)\right|\\
&=\left|\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(r(k))(1-\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(g))\right|\\
&\leq\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(r(k))\left|1-\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(g)\right|\\&<\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(r(k))\cdot\frac{\varepsilon}{2}=\frac{\varepsilon}{2}<\varepsilon,
\end{align*}
where the fourth inequality follows from the amenability of $\mathcal{G}_{i}$, and the fifth equality comes from the partition of unity in Proposition \ref{prop 5.1}.
For any $k\in K\subseteq K'$,
\begin{align*}
\sum\limits_{g\in\mathcal{G}_{r(k)}}\left|\mu(g)-\mu(gk)\right|&\leq\sum\limits_{g\in\mathcal{G}_{r(k)}}\left|\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left(\phi_{i}^{p}(s(g))\mu_{i}(g)-\phi_{i}^{p}(s(gk))\mu_{i}(gk)\right)\right|\\
&\leq\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\sum\limits_{g\in\mathcal{G}_{r(k)}}\phi_{i}^{p}(r(k))\left|\mu_{i}(g)-\mu_{i}(gk)\right|+\mu_{i}(gk)\left|\phi_{i}^{p}(r(k))-\phi_{i}^{p}(s(k))\right|\\
&=\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(r(k))\cdot\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\left|\mu_{i}(g)-\mu_{i}(gk)\right|\\&+\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left|\phi_{i}(r(k))-\phi_{i}(s(k))\right|\cdot\left|\sum\limits_{j=0}^{p-1}\phi_{i}(r(k))^{j}\phi_{i}(s(k))^{p-1-j}\right|\cdot\sum\limits_{g\in(\mathcal{G}_{i})_{r(k)}}\mu_{i}(gk)\\&<\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\phi_{i}^{p}(r(k))\cdot\frac{\varepsilon}{2}+p\cdot\sum\limits_{i=0}^{f([(R+1)^{c}]R+1)}\left|\phi_{i}(r(k))-\phi_{i}(s(k))\right|\\
&<\frac{\varepsilon}{2}+\frac{\varepsilon}{2}=\varepsilon,
\end{align*}
where the third equality comes from the equality $\left|a^{p}-b^{p}\right|=\left|a-b\right|\cdot\left|\sum\limits_{j=0}^{p-1}a^{j}b^{p-1-j}\right|$ for all $a, b\geq 0$, the fourth inequality is due to the amenability of $\mathcal{G}_{i}$, and the final inequality follows from the item (3) of Proposition \ref{prop 5.1}.
\end{proof}
\begin{example}
Let $f: \mathbb{R}^{+}\rightarrow\mathbb{N}$ be a strictly increasing function such that $f\preceq x^{\alpha}(0<\alpha<1)$, and let $X=\bigoplus\limits_{n=1}^{\infty}f(n)\mathbb{Z}$, which is endowed with the direct sum metric. Then $X$ is a discrete metric space with slow asymptotic dimension growth $f$, a subexponential growth. Moreover, $X$ has bounded geometry since the sequence $\{f(n)\}$ is (strictly) increasing and every element of $X$ has only finitely many nonzero terms. According to Theorem \ref{th 4.18}, $X$ has property $A$. Hence, by Corollary \ref{cor 4.19}, the corresponding coarse groupoid $G(X)$ is amenable.
\end{example}
\section*{Acknowledgments}
We thank Jiawen Zhang for many valuable suggestions on asymptotic dimension growth, Kang Li and Zehong Huang for their helpful comments on dynamic asymptotic dimension, and Jianchao Wu for discussions on coarse groupoids.
\begin{thebibliography}{WXY21b}
\bibitem[ALSS21]{ALSS21}
M. Amini, K. Li, D. Sawicki, and A. Shakibazadeh,
\newblock{Dynamic asymptotic dimension for actions of virtually cyclic groups},
\newblock{\em Proc. Edinb. Math. Soc. (2)}, {\bf 64}(2): 364--372, 2021.
\bibitem[AR00]{AR00}
C. Anantharaman-Delaroche and J. Renault,
\newblock Amenable groupoids,
\newblock Monogr. Enseign. Math., {\bf 36} [Monographs of L'Enseignement Mathématique]
L'Enseignement Mathématique, Geneva, 2000.
\bibitem[ANWZ18]{ANWZ18}
G. Arzhantseva, G. A. Niblo, N. Wright, and J. Zhang,
\newblock A characterization for asymptotic dimension growth,
\newblock{\em Algebr. Geom. Topol.}, {\bf 18}(1): 493--524, 2018.
\bibitem[Bar16]{Bar16}
A. Bartels,
\newblock On proofs of the Farrell-Jones conjecture,
\newblock Springer Proc. Math. Stat.,{\bf 184}, Springer, [Cham], 1--31, 2016.
\bibitem[BLR08]{BLR08}
A. Bartels, W. L\"uck, and H. Reich,
\newblock Equivariant covers for hyperbolic groups,
\newblock{\em Geom. Topol.}, {\bf 12}(3): 1799--1882, 2008.
\bibitem[BHS17]{BHS17}
J. Behrstock, M. F. Hagen, and A. Sisto,
\newblock Asymptotic dimension and small-cancellation for
hierarchically hyperbolic spaces and groups,
\newblock{\em Proc. Lond. Math. Soc.}, {\bf 114}: 890--926, 2017.
\bibitem[Bel05]{Bel05}
G. Bell,
\newblock Growth of the asymptotic dimension function for groups,
\newblock{arxiv: 0504577v1}, 2005.
\bibitem[BD01]{BD01}
G Bell, A Dranishnikov,
\newblock On asymptotic dimension of groups,
\newblock{\em Algebr. Geom. Topol.}, {\bf 1}: 57--71, 2001.
\bibitem[BD11]{BD11}
G. Bell and A. Dranishnikov,
\newblock Asymptotic dimension in B\c{e}dlewo,
\newblock{\em Topology Proc.}, {\bf 38}: 209--236, 2011.
\bibitem[BF08]{BF08}
G. Bell and K. Fujiwara,
\newblock The asymptotic dimension of a curve graph is finite,
\newblock{\em J. Lond. Math. Soc. (2)}, {\bf 77}(1): 33--50, 2008.
\bibitem[BBF15]{BBF15}
M. Bestvina, K. Bromberg, and K. Fujiwara,
\newblock Constructing group actions on quasi-trees and applications to mapping class groups,
\newblock{\em Publ. Math. Inst. Hautes Études Sci.}, {\bf 122}: 1--64, 2015.
\bibitem[Bon24]{Bon24}
C. B\"onicke,
\newblock{On the dynamic asymptotic dimension of \'etale groupoids}
\newblock{\em Math. Z.}, {\bf 307}(1), Paper No. 16, 2024.
\bibitem[BDGW23]{BDGW23}
C. B\"onicke, C. Dell'Aiera, J. Gabe, and R. Willett,
\newblock{Dynamic asymptotic dimension and Matui's HK conjecture},
\newblock{\em Proc. Lond. Math. Soc. (3)}, {\bf 126}(4): 1182--1253, 2023.
\bibitem[BL24]{BL24}
C. B\"onicke and K. Li,
\newblock{Nuclear dimension of subhomogeneous twisted groupoid $\rm C^*$-algebras and dynamic asymptotic dimension},
\newblock{\em Int. Math. Res. Not. IMRN}, {\bf 2024}(16): 11597--11610, 2024.
\bibitem[BO08]{BO08}
N. P. Brown and N. Ozawa,
\newblock $\rm C^{*}$-algebras and finite-dimensional approximations,
\newblock{\em Grad. Stud. Math.}, {\bf 88},
American Mathematical Society, RI, 2008.
\bibitem[CFP96]{CFP96}
J. W. Cannon, W. J. Floyd, and W. R. Parry,
\newblock Introductory notes on Richard Thompson’s
groups,
\newblock{\em Enseign. Math. (2)}, {\bf 42}(3-4): 215--256, 1996.
\bibitem[CFY09]{CFY08}
S. Chang, S. Ferry, and G. Yu,
\newblock Bounded rigidity of manifolds and asymptotic dimension growth,
\newblock{\em J. $K$-Theory}, {\bf 1}(1): 129--144, 2008.
\bibitem[Con94]{ConnesNCG}
A. Connes,
\newblock {Noncommutative geometry},
\newblock Academic Press, Inc., San Diego, CA, 1994.
\bibitem[CJMST23]{CJMST23}
C. Conley, S. Jackson, A. Marks, B. Seward, and R. Tucker-Drob,
\newblock{Borel asymptotic dimension and hyperfinite equivalence relations},
\newblock{\em Duke Math. J.}, {\bf 172}(16): 3175--3226, 2023.
\bibitem[CDGHV24]{CDGHV24}
K. Courtney, A. Duwenig, M. C. Georgescu, A. Huef, and M. G. Viola,
\newblock{Alexandrov groupoids and the nuclear dimension of twisted groupoid $\rm C^{*}$-algebras},
\newblock{\em J. Funct. Anal.}, {\bf 286}(9), Paper No. 110372, 2024.
\bibitem[Chu18]{Chu18}
Y. C. Chung.
\newblock{Quantitative $K$-theory for Banach algebras},
\newblock{\em J. Funct. Anal.}, {\bf 274}(1): 278--340, 2018.
\bibitem[DS18]{DS18}
R. J. Deeley and K. R. Strung,
\newblock{Nuclear dimension and classification of $\rm C^{*}$-algebras associated to Smale spaces},
\newblock{\em Trans. Amer. Math. Soc.}, {\bf 370}(5): 3467--3485, 2018.
\bibitem[Dra00]{Dra00}
A. Dranishnikov,
\newblock Asymptotic topology,
\newblock {\em Russian Math. Surveys}, {\bf 55}(6): 1085--1129, 2000.
\bibitem[Dra06]{Dra06}
A. Dranishnikov,
\newblock Groups with a polynomial dimension growth,
\newblock {\em Geom. Dedicata}, {\bf 119}:1--15, 2006.
\bibitem[DS11]{DS11}
A. Dranishnikov and M. Sapir,
\newblock On the dimension growth of groups,
\newblock{\em J. Algebra}, {\bf 347}: 23--39, 2011;
\newblock corrigendum to "On the dimension growth of groups".
\newblock{\em J. Algebra}, {\bf 370}, 407--409, 2012.
\bibitem[FP21]{FP21}
K. Fujiwara, and P. Papasoglu,
\newblock Asymptotic dimension of planes and planar graphs,
\newblock{\em Trans. Amer. Math. Soc.}, {\bf 374}12: 8887--8901, 2021.
\bibitem[FS24]{FS24}
K. Fujiwara and S. Schleimer,
\newblock Asymptotic dimensions of the arc graphs and disk graphs,
\newblock{\em Algebr. Geom. Topol.}, {\bf 24}(5): 2655--2672, 2024.
\bibitem[Gro93]{Gro93}
M.~Gromov,
\newblock Asymptotic invariants of infinite groups,
\newblock In {\em Geometric group theory, {V}ol. 2 ({S}ussex, 1991)}, {\bf
182} of {\em London Math. Soc. Lecture Note Ser.}, pages 1--295. Cambridge
Univ. Press, Cambridge, 1993.
\bibitem[Gro03]{Gro03}
M.~Gromov,
\newblock Random walk in random groups,
\newblock{\em Geom. Funct. Anal.}, {\bf 13}(1): 73--146, 2003.
\bibitem[GWY17]{GWY17}
E. Guentner, R. Willett, and G. Yu,
\newblock Dynamic asymptotic dimension: relation to dynamics, topology, coarse geometry, and $\rm C^{*}$-algebras,
\newblock {\em Math. Ann.},{\bf 367}(1-2): 785--829, 2017.
\bibitem[GWY24]{GWY24}
E. Guentner, R. Willett, and G. Yu,
\newblock Dynamical complexity and controlled operator {$K$}-theory,
\newblock {\emph{Ast\'{e}risque}}, {\bf 451}, 2024.
\bibitem[HW24]{HW24}
Z. He, and S. Wei,
\newblock A note on the nuclear dimension of Cuntz-Pimsner $\rm C^{*}$-algebras associated with minimal shift spaces,
\newblock{\em Canad. J. Math.}, {\bf 76}(1): 104--125, 2024.
\bibitem[Hig00]{Hig00}
N. Higson,
\newblock Bivariant $K$-theory and the Novikov conjecture,
\newblock{\em Geom. Funct. Anal.}, {\bf 10}(3): 563--581, 2000.
\bibitem[HR00]{HR00}
N. Higson and J. Roe,
\newblock Amenable group actions and the Novikov conjecture,
\newblock{\em J. Reine Angew. Math.}, {\bf 519}: 143--153, 2000.
\bibitem[KKLRU21]{KKLRU21}
M. Kennedy, S.-J. Kim, X. Li, S. Raum, D. Ursu
\newblock{The ideal intersection property for essential groupoid $\rm C^{*}$-algebras}
\newblock{arXiv:2107.03980, 2021}
\bibitem[Ker20]{Ker20}
D. Kerr,
\newblock{Dimension, comparison, and almost finiteness},
\newblock{\em J. Eur. Math. Soc.}, {\bf 22}(11): 3697--3745, 2020.
\bibitem[LLW23]{LLW23}
K. Li, H.-C. Liao, and W. Winter,
\newblock{The diagonal dimension of sub-$\rm C^*$-algebras},
\newblock{arXiv: 2303.16762, 2023}.
\bibitem[NY12]{NY12}P. W. Nowak and G. Yu,
{Large Scale Geometry}, EMS Textbooks in Mathematics,
European Mathematical Society, Zürich, 2012.
\bibitem[Opp14]{Opp14}
I. Oppenheim,
\newblock An intermediate quasi-isometric invariant between subexponential asymptotic dimension growth and Yu's property A,
\newblock{\em Internat. J. Algebra Comput.}, {\bf 24}(6): 909--922, 2014.
\bibitem[Osi05]{Osi05}
D. Osin,
{Asymptotic dimension of relatively hyperbolic groups},
\emph{Int. Math. Res. Not.}, {\bf 2005}(35): 2143--2161, 2005.
\bibitem[Oza12]{Oza12}
N. Ozawa,
\newblock Metric spaces with subexponential asymptotic dimension growth,
\newblock{\em Internat. J. Algebra Comput.}, {\bf 22}(2): 1250011, 2012.
\bibitem[Ren80]{Ren80}
J. Renault,
\newblock A groupoid approach to $\rm C^{*}$-algebras,
\newblock Lecture Notes in Math., {\bf 793}, Springer, Berlin, 1980.
\bibitem[OY19]{OY19}
H. Oyono-Oyono and G. Yu,
\newblock Quantitative $K$-theory and the K\"unneth formula for operator algebras,
\newblock{\em J. Funct. Anal.}, {\bf 277}(7): 2003--2091, 2019.
\bibitem[RS12]{RS12}
M. R{\o}rdam and A. Sierakowski,
\newblock Purely infinite $\rm C^{*}$-algebras arising from crossed products,
\newblock{\em Ergod. Theory Dynam. Syst.}, {\bf 32}(1): 273--293, 2012.
\bibitem[Roe93]{Roecoarse}
J. Roe,
\newblock Coarse cohomology and index theory on complete {R}iemannian
manifolds,
\newblock {\em Mem. Amer. Math. Soc.}, {\bf 104}(497):x+90, 1993.
\bibitem[Roe03]{Roe03}
J. Roe,
\newblock Lectures on Coarse Geometry,
\newblock{\em Univ. Lecture Ser.}, {\bf 31},
\newblock{American Mathematical Society}, Providence, RI, 2003.
\bibitem[Roe05]{Roe05}
J. Roe,
\newblock Hyperbolic groups have finite asymptotic dimension,
\newblock{\em Proc. Amer. Math. Soc.}, {\bf 133}(9): 2489--2490, 2005.
\bibitem[Smi07]{Smi07}
J. Smith,
\newblock The asymptotic dimension of the first Grigorchuk group is infinity,
\newblock{\em Rev. Mat. Complut.}, {\bf 20}(1): 119--121, 2007.
\bibitem[STY02]{STY02}
G. Skandalis, J-L Tu, and G. Yu,
\newblock The coarse Baum–Connes conjecture and groupoids,
\newblock{\em Topology}, {\bf 41}(4): 807--834, 2002.
\bibitem[Tse23]{Tse23}
P. Tselekidis,
\newblock Asymptotic dimension of graphs of groups and one-relator groups,
\newblock{\em Algebr. Geom. Topol.}, {\bf 23}(8): 3587--3613, 2023.
\bibitem[WWZZ23]{WWZZ23}
H. Wang, Y. Wang, J. Zhang, and D. Zhou,
\newblock{$L^p$ coarse Baum-Connes conjecture via $C_0$ coarse geometry},
\newblock{arXiv:2311.05333, 2023}.
\bibitem[WWZZ24]{WWZZ24}
H. Wang, Y. Wang, J. Zhang, and D. Zhou,
\newblock Persistence approximation property for $L^p$ operator algebras,
\newblock{\em Chin. Ann. Math. Ser. B}, {\bf 45}(6): 869--904, 2024.
\bibitem[Tu99]{Tu99}
J-L Tu,
\newblock La conjecture de Baum-Connes pour les feuilletages moyennables,
\newblock{\em K-theory}, {\bf 17}(3): 215--264, 1999.
\bibitem[WY20]{Willett_Yu_Book_higher_index_theory}
R. Willett and G. Yu,
\newblock { Higher index theory}, volume 189 of {\em Cambridge Studies in
Advanced Mathematics},
\newblock Cambridge University Press, Cambridge, 2020.
\bibitem[Wil09]{Wil09}
R. Willett,
\newblock {Some notes on Property A}, in Limits of graphs in group theory and computer science,
\newblock EPFL Press, Lausanne, 191--281, 2009.
\bibitem[MW20]{MW20}
X. Ma and J. Wu,
\newblock Almost elementariness and fiberwise amenability for \'etale groupoids,
\newblock {arxiv:2011.01182v2}, 2020.
\bibitem[Wri12]{Wri12}
N. Wright,
\newblock Finite asymptotic dimension for CAT(0) cube complexes,
\newblock{\em Geom. Topol.}, {\bf 16}(1): 527--554, 2012.
\bibitem[Yu97]{Yu_zero_spectrum}
G. Yu,
\newblock Zero-in-the-spectrum conjecture, positive scalar curvature and
asymptotic dimension,
\newblock {\em Invent. Math.}, {\bf 127}(1):99--126, 1997.
\bibitem[Yu98]{Yu98}
G. Yu,
\newblock The {N}ovikov conjecture for groups with finite asymptotic dimension,
\newblock {\em Ann. of Math. (2)}, {\bf 147}(2):325--355, 1998.
\bibitem[Yu00]{Yu00}
G. Yu,
\newblock The coarse {B}aum-{C}onnes conjecture for spaces which admit a
uniform embedding into {H}ilbert space,
\newblock {\em Invent. Math.}, {\bf 139}(1):201--240, 2000.
\bibitem[ZZ21]{ZZ21}
J. Zhang and D. Zhou,
{$L^{p}$ coarse Baum–Connes conjecture and $K$-theory for $L^{p}$ Roe algebras},
\emph{J. Noncommut. Geom.}, {\bf 15}(4): 1285--1322, 2021.
\end{thebibliography}
\end{document}