Boundedness of operators generated by fractional semigroups associated with Schr\"odinger operators on Campanato type spaces via $T1$ theorem
Extracted source text ยท arXiv 2105.03717
%%%%%%%%%%%%%%%%%%%%%%% file template.tex %%%%%%%%%%%%%%%%%%%%%%%%%
%
% This is a general template file for the LaTeX package SVJour3
% for Springer journals. Springer Heidelberg 2010/09/16
%
% Copy it to a new file with a new name and use it as the basis
% for your article. Delete % signs as needed.
%
% This template includes a few options for different layouts and
% content for various journals. Please consult a previous issue of
% your journal as needed.
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% First comes an example EPS file -- just ignore it and
% proceed on the \documentclass line
% your LaTeX will extract the file if required
\begin{filecontents*}{example.eps}
%!PS-Adobe-3.0 EPSF-3.0
%%BoundingBox: 19 19 221 221
%%CreationDate: Mon Sep 29 1997
%%Creator: programmed by hand (JK)
%%EndComments
gsave
newpath
20 20 moveto
20 220 lineto
220 220 lineto
220 20 lineto
closepath
2 setlinewidth
gsave
.4 setgray fill
grestore
stroke
grestore
\end{filecontents*}
%
\RequirePackage{fix-cm}
%
%\documentclass{svjour3} % onecolumn (standard format)
%\documentclass[smallcondensed]{svjour3} % onecolumn (ditto)
\documentclass[smallextended]{svjour3} % onecolumn (second format)
%\documentclass[twocolumn]{svjour3} % twocolumn
%
\smartqed % flush right qed marks, e.g. at end of proof
%
\usepackage{graphicx}
\usepackage{mathptmx} % use Times fonts if available on your TeX system
%
% insert here the call for the packages your document requires
\usepackage{latexsym}
\usepackage{amsmath,amssymb,txfonts}
\usepackage{amssymb}
\usepackage{amsmath}
\usepackage{mathrsfs}
\usepackage{amsmath,amssymb}
\usepackage{color}
\usepackage{extarrows}
\allowdisplaybreaks[4]
%
% please place your own definitions here and don't use \def but
% \newcommand{}{}
%
\newcommand{\red}[1]{\textcolor{red}{#1}}
% Insert the name of "your journal" with
% \journalname{myjournal}
%
\begin{document}
\title{Boundedness of operators generated by fractional semigroups associated with Schr\"odinger operators on Campanato type spaces via $T1$ theorem
\thanks{P. Li was financially supported by the National Natural Science Foundation of China (No. 12071272) and Shandong Natural Science Foundation of China (Nos. ZR2020MA004, ZR2017JL008). C. Zhang was supported by the National Natural Science Foundation of China (No. 11971431), the Zhejiang Provincial Natural Science Foundation of China(Grant No. LY18A010006) and
the first Class Discipline of Zhejiang-A(Zhejiang Gongshang University-Statistics).}
\thanks{Pengtao Li is the corresponding author.}
}
\subtitle{}
\titlerunning{Boundedness of operators on Campanato type spaces} % if too long for running head
\author{Zhiyong Wang \and Pengtao Li* \and Chao Zhang }
\authorrunning{Z. Wang \and P. Li\and C. Zhang}
\institute{Zhiyong Wang \at
School of Mathematics and Statistics, Qingdao University, Qingdao, 266071, China \\
\email{zywang199703@163.com} % \\
% \emph{Present address:} of F. Author % if needed
\and
Pengtao Li \at
School of Mathematics and Statistics, Qingdao University, Qingdao, 266071, China\\
\email{ptli@qdu.edu.cn}
\and
Chao Zhang \at
School of Statistics and Mathematics, Zhejiang Gongshang University, Hangzhou, 310018, China\\
\email{zaoyangzhangchao@163.com} }
\date{Received: date / Accepted: date}
% The correct dates will be entered by the editor
\maketitle
\begin{abstract}
Let $\mathcal{L}=-\Delta+V$ be a Schr\"{o}dinger operator, where the nonnegative potential $V$ belongs to the reverse H\"{o}lder class $B_{q}$.
By the aid of the subordinative formula, we estimate the regularities of the fractional heat semigroup, $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0},$
associated with $\mathcal{L}$. As an application, we obtain the $BMO^{\gamma}_{\mathcal{L}}$-boundedness of the maximal function, and the Littlewood-Paley $g$-functions associated with $\mathcal{L}$ via $T1$ theorem, respectively.
\end{abstract}
\keywords{Schr\"odinger ooperator\and fractional heat semigroups \and$T1$ theorem \and Campanato type space}
% \PACS{PACS code1 \and PACS code2 \and more}
\subclass{35J10 \and 42B20 \and 42B30}
\section{Introduction}
In the research of harmonic analysis and partial differential equations, the maximal operators and Littelwood-Paley $g$-functions paly an important role and were investigated by many mathematicians extensively.
For any integrable function $f$ on $\mathbb R^{n}$, the Hardy-Littlewood maximal operator is defined as
$$M(f)(x):=\sup_{Q}\frac{1}{|Q|}\int_{Q}|f(y)|dy,$$
where the supremum is taken over all cubes $Q\subset \mathbb R^{n}$. For $f\in BMO(\mathbb R^{n})$, Bennett-DeVore-Sharpley proved in \cite{BDS} that $M(f)$ is either infinite or belongs to $BMO(\mathbb R^{n})$. The boundedness result in \cite{BDS} can be extended to other maximal operators. For example, let $-\Delta$ be the Laplace operator: $\displaystyle \Delta=\sum^{n}_{i=1}\frac{\partial^{2}}{\partial x_{i}^{2}}$. Denote by $M_{\Delta}$ and $g$ the maximal operator and Littlewood-Paley $g$-function generated by the heat semigroup $\{e^{-t(-\Delta)}\}_{t>0}$, respectively, i.e.,
\begin{equation}\label{eq-1.1}
\left\{\begin{aligned}
\displaystyle &M_{\Delta}(f)(x):=\sup_{t>0}|e^{-t(-\Delta)}(f)(x)|;\\
&g(f)(x):=\Bigg(\int_{0}^{\infty}|e^{-t(-\Delta)}(f)(x)|^{2}\frac{dt}{t^{n+1}}\Bigg)^{1/2}.
\end{aligned}\right.
\end{equation}
Due to the mean value on ``large" cubes may be infinite, the $BMO(\mathbb R^{n})$-boundedness of $M_{\Delta}$ or $g$ holds if $M_{\Delta}(f)<\infty$ or $g(f)<\infty$ for $f\in BMO(\mathbb R^{n})$.
However, if the Laplacian $-\Delta$ is replaced by other second-order differential operators, the situation becomes different. Consider the Schr\"{o}dinger $\mathcal{L}=-\Delta+V$ in $\mathbb{R}^{n},\ \ n\geq 3,$
where $V$ is a nonnegative potential belonging to the reverse H\"{o}lder class $B_{q}$ for some $q>n/2$. Here a nonnegative potential $V$ is said to belong to $B_{q}$ if there exists $C>0$ such that for every ball $B$,
$$\Bigg(\frac{1}{|B|}\int_{B}V^{q}(x)dx\Bigg)^{1/q}\leq \frac{C}{|B|}\int_{B}V(x)dx.$$
In \cite{BDS}, the authors pointed out that for $f\in BMO(\mathbb R^{n})$ a supremum of averages of $f$ over ``large" cubes may be infinite, see \cite[page 610]{BDS}. In 2005, Dziuba\'{n}ski et.al. \cite{DGMTZ} proved the square functions associated with Schr\"odinger operators are bounded on the BMO type space $BMO_{\mathcal{L}}(\mathbb R^{n})$ related with $\mathcal{L}$ which is distinguished from the case of $BMO(\mathbb R^{n})$. See also \cite{LL} and \cite{wyx} for similar results in the setting of Heisenberg groups and stratified Lie groups.
Let $H=-\Delta+|x|^{2}$ be the harmonic oscillator. In \cite{bet}, Betancor et al. introduced a $T1$ criterion for Calder\'{o}n-Zygmund operators related to $H$ on the BMO type space $BMO_{H}(\mathbb{R}^{n})$. Later, Ma et al. in \cite{ma2} generalized the $T1$ criterion to the case of Campanato type spaces $BMO_{\mathcal{L}}^{\gamma}(\mathbb{R}^{n})$ related with $\mathcal{L}$. As applications, the authors in \cite{ma2} proved that the maximal operators associated with the heat
semigroup $\{e^{-t\mathcal{L}}\}_{t>0}$ and with the generalized Poisson operators
$\{P^{\sigma}_{t}\}_{t>0}(0<\sigma<1)$, the Littlewood-Paley g-functions given in terms of the heat and the Poisson semigroups are bounded on $BMO^{\gamma}_{\mathcal{L}}(\mathbb R^{n})$.
Notice that for $\sigma\in(0,1)$, the generalized Poisson operator
$\{P^{\sigma}_{t}\}_{t>0}$ is expressed as
\begin{equation}\label{eq-1.2}
P^{\sigma}_{t}(f)(x):=\frac{t^{2\sigma}}{4^{\sigma}\Gamma(\sigma)}\int^{\infty}_{0}e^{-\frac{t^{2}}{4r}}e^{-r\mathcal{L}}(f)\frac{dr}{r^{1+\sigma}}.
\end{equation}
Specially, for $\sigma=1/2$, $\{P^{1/2}_{t}\}_{t>0}$ is corresponding to the Poisson semigroup $\{e^{-t\mathcal{L}^{1/2}}\}_{t>0}$ associated with $\mathcal{L}$.
The main purpose of this paper is to derive the pointwise estimate and regularity properties of the fractional heat semigroup $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0}$, $\alpha>0$, to prove the boundedness of the maximal function and the Littlewood-Paley $g$-functions generated by $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0}$ on $BMO^{\gamma}_{\mathcal{L}}(\mathbb R^{n})$ via $T1$ theorem, respectively.
When $\mathcal{L}=-\Delta$, the kernels of the fractional heat semigroup $\{e^{-t(-\Delta)^{\alpha}}\}_{t>0}$ can be defined via the Fourier transform, i.e.,
\begin{equation}\label{1.4}
K_{\alpha, t}(x)=(2\pi)^{-n/2}\int_{\mathbb R^n}e^{ix\cdot \xi-t|\xi|^{2\alpha}}d\xi.
\end{equation}
For $\mathcal{L}=-\Delta+V$ with $V>0$, the kernels of fractional heat semigroups $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0}$, $\alpha\in(0,1)$, can not be defined via (\ref{1.4}).
However, for $\alpha>0$, the subordinative formula (cf.\cite{gri}) indicates that
\begin{equation}\label{p-1}
K^{\mathcal{L}}_{\alpha,t}(x,y)=\int^{\infty}_{0}\eta^{\alpha}_{t}(s)K^{\mathcal{L}}_{s}(x,y)ds,
\end{equation}
where $\eta^{\alpha}_{t}(\cdot)$ is a continuous function on $(0,\infty)$ satisfying (\ref{eq-2.1}) below.
In \cite{Li2}, the identity (\ref{p-1}) was applied to estimate $K^{\mathcal{L}}_{\alpha,t}(\cdot,\cdot)$ via the heat kernel $K^{\mathcal{L}}_{t}(\cdot,\cdot)$, see Proposition \ref{pro2.51}. Specially, for $\alpha=1/2$, the estimates of $K^{\mathcal{L}}_{\alpha,t}(\cdot,\cdot)$ goes back to those of the Poisson kernel $P_{t}^{\mathcal{L}}(\cdot,\cdot)$, see \cite[Lemma 3.9]{duo}.
We point out that, compared with the case of $\{P^{\sigma}_{t}\}_{t>0}$, some new regularity estimates should be introduced to prove the $BMO^{\gamma}_{\mathcal{L}}$-boundedness of the maximal function and Littlewood-Paley $g$-functions generated by $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0}$. Let $E=L^{\infty}((0,\infty), dt)$. It follows from (\ref{eq-1.2}) and the Minkowski integral inequality that
$$\|P^{\sigma}_{t}(f)\|_{E}\leq C_{\sigma}\int^{\infty}_{0}t^{2\sigma}e^{-\frac{t^{2}}{4r}}\|e^{-r\mathcal{L}}(f)\|_{E}\frac{dr}{r^{1+\sigma}}.$$
The fact that
$$\int^{\infty}_{0}t^{2\sigma}e^{-\frac{t^{2}}{4r}}\frac{dr}{r^{1+\sigma}}<\infty$$
ensures that the $BMO^{\gamma}_{\mathcal{L}}$-boundedness of the maximal function
$\sup_{t>0}|P^{\sigma}_{t}f(x)|$
can be deduced from that of the heat maximal function $\sup_{t>0}|e^{-t\mathcal{L}}f(x)|$, see \cite[Proposition 4.7]{ma2}. However, we can see from the identity (\ref{p-1}) that this method is not applicable to the case $\{e^{-t\mathcal{L}^{\alpha}}\}_{t>0}$.
In this paper, we get the following results:
\begin{itemize}
\item [$\bullet$]
In Section \ref{sec-2.1}, let $\displaystyle \nabla_{x}=\Big({\partial}/{\partial x_{1}},{\partial}/{\partial x_{2}},\ldots,{\partial}/{\partial x_{n}}\Big)$. By the perturbation theory for semigroups of operators, we deduce the pointwise estimates and the H\"older type estimates of the kernels:
$$\left\{\begin{aligned}
&\Big|\nabla_{x}(K_{t}(x-y)-K^{\mathcal{L}}_{t}(x,y))\Big|,\\
&\Big|t^{m}\partial^{m}_{t}(K_{t}(x-y)-K^{\mathcal{L}}_{t}(x,y))\Big|,\\
\end{aligned}\right.$$
see Lemmas \ref{lem1-1}, \ref{lem2-2} and Proposition \ref{pro2.5}, respectively.
\item [$\bullet$]In Section \ref{sec-2.2}, we use (\ref{1.4}) to obtain the corresponding estimates for
$$\left\{\begin{aligned}
&\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big|,\\
&\Big|t^{1/2\alpha}\nabla_{x}(K_{\alpha,t}(x-y)-K^{\mathcal{L}}_{\alpha, t}(x,y))\Big|,\\
&\Big|t^{m}\partial^{m}_{t}(K_{\alpha,t}(x-y)-K^{\mathcal{L}}_{\alpha, t}(x,y))\Big|,\\
\end{aligned}\right.$$
see Propositions \ref{pro1}--\ref{pro2-2}, respectively.
\item [$\bullet$]
In Section \ref{sec-4}, as applications of the regularity estimates obtained in Section \ref{sec-3}, we use the $T1$ criterion established in \cite{ma2} to prove the boundedness on Campanato type spaces $BMO^{\gamma}_{\mathcal{L}}(\mathbb R^{n})$ of the following maximal operator and $g$-functions:
$$\left\{
\begin{aligned}
&M^{\alpha}_{\mathcal{L}}f(x):=\sup_{t>0}|e^{-t\mathcal{L}^{\alpha}}f(x)|;\\
&g^{\mathcal{L}}_{\alpha}(f)(x):=\Bigg(\int^{\infty}_{0}|D^{\mathcal{L},m}_{\alpha,t}(f)(x)|^{2}\frac{dt}{t}\Bigg)^{1/2};\\
& \widetilde{g}^{\mathcal{L}}_{\alpha}f(x):=\Bigg(\int^{\infty}_{0}|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(f)(x)|^{2}\frac{dt}{t}\Bigg)^{1/2},
\end{aligned}
\right.$$
see Theorems \ref{the3}--\ref{the3.2}, respectively, where $D^{\mathcal{L},m}_{\alpha,t}$ and $\widetilde{D}^{\mathcal{L}}_{\alpha,t}$ are the operators with the integral kernels
\begin{equation}\label{eq-1.3}
\begin{cases}
&D^{\mathcal{L},m}_{\alpha,t}(x,y):=t^{m}\partial_{t}^{m}K^{\mathcal{L}}_{\alpha,t}(x,y),\\
&\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y):=t^{1/2\alpha}\nabla_{x}K^{\mathcal{L}}_{\alpha,t}(x,y),
\end{cases}
\end{equation}
respectively.
\end{itemize}
{\it Notations.} We will use $c$ and $C$ to denote the positive constants, which are independent of main parameters and may be different at each occurrence. By $B_{1}\sim B_{2}$, we mean that there exists a constant $C>1$ such that $C^{-1}\leq B_{1}/B_{2}\leq C.$
\section{Preliminaries}\label{sec-2}
\subsection{Schr\"odinger operators and function spaces} In this paper, let $ \delta_{0}=2- n/q$. At first, we list some properties of the potential $V$ which will be used in the sequel.
\begin{lemma}\label{lem2.4}{\rm (\cite[Lemma1.2]{she})}
\item{\rm (i)} For $0<r<R<\infty$,
$$\frac{1}{r^{n-2}}\int_{B(x,r)}V(y)dy\leq C\Big(\frac{r}{R}\Big)^{\delta}\frac{1}{R^{n-2}}\int_{B(x,R)}V(y)dy.$$
\item{\rm (ii)}
$r^{2-n}\int_{B(x,r)}V(y)dy=1$ if $r=\rho(x)$.
$r\sim \rho(x)$ if and only if
$r^{2-n}\int_{B(x,r)}V(y)dy\sim 1.$
\end{lemma}
\begin{lemma}\label{lem2.5}{\rm (\cite[Lemma 1.4]{she})}
\item{\rm (i)} There exist $C>0$ and $k_{0}\geq1$ such that for all $x,y\in \mathbb{R}^{n}$,
$$C^{-1}\rho(x)\Big(1+|x-y|/\rho(x)\Big)^{-k_{0}}\leq \rho(y)\leq C\rho(x)\Big(1+|x-y|/\rho(x)\Big)^
{k_{0}/(1+k_{0})}.$$
In particular, $\rho(y)\sim \rho(x)$ if $|x-y|<C\rho(x)$.
\item{\rm (ii)} There exists $l_{0}>1$ such that
$$\int_{B(x,R)}\frac{V(y)}{|x-y|^{n-2}}dy\leq \frac{C}{R^{n-2}}\int_{B(x,R)}V(y)dy\leq C\Big(1+
\frac{R}{\rho(x)}\Big)^{l_{0}}.$$
\end{lemma}
\begin{lemma}\label{lem2.6}{\rm (\cite[Corollary 2.8]{DZ2})}
For every nonnegative Schwarz function $\omega$, there exist $\delta>0$ and $C>0$ such that
\begin{equation}\notag
\int_{\mathbb{R}^{n}}t^{-n/2}\omega(|x-y|/\sqrt{t})V(y)dy \leq
\begin{cases}
Ct^{-1}(\sqrt{t}/\rho(x))^{\delta}, & t< \rho(x)^{2}; \\
Ct^{-1}(\sqrt{t}/\rho(x))^{l_{0}}, & t\geq{\rho(x)^{2}},
\end{cases}
\end{equation}
where $l_{0}$ is the constant given in Lemma \ref{lem2.5}.
\end{lemma}
It is well known that the classical Hardy space $H^{1}(\mathbb{R}^{n})$ can be defined via the maximal function $\sup_{t>0}|e^{-t(-\Delta)}f(x)|$ (cf. \cite{15}). In this sense, we can say that the Hardy
space $H^{1}(\mathbb{R}^{n})$ is the Hardy space associated with $-\Delta$. Since 1990s, the theory of Hardy spaces associated with operators on $\mathbb R^{n}$ has been investigated extensively. In \cite{DZ3}, Dziuba\'{n}ski and Zienkiewicz introduced the Hardy space $H^{1}_{\mathcal{L}}(\mathbb R^{n})$ related to Schr\"odinger operators $\mathcal{L}$ and obtained the atomic characterization and the Riesz transform characterization of $H^{1}_{\mathcal{L}}(\mathbb R^{n})$ via local Hardy spaces. By the aid of Campanato type spaces, the spaces $H^{p}_{\mathcal{L}}(\mathbb{R}^{n})(0<p\le 1)$ were introduced by Dziuba\'{n}ski and Zienkiewicz \cite{DZ2}. In recent years, the results of \cite{DZ3,DZ2} have been extended to other second-ordered differential operators, and various function spaces associated to operators have been established. For further information, we refer the reader to \cite{BDY,wyx,jia,yang} and the references therein.
For a Schr\"{o}dinger operator $\mathcal{L}$, let $\{e^{-t\mathcal{L}}\}_{t>0}$ be the heat semigroup generated by $\mathcal{L}$ and denote by $K^{\mathcal{L}}_{t}(\cdot,\cdot)$ the integral kernels of $e^{-t\mathcal{L}}$. Because the potential $V\geq 0$, the Feynman-Kac formula implies that
$$0\leq K^{\mathcal L}_{t}(x,y)\leq K_{t}(x-y):=(4\pi t)^{-n/2}e^{-\frac{|x-y|^{2}}{4t}}.$$
The Hardy type spaces $H^{p}_{\mathcal{L}}(\mathbb{R}^{n})$, $0<p\leq 1$, are defined as follows (cf. \cite{DZ2}):
\begin{definition}
For $0<p\leq 1$, the Hardy type space $H^{p}_{\mathcal{L}}(\mathbb{R}^{n})$ is defined as the completion of the space of compactly supported $L^{1}(\mathbb R^{n})$)-functions such that the maximal function
$$M_{\mathcal{L}}(f)(x):=\sup_{t>0}|e^{-t\mathcal{L}}(f)(x)|$$
belongs to $L^{p}(\mathbb{R}^{n})$. The quasi-norm in $H^{p}_{\mathcal{L}}(\mathbb{R}^{n})$ is defined as
$\|f\|_{H^{p}_{\mathcal{L}}}:=\|M_{\mathcal{L}}(f)\|_{L^{p}}.$
\end{definition}
Let $f$ be a locally integrable function on $\mathbb{R}^{n}$ and $B=B(x,r)$ be a ball. Denote by $f_{B}$ the mean of $f$ on $B$, i.e., $f_{B}=|B|^{-1}\int_{B}f(y)dy$. Let
\begin{equation*}
f(B,V):=\begin{cases}
f_{B}, & r<\rho(x);\\
0, & r\geq\rho(x),
\end{cases}
\end{equation*}
where the auxiliary function $\rho(\cdot)$ is defined as
$$\rho(x):=\sup\Bigg\{r>0:\frac{1}{r^{n-2}}\int_{B(x,y)}V(y)dy\leq 1\Bigg\}.$$
\begin{definition}\label{def1.1}
Let $0<\gamma\leq 1$. The Campanato type space $BMO^{\gamma}_{\mathcal{L}}(\mathbb{R}^{n})$ is defined as the set of all locally integrable functions $f$ satisfying
$$\|f\|_{BMO^{\gamma}_{\mathcal{L}}}:=\sup_{B\subset\mathbb{R}^{n}}\Bigg\{\frac{1}{|B|^{1+\gamma/n}}\int_{B}|f(x)-f(B,V)|dx\Bigg\}<\infty.$$
\end{definition}
The dual space of $H^{n/(n+\gamma)}_{\mathcal{L}}(\mathbb{R}^{n})$, $0\leq\gamma< 1$, is the Campanato type space $BMO^{\gamma}_{\mathcal{L}}(\mathbb{R}^{n})$ (cf. \cite[Theorem 4.5]{ma1}).
\subsection{The $T1$ criterion on Campanato type spaces}We denote by $L^{p}_{c}(\mathbb{R}^{n})$ the set of functions $f\in L^{p}(\mathbb{R}^{n})$, $1\leq p\leq\infty$, whose support $\text{ supp }(f)$ is a compact subset of $\mathbb{R}^{n}$.
\begin{definition}
Let $0\leq \beta<n$, $1<p\leq q<\infty$ with $1/q=1/p-\beta/n$. Let $T$ be a bounded linear operator from $L^{p}(\mathbb{R}^{n})$ into $L^{q}(\mathbb{R}^{n})$ such that
$$Tf(x)=\int_{\mathbb{R}^{n}}K(x,y)f(y)dy,\ \ \ f\in L^{p}_{c}(\mathbb{R}^{n})\ and\ a.e.\ x\notin \text{supp}(f).$$
We shall say that $T$ is a $\beta$-Schr\"{o}dinger-Calder\'{o}n-Zygmund operator with regularity exponent $\delta>0$ if there exists a constant $C>0$ such that
\begin{itemize}
\item[(i)] $\displaystyle |K(x,y)|\leq \frac{C}{|x-y|^{n-\beta}}\Big(1+\frac{|x-y|}{\rho(x)}\Big)^{-N}$ for all $N>0$ and $x\neq y$;
\item[(ii)] $\displaystyle |K(x,y)-K(x,z)|+|K(y,x)-K(z,x)|\leq C\frac{|y-z|^{\delta}}{|x-y|^{n-\beta+\delta}}$ when $|x-y|>2|y-z|$.
\end{itemize}
\end{definition}
The following $T1$ type criterions on Campanato type spaces were established by Ma et al. \cite{ma2}.
\begin{theorem}\label{the1}{\rm (\cite[Theorem1.1]{ma2})}
Let $T$ be a $\beta$-Schr\"{o}dinger-Calder\'{o}n-Zygmund operator, $\beta\geq 0$, $0< \beta+\gamma<\min\{1,\delta\}$, with smoothness exponent $\delta$. Then $T$ is a bounded operator from $BMO^{\gamma}_{\mathcal{L}}(\mathbb R^{n})$ into $BMO^{\gamma+\beta}_{\mathcal{L}}(\mathbb R^{n})$ if and only if there exists a constant
$C$ such that
$$\Big(\frac{\rho(x)}{r}\Big)^{\gamma}\frac{1}{|B|^{1+\beta/n}}\int_{B}|T1(y)-(T1)_{B}|dy\leq C$$
for every ball $B(x,r)$, $x\in\mathbb{R}^{n}$ and $0<r\leq \rho(x)/2$.
\end{theorem}
When $\gamma=0$, the authors in \cite{ma2} also proved
\begin{theorem}\label{the2}{\rm (\cite[Theorem1.2]{ma2})}
Let $T$ be a $\beta$-Schr\"{o}dinger-Calder\'{o}n-Zygmund operator, $0\leq \beta<\min\{1,\delta\}$, with smoothness exponent $\delta$. Then $T$ is a bounded operator from $BMO_{\mathcal{L}}(\mathbb R^{n})$ into $BMO^{\beta}_{\mathcal{L}}(\mathbb R^{n})$ if and only if there exists a constant
$C$ such that
$$\log\Big(\frac{\rho(x)}{r}\Big)\frac{1}{|B|^{1+\beta/n}}\int_{B}|T1(y)-(T1)_{B}|dy\leq C$$
for every ball $B(x,r)$, $x\in\mathbb{R}^{n}$ and $0<r\leq \rho(x)/2$.
\end{theorem}
\begin{lemma}\label{lem4.1}{\rm (\cite[Remark 4.1]{ma2})}
Theorems \ref{the1}\ and \ \ref{the2} can be also stated in a vector-valued setting. If $Tf$ takes values in a Banach space $\mathbb{B}$ and the absolute values in the conditions are replaced by the norm in $\mathbb{B}$, then both results hold.
\end{lemma}
\section{Regularity estimates}\label{sec-3}
\subsection{Regularities of heat kernels}\label{sec-2.1} By the fundamental solution of Schr\"{o}dinger operators, Dziuba\'{n}ski and Zienkiewicz proved that the heat kernel $K^{\mathcal{L}}_{t}(\cdot,\cdot)$ satisfies the following estimates.
\begin{lemma}\label{pro2.2}
\item{\rm (i)} {\rm (\cite[Theorem 2.11]{DZ3})} For any $N>0$, there exist constants $C_{N},c>0$ such that
$$|K^{\mathcal{L}}_{t}(x,y)|\leq C_{N}t^{-n/2}e^{-c|x-y|^{2}/t}\Big(1+\frac{\sqrt{t}}{\rho(x)}+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii)} {\rm (\cite[Theorem \ 4.11]{DZ1})} Assume that $0<\delta\leq \min\{1,\delta_{0}\}.$ For any $N>0$, there exist constants $C_{N},c>0$ such that for all $|h|<\sqrt{t}$,
$$\Big|K^{\mathcal{L}}_{t}(x+h,y)-K^{\mathcal{L}}_{t}(x,y)\Big|\leq C_{N}\Big(\frac{|h|}{\sqrt{t}}\Big)^{\delta}t^{-n/2}e^{-c|x-y|^{2}/t}\Big(1+\frac{\sqrt{t}}{\rho(x)}+
\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}.$$
\end{lemma}
\begin{lemma}\label{lem1}{\rm (\cite[Proposition 2.16]{DZ2})}
There exist constants $C,c>0$ such that for $x,y\in\mathbb{R}^{n}$ and $t>0,$
$$\Big|K^{\mathcal{L}}_{t}(x,y)-K_{t}(x-y)\Big|\leq C \Big(\frac{\sqrt{t}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/2}e^{-c|x-y|^{2}/t}.$$
\end{lemma}
In \cite{duo}, under the assumption that $V\in B_{q}, q>n$, X. Duong, L. Yan and C. Zhang obtained the following regularity estimate for the kernel $K^{\mathcal{L}}_{t}(\cdot,\cdot)$.
\begin{lemma}\label{le3.8}{\rm (\cite[Lemma 3.8]{duo})}
Suppose that $V\in B_{q}$ for some $q>n$. For any $N>0$, there exist constants $C >0$ and $c>0$ such that
for all $x,y\in\mathbb R^n$ and $t>0,$
\begin{eqnarray}\label{e3.11}
| \nabla_x { K}^{\mathcal{L}}_t(x,y)|
\leq C t^{-(n+1)/2}e^{-c{{|x-y|}^2}/{ t}}\left(1+\frac{\sqrt{t}}{\rho(x)}
+\frac{\sqrt{t}}{\rho(y)}\right)^{-N}.
\end{eqnarray}
\end{lemma}
By the perturbation theory for semigroups of operators,
\begin{eqnarray}\label{p-2}
% \nonumber to remove numbering (before each equation)
K_{t}(x-y)-K_{t}^{\mathcal{L}}(x,y) &=& \int^{t}_{0}\int_{\mathbb{R}^{n}}K_{t-s}(w-x)V(w)K^{\mathcal{L}}_{s}(w,y)dwds \\
&=& \int^{t/2}_{0}\int_{\mathbb{R}^{n}}K_{t-s}(w-x)V(w)K^{\mathcal{L}}_{s}(w,y)dwds\nonumber \\
&&+ \int^{t/2}_{0}\int_{\mathbb{R}^{n}}K_{s}(w-x)V(w)K^{\mathcal{L}}_{t-s}(w,y)dwds.\nonumber
\end{eqnarray}
Similar to \cite[Proposition 2.16]{DZ2}, we can prove the following lemma.
\begin{lemma}\label{lem1-1}
Suppose that $V\in B_{q}$ for some $q>n$. There exist constants $C,c>0$ such that
$$\Big|\nabla_{x}K_{t}(x-y)-\nabla_{x}K^{\mathcal{L}}_{t}(x,y)\Big|\leq Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\min\Bigg\{\Big(\frac{\sqrt{t}}{\rho(x)}\Big)^{\delta_{0}},\
\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}\Bigg\}.$$
\end{lemma}
\begin{proof}
If $t\geq\rho(y)^{2}$, it is easy to see that
$$\Big|\nabla_{x}K_{t}(x-y)-\nabla_{x}K^{\mathcal{L}}_{t}(x,y)\Big|\leq Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}.$$
If $t<\rho(y)^{2}$, by (\ref{p-2}), we get
\begin{eqnarray}\notag
% \nonumber to remove numbering (before each equation)
\Big|\nabla_{x}K_{t}(x-y)-\nabla_{x}K^{\mathcal{L}}_{t}(x,y)\Big| &=& \Big|\int^{t}_{0}\int_{\mathbb{R}^{n}}\nabla_{x}K_{t-s}(x-w)V(w)K^{\mathcal{L}}_{s}(w,y)dwds\Big| \\ \label{12}
&\leq&I_{1}+I_{2},
\end{eqnarray}
where
\begin{equation*}
\left\{ \begin{aligned}
I_{1}&:= \int^{t/2}_{0}\int_{\mathbb{R}^{n}}|\nabla_{x}K_{t-s}(x-w)|V(w)K^{\mathcal{L}}_{s}(w,y)dwds;\\
I_{2}&:=\int^{t/2}_{0}\int_{\mathbb{R}^{n}}|\nabla_{x}K_{s}(x-w)|V(w)K^{\mathcal{L}}_{t-s}(w,y)dwds.
\end{aligned}\right.
\end{equation*}
For $I_{1}$, using Lemmas \ref{le3.8} and \ref{lem2.6}, we get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
I_{1} &=&\int^{t/2}_{0}\int_{|w-y|<|x-y|/2}|\nabla_{x}K_{t-s}(x-w)|V(w)K^{\mathcal{L}}_{s}(w,y)dwds\\
&&+\int^{t/2}_{0}\int_{|w-y|\geq|x-y|/2}|\nabla_{x}K_{t-s}(x-w)|V(w)K^{\mathcal{L}}_{s}(w,y)dwds \\
&\leq& Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\int^{t/2}_{0}\int_{|w-y|<|x-y|/2}V(w)s^{-n/2}e^{-c|w-y|^{2}/s}dwds\\
&&+ Ct^{-(n+1)/2}\int^{t/2}_{0}\int_{|w-y|\geq|x-y|/2}V(w)s^{-n/2}e^{-c(|w-y|^{2}+|x-y|^{2})/s}dwds \\
&\leq& Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\int^{t/2}_{0}\frac{1}{s}\Big(\frac{\sqrt{s}}{\rho(y)}\Big)^{\delta_{0}}ds \\
&=& Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}.
\end{eqnarray*}
Similar to $I_{1}$, for the term $I_{2}$, we can obtain
$$I_{2}\leq Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\Big(\frac{\sqrt{t}}{\rho(x)}\Big)^{\delta_{0}}.$$
It follows from (i) of Lemma \ref{lem2.5} that
$$\frac{\sqrt{t}}{\rho(x)}\leq C\Bigg(1+\frac{|x-y|}{\sqrt{t}}\frac{\sqrt{t}}{\rho(y)}\Bigg)^{l_{0}}\frac{\sqrt{t}}{\rho(y)}\leq
C_{\varepsilon}e^{\varepsilon|x-y|^{2}/t}\frac{\sqrt{t}}{\rho(y)},$$
where $\varepsilon>0$ is an arbitrary small constant. Hence
$$I_{2}\leq Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}.$$
\end{proof}
\begin{lemma}\label{lem2}{\rm (\cite[Proposition 2.17]{DZ2})}
Let $0<\delta<\min\{1,\delta_{0}\}$. For $|y-z|<\min\{C\rho(x),|x-y|/4\},$ there exist constants $C,c>0$ such that
$$\Big|\Big(K^{\mathcal{L}}_{t}(x,y)-K_{t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{t}(x,z)-K_{t}(x-z)\Big)\Big|\leq C\Big(\frac{|y-z|}{\rho(y)}\Big)^{\delta}t^{-n/2}e^{-c|x-y|^{2}/t}.$$
\end{lemma}
\begin{lemma}\label{lem2-2}
Suppose that $V\in B_{q}$ for some $q>n$. Let $\delta_{1}=1-n/q$. If $|u|\leq \min\{|x-y|/4, C\rho(x)\}$, there exist constants $C,c>0$ such that for $0<\delta'<\delta_{1}$,
\begin{multline*}
\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x,y)-\nabla_{x}K_{t}(x-y)\Big)-\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)-\nabla_{x}K_{t}(x+u-y)\Big)\Big|\\
\quad\leq Ct^{-(n+1)/2}e^{-c|x-y|^{2}/t}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta'}.
\end{multline*}
\end{lemma}
\begin{proof}
We prove this lemma by the same argument as Lemma \ref{lem1-1}. It is enough to verify that
\begin{eqnarray}\label{13}
&&\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x,y)-\nabla_{x}K_{t}(x-y)\Big)-\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)-\nabla_{x}K_{t}(x+u-y)\Big)\Big|\\
&&\quad\leq C_{\epsilon}t^{-(n+1)/2}e^{\epsilon|x-y|^{2}/t}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}},\nonumber
\end{eqnarray}
where $\epsilon>0$ is an arbitrary small constant. In fact, under the condition $|u|<|x-y|/4$, it is easy to see that $|x-y|\sim |x+u-y|$. We can deduce from Lemma \ref{le3.8} that
\begin{eqnarray}\label{eq-10}
&&\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x,y)-\nabla_{x}K_{t}(x-y)\Big)-\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)-\nabla_{x}K_{t}(x+u-y)\Big)\Big|\\
&&\quad\leq \Big|\nabla_{x}K^{\mathcal{L}}_{t}(x,y)\Big|+\Big|\nabla_{x}K_{t}(x-y)\Big|+\Big|\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)\Big|+\Big|\nabla_{x}K_{t}(x+u-y)\Big|\nonumber\\
&&\quad\leq C t^{-(n+1)/2}e^{-c{{|x-y|}^2}/t}.\nonumber
\end{eqnarray}
Then, for $\delta'\in(0,\delta_{1})$, it follows from (\ref{13})\ and (\ref{eq-10}) that
\begin{eqnarray*}
&&\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x,y)-\nabla_{x}K_{t}(x-y)\Big)-\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)-\nabla_{x}K_{t}(x+u-y)\Big)\Big|\\
&&\quad\leq C \Bigg\{t^{-(n+1)/2}e^{\epsilon|x-y|^{2}/t}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Bigg\}^{\delta'/\delta_{1}}
\Bigg\{t^{-(n+1)/2}e^{-c{{|x-y|}^2}/t}\Bigg\}^{1-\delta'/\delta_{1}},
\end{eqnarray*}
which gives the desired estimate.
Now we prove (\ref{13}). Since the case for $|u|\geq\rho(y)$ is trivial, so we may assume $|u|<\rho(y)$. If $t\leq 2|u|^{2}$, the required estimate follows from Lemma \ref{lem1-1}. Hence we consider the case $t>2|u|^{2}$ only. Recall that for the classical heat kernel $K_{t}(\cdot)$, it holds
$$\Big|\nabla^{2}_{x}K_{t}(x)\Big|\leq Ct^{-n/2-1}e^{-c|x|^{2}/t}.$$
A direct computation gives
\begin{equation}\label{14}
\Big|\nabla_{x}K_{t}(x+u)-\nabla_{x}K_{t}(x)\Big|\leq C|u|t^{-n/2-1},
\end{equation}
and for $|u|\leq |x|/2$,
\begin{equation}\label{15}
\Big|\nabla_{x}K_{t}(x+u)-\nabla_{x}K_{t}(x)\Big|\leq C|u|t^{-n/2-1}e^{-c|x|^{2}/t}.
\end{equation}
Similar to (\ref{12}), we split $$\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x,y)-\nabla_{x}K_{t}(x-y)\Big)-\Big(\nabla_{x}K^{\mathcal{L}}_{t}(x+u,y)-\nabla_{x}K_{t}(x+u-y)\Big)\Big|\leq J_{1}+J_{2},$$ where
\begin{equation*}
\left\{ \begin{aligned}
J_{1}&:=\int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big|\nabla_{x}K_{t-s}(w-(x+u))-\nabla_{x}K_{t-s}(w-x)\Big|V(w)K^{\mathcal{L}}_{s}(w,y)dwds;\\
J_{2}&:=\int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big|\nabla_{x}K_{s}(w-(x+u))-\nabla_{x}K_{s}(w-x)\Big|V(w)K^{\mathcal{L}}_{t-s}(w,y)dwds.
\end{aligned}\right.
\end{equation*}
For $J_{1}$, if $t<2\rho(y)^{2}$, using Lemma \ref{lem2.6} and (\ref{14}), we get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
J_{1} &\leq& C|u|t^{-n/2-1}\int^{t/2}_{0}\int_{\mathbb{R}^{n}}V(w)s^{-n/2}e^{-c|y-w|^{2}/s}dwds \\
&\leq& C|u|t^{-n/2-1}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}} \\
&\leq& Ct^{-(n+1)/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}.
\end{eqnarray*}
If $t\geq 2\rho(y)^{2}$, applying Lemmas \ref{lem2.6} and \ref{pro2.2}, we have
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
J_{1} &\leq&C|u|t^{-n/2-1}\int^{t/2}_{0}\int_{\mathbb{R}^{n}}V(w)s^{-n/2}e^{-c|y-w|^{2}/s}\Big(1+\frac{\sqrt{s}}{\rho(y)}\Big)^{-N}dwds \\
&\leq&C|u|t^{-n/2-1}\Bigg\{\int^{\rho(y)^{2}}_{0}\frac{1}{s}\Big(\frac{\sqrt{s}}{\rho(y)}\Big)^{\delta_{0}}ds+\int^{t/2}_{\rho(y)^{2}}
\frac{1}{s}\Big(\frac{\sqrt{s}}{\rho(y)}\Big)^{l_{0}-N}ds\Bigg\} \\
&\leq&C|u|t^{-n/2-1} \\
&\leq& Ct^{-(n+1)/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}},
\end{eqnarray*}
where $N$ is chosen large enough satisfying $N>l_{0}$.
To estimate $J_{2}$, we use Lemma \ref{pro2.2} and write $J_{2}\leq C(J_{2,1}+J_{2,2}+J_{2,3})$, where
\begin{equation*}
\left\{ \begin{aligned}
J_{2,1}&:=t^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{|u|^{2}}_{0}\int_{\mathbb{R}^{n}}\Big|\nabla_{x}K_{s}(w-(x+u))-\nabla_{x}K_{s}
(w-x)\Big|V(w)dwds;\\
J_{2,2}&:=t^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{|u|^{2}}\int_{|w-x|<2|u|}\Big|\nabla_{x}K_{s}(w-(x+u))-\nabla_{x}K_{s}
(w-x)\Big|V(w)dwds;\\
J_{2,3}&:=t^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{|u|^{2}}\int_{|w-x|\geq2|u|}\Big|\nabla_{x}K_{s}(w-(x+u))-\nabla_{x}K_{s}
(w-x)\Big|V(w)dwds.
\end{aligned}\right.
\end{equation*}
Notice that $\rho(x+u)\sim\rho(x)$ as $|u|\leq \rho(x)$. It holds
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
J_{2,1} &\leq& Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{|u|^{2}}_{0}\frac{1}{s^{3/2}}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}ds \\
&\leq&Ct^{-n/2}|u|^{-1}\Big(\frac{|u|}{\rho(y)}\Big)^{-1}\Big(\frac{\sqrt{t}}{|u|}\Big)^{-1}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}
\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{0}}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}} \\
&\leq& Ct^{-(n+1)/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}},
\end{eqnarray*}
where in the last inequality we have used the fact that $\delta_{0}=2-n/q$. By Lemmas \ref{lem2.4}\ and \ \ref{lem2.5}, we apply (\ref{14}) to get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
J_{2,2} &\leq& Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{|u|^{2}}\int_{|w-x|<2|u|}|u|s^{-n/2-1}V(w)dwds \\
&\leq& Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{|u|^{2}}|u|^{n-1}s^{-n/2-1}\Big(\frac{|u|}{\rho(x)}\Big)^{\delta_{0}}ds \\
&\leq& Ct^{-n/2}|u|^{-1}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{0}}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}} \\
&\leq& Ct^{-(n+1)/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}}.
\end{eqnarray*}
For $J_{2,3}$, if $t\leq 2\rho(x)^{2}$, it can be deduced from Lemma \ref{lem2.6} and (\ref{15}) that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
J_{2,3} &\leq&Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{|u|^{2}}\int_{|w-x|\geq 2|u|}|u|s^{-n/2-1}e^{-c|w-x|^{2}/s}V(w)dwds \\
&\leq&Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}|u|\int^{t/2}_{|u|^{2}}\frac{1}{s^{2}}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}ds \\
&\leq&Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}|u|^{-1}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{0}}
\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}} \\
&\leq& Ct^{-(n+1)/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}}.
\end{eqnarray*}
If $t>2\rho(x)^{2}$, then
\begin{eqnarray}\notag
% \nonumber to remove numbering (before each equation)
J_{2,3} &\leq& Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{\rho(x)^{2}}_{|u|^{2}}\int_{|w-x|\geq 2|u|}\Big|\nabla_{x}K_{s}(w-(x+u))-\nabla_{x}K_{s}(w-x)\Big|V(w)dwds \\ \notag
&&+ Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{\rho(x)^{2}}\int_{|w-x|\geq 2|u|}|u|s^{-n/2-1}e^{-c|w-x|^{2}/s}V(w)dwds\\ \notag
&\leq& Ct^{-(n+1)/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}}\\
\label{e-1}
&&+Ct^{-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int^{t/2}_{\rho(x)^{2}}\frac{|u|}{s}\frac{1}{s}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{l_{0}}
ds \\ \notag
&\leq& Ct^{-(n+1)/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta_{1}}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+1}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{\delta_{0}} \\ \notag
&&+Ct^{-(n+1)/2}\frac{|u|}{\rho(y)}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N-1+l_{0}}\Big(\frac{\rho(y)}{\rho(x)}\Big)^{l_{0}},
\end{eqnarray}
where in (\ref{e-1}) we have used the estimate obtained for $t\leq 2\rho(x)^{2}$ and Lemma \ref{lem2.6} for $s\geq \rho(x)^{2}$.
By Lemma \ref{lem2.5},
$$\frac{\rho(y)}{\rho(x)}\leq C\Big(1+\frac{|x-y|}{\sqrt{t}}\frac{\sqrt{t}}{\rho(y)}\Big)^{m_{0}}\leq C_{\epsilon}e^{\epsilon|x-y|^{2}/t}
\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{m_{0}},$$
where $\epsilon>0$ is an arbitrary small constant. Choosing $N$ large enough in the estimates of $J_{2,1},J_{2,2}$ and $J_{2,3}$, we obtain (\ref{13}) and hence Lemma \ref{lem2-2} is proved.
\end{proof}
We can obtain the following estimates, which generalize \cite[Lemmas 3.7\ and \ 3.8]{Zhang}. We also refer to \cite[(57)]{LL} for the case $m=1$ in the setting of Heisenberg groups.
\begin{proposition}\label{pro2.5}
\item{\rm (i)} There exist constants $C,c>0$ such that
$$\Big|t^{m}\partial_{t}^{m}K^{\mathcal{L}}_{t}(x,y)-t^{m}\partial_{t}^{m}K_{t}(x-y)\Big|\leq Ct^{-n/2}e^{-c|x-y|^{2}/t}\min\Bigg\{\Big(\frac{\sqrt{t}}{\rho(x)}\Big)^{\delta_{0}},
\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}\Bigg\}.$$
\item{\rm (ii)} For every $0<\delta<\min\{1,\delta_{0}\}$ and $|y-z|<\min\{\rho(x),|x-y|/4\}$, there exist constants $C,c>0$ such that
\begin{multline*}
\Bigg|\Big(t^{m}\partial_{t}^{m}K^{\mathcal{L}}_{t}(x,y)-t^{m}\partial_{t}^{m}K_{t}(x-y)\Big)
-\Big(t^{m}\partial_{t}^{m}K^{\mathcal{L}}_{t}(x,z)-t^{m}\partial_{t}^{m}K_{t}(x-z)\Big)\Bigg|\\ \leq C\Big(\frac{|y-z|}{\rho(y)}\Big)^{\delta}t^{-n/2}e^{-c|x-y|^{2}/t}.
\end{multline*}
\end{proposition}
\begin{proof}For $t>0$ and $m\in\mathbb{Z}_{+}$, define
$Q^{\mathcal{L}}_{t,m}(x,y):= t^{m}\partial_{t}^{m}K^{\mathcal{L}}_{t}(x,y)$
and $Q_{t,m}(x-y):= t^{m}\partial_{t}^{m}K_{t}(x-y).$
The proof of (i) is similar to \cite[Lemmas 4.10\ and \ 4.11]{wyx}, so we omit the details.
For (ii), by (\ref{p-2}), we get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big(K_{t}(x+u-y)-K^{\mathcal{L}}_{t}(x+u,y)\Big)-\Big(K_{t}(x-y)-K_{t}^{\mathcal{L}}(x,y)\Big)\\ &&= \int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big(K_{t-s}(w-(x+u))-K_{t-s}(w-x)\Big)V(w)K^{\mathcal{L}}_{s}(w,y)dwds\\
&&\quad + \int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big(K_{s}(w-(x+u))-K_{s}(w-x)\Big)V(w)K^{\mathcal{L}}_{t-s}(w,y)dwds.
\end{eqnarray*}
Similar to \cite[Proposition 4.8]{wyx}, we use a direct calculus to deduce that
\item{\rm (1)} If $m$ is even with $m\geq 2$, there exists a sequence of coefficients $\{C_{m,j}\}_{m\geq 2,2\leq j\leq m/2}$ such that
\begin{eqnarray}\label{eq-2.2}
% \nonumber to remove numbering (before each equation)
&&t^{m}\frac{d^{m}}{dt^{m}}\Big\{\Big(K_{t}(x+u-y)-K^{\mathcal{L}}_{t}(x+u,y)\Big)-\Big(K_{t}(x-y)-K_{t}^{\mathcal{L}}(x,y)\Big)\Big\}\nonumber\\
&&= \frac{m+1}{2}(E_{1}+E_{2})
+\sum^{m/2}_{j=2}(C_{m-1,j-1}+C_{m-1,j})\Big(E^{j}_{3,1}+E_{3,2}^{j}\Big)
+E_{4}+E_{5}.
\end{eqnarray}
\item{\rm (2)} If $m$ is odd with $m\geq 3$, there exists a sequence of coefficients $\{C_{m,j}\}_{m\geq 3,2\leq j\leq [m/2]}$ such that
\begin{eqnarray}\label{eq-2.3}
% \nonumber to remove numbering (before each equation)
&&t^{m}\frac{d^{m}}{dt^{m}}\Big\{\Big(K_{t}(x+u-y)-K^{\mathcal{L}}_{t}(x+u,y)\Big)-\Big(K_{t}(x-y)-K_{t}^{\mathcal{L}}(x,y)\Big)\Big\}\nonumber\\
&&=\frac{m+1}{2}(E_{1}+E_{2})+ E_{4}-E_{5}
+\sum^{[m/2]}_{j=2}(C_{m-1,j-1}+C_{m-1,j})\Big(E^{j}_{3,1}+E^{j}_{3,2}\Big)\nonumber\\
&&+2C_{m,[m/2]}\frac{d^{[m/2]}}{dt^{[m/2]}}\Big(K_{t/2}(w-(x+u))-K_{t/2}(w-x)\Big)V(w)\frac{d^{[m/2]}}{dt^{[m/2]}}K^{\mathcal{L}}_{t/2}(w,y).
\end{eqnarray}
Here in the above (\ref{eq-2.2})\ and \ (\ref{eq-2.3}),
\begin{equation*}
\left\{ \begin{aligned}
E_{1}:&=\int_{\mathbb{R}^{n}}t\Big\{t^{m-1}\frac{d^{m-1}}{dt^{m-1}}\Big(K_{t/2}(w-x-u)-K_{t/2}(w-x)\Big)\Big\}V(w)K^{\mathcal{L}}_{t/2}(w,y)dw;\\
E_{2}:&=\int_{\mathbb{R}^{n}}t\Big(K_{t/2}(w-x-u))-K_{t/2}(w-x)\Big)V(w)\Big(t^{m-1}\frac{d^{m-1}}{dt^{m-1}}K^{\mathcal{L}}_{t/2}(w,y)\Big)dw;\\
E_{3,1}^{j}:&=\int_{\mathbb{R}^{n}}
t\Big\{t^{m-j}\frac{d^{m-j}}{dt^{m-j}}\Big(K_{t/2}(w-x-u)-K_{t/2}(w-x)\Big)\Big\}V(w)\Big(t^{j-1}\frac{d^{j-1}}{dt^{j-1}}K^{\mathcal{L}}_{t/2}(w,y)\Big)dw;\\
E_{3,2}^{j}:&=\int_{\mathbb{R}^{n}}
t\Big\{t^{j-1}\frac{d^{j-1}}{dt^{j-1}}\Big(K_{t/2}(w-x-u)-K_{t/2}(w-x)\Big)\Big\}V(w)\Big(t^{m-j}\frac{d^{m-j}}{dt^{m-j}}K^{\mathcal{L}}_{t/2}(w,y)\Big)dw;\\
E_{4}:&=\int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big(t^{m}\frac{d^{m}}{dt^{m}}(K_{t-s}(w-x-u)-K_{t-s}(w-x)\Big)V(w)K^{\mathcal{L}}_{s}(w,y)dwds;\\
E_{5}:&=\int^{t/2}_{0}\int_{\mathbb{R}^{n}}\Big(K_{s}(w-x-u)-K_{s}(w,x)\Big)V(w)\Big(t^{m}\frac{d^{m}}{dt^{m}}K^{\mathcal{L}}_{t-s}(w,y)\Big)dwds.
\end{aligned}\right.
\end{equation*}
Below, for the sake of simplicity, we only estimate $E_{1}, E_{4}, E_{5}$. The estimations for $E_{2}$, $E_{3,1}^{j}$, $E^{j}_{3,2}$ are similar, and so we omit the details. By the mean value theorem, we know that there exist constants $C, c$ such that
\begin{equation}\label{24}
\Big|Q_{t,m}(x+u)-Q_{t,m}(x)\Big|\leq
\left\{\begin{aligned}
&C|u|t^{-(n+1)/2},&\quad \forall\ x, u \in\mathbb R^{n},\ t\in(0,\infty);\\
&C|u|t^{-(n+1)/2}e^{-c|x|^{2}/t},&\quad |u|\leq |x|/2,\ t>0.
\end{aligned}\right.
\end{equation}
We divide $E_{1}$ as $E_{1}\leq E_{1,1}+E_{1,2}$, where
\begin{equation*}
\left\{ \begin{aligned}
E_{1,1}&:=t\int_{|w-x|<2u}\Big|Q_{t/2,m-1}(w-(x+u))-Q_{t/2,m-1}(w-x)\Big|V(w)K^{\mathcal{L}}_{t/2}(w,y)dw;\\
E_{1,2}&:=t\int_{|w-x|\geq2u}\Big|Q_{t/2,m-1}(w-(x+u))-Q_{t/2,m-1}(w-x)\Big|V(w)K^{\mathcal{L}}_{t/2}(w,y)dw.
\end{aligned}\right.
\end{equation*}
If $t<2\rho(y)^{2}$, for $E_{1,1}$, By (\ref{24}), Lemma \ref{lem2.6} (i) and Lemma \ref{pro2.2}, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
E_{1,1} &\leq& Ct^{1-n/2}\int_{|w-x|<2|u|}|u|t^{-(n+1)/2}V(w)e^{-c|w-y|^{2}/t}dw \\
&\leq& Ct^{1-n/2}|u|\frac{1}{t^{3/2}}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}} \\
&=&Ct^{-n/2}\frac{|u|}{\sqrt{t}}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}}\leq Ct^{-n/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta}.
\end{eqnarray*}
Similar to $E_{1,1}$, by (\ref{24}) again, we get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
E_{1,2} &\leq&Ct^{1-n/2}\int_{|w-x|\geq 2|u|}|u|t^{-(n+1)/2}e^{-c|w-y|^{2}/t}V(w)dw \\
&\leq&Ct^{1-n/2}\frac{|u|}{t^{3/2}}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{\delta_{0}} \\
&\leq&Ct^{-n/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta}.
\end{eqnarray*}
If $t>2\rho(y)^{2}$, for $E_{1,1}$, by (\ref{24}), Lemma \ref{lem2.6} (ii) and Lemma \ref{pro2.2}, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
E_{1,1} &\leq& Ct^{1-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}\int_{|w-x|<2|u|}|u|t^{-(n+1)/2}V(w)e^{-c|w-y|^{2}/t}dw \\
&\leq& Ct^{1-n/2}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N}|u|\frac{1}{t^{3/2}}\Big(\frac{\sqrt{t}}{\rho(y)}\Big)^{l_{0}} \\
&\leq& Ct^{-n/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+l_{0}}.
\end{eqnarray*}
Similar to $E_{1,1}$, we can also choose $N$ large enough such that
$$E_{1,2}\leq Ct^{-n/2}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta}\Big(1+\frac{\sqrt{t}}{\rho(y)}\Big)^{-N+l_{0}}.$$
For $E_{4}$ and $E_{5}$, similar to \cite[Lemma 10]{LL}, by Lemma \ref{pro2.2} and (\ref{24}), we can obtain
\begin{equation}\label{23}
|E_{4}+E_{5}|\leq C_{\epsilon}t^{-n/2}e^{\epsilon|x-y|^{2}/t}\Big(\frac{|u|}{\rho(y)}\Big)^{\delta},
\end{equation}
where $\epsilon>0$ is an arbitrary small constant. Hence Proposition \ref{pro2.5} is proved.
\end{proof}
\subsection{Fractional heat kernels associated with $\mathcal{L}$}\label{sec-2.2}
In the following, we will derive some regularity estimates for the fractional heat kernels related with $\mathcal{L}$. For $\alpha\in(0,1)$, the fractional power of $\mathcal{L}$, denoted by $\mathcal{L}^{\alpha}$, is defined as
\begin{equation}\label{7}
\mathcal{L}^{\alpha}:=\frac{1}{\Gamma(-\alpha)}\int^{\infty}_{0}\Big(e^{-t\sqrt{\mathcal{L}}}f(x)-f(x)\Big)\frac{dt}{t^{1+2\alpha}}\quad \forall\ f\in L^{2}(\mathbb{R}^{n}).
\end{equation}
We use the subordinative formula to express the integral kernel $K^{\mathcal{L}}_{\alpha,t}(\cdot,\cdot)$ of $e^{-t\mathcal{L}^{\alpha}}$ as (cf. \cite{gri})
\begin{equation*}
K^{\mathcal{L}}_{\alpha,t}(x,y)=\int^{\infty}_{0}\eta^{\alpha}_{t}(s)K^{\mathcal{L}}_{s}(x,y)ds,
\end{equation*}
where $\eta^{\alpha}_{t}(\cdot)$ satisfies
\begin{equation}\label{eq-2.1}
\left\{\begin{aligned}
&\eta^{\alpha}_{t}(s)=1/t^{1/\alpha}\eta_{1}^{\alpha}(s/t^{1/\alpha});\\
&\eta^{\alpha}_{t}(s)\leq t/s^{1+\alpha}\quad \forall\ s,t>0;\\
&\int^{\infty}_{0}s^{-r}\eta^{\alpha}_{1}(s)ds<\infty, \ \ r>0;\\
&\eta^{\alpha}_{t}(s)\simeq t/s^{1+\alpha}\ \ \forall\ s\geq t^{1/\alpha}>0.
\end{aligned}\right.
\end{equation}
By the subordinative formula (\ref{p-1}) and Lemma \ref{pro2.2}, Li et al. in \cite{Li2} proved the following estimates for $K^{\mathcal{L}}_{\alpha,t}(\cdot,\cdot)$.
\begin{proposition}\label{pro2.51}{\rm \cite[Propositions 3.1\ and \ 3.2]{Li2}}
Let $0<\alpha<1$.
\item{\rm (i)} For any $N>0$, there exists a constant $C_{N}>0$ such that
$$\Big|K^{\mathcal L}_{\alpha,t}(x,y)\Big|\leq \frac{C_{N}t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii)} Let $0<\delta\leq \min\{1,\delta_{0}\}$. For any $N>0$, there exists a constant $C_{N}>0$ such that for all $|h|\leq \sqrt{t^{1/\alpha}}$,
$$\Big|K^{\mathcal L}_{\alpha,t}(x+h,y)-K^{\mathcal L}_{\alpha,t}(x,y)\Big|\leq C_{N}\Big(\frac{|h|}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}
\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x}+
\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\end{proposition}
For the kernels $\widetilde{D}^{\mathcal L}_{\alpha,t}(\cdot,\cdot)$ and $D^{\mathcal{L},m}_{\alpha,t}(\cdot,\cdot), m\in\mathbb{Z}_{+}, t>0$, defined by (\ref{eq-1.3}),
the following regularity estimates were obtained by Li et al. in \cite{Li2}.
\begin{proposition}\label{pro3.12}{\rm (\cite[Proposition 3.3]{Li2})} Let $0<\alpha<1$.
\item{\rm (i)} For every $N$, there is a constant $C_{N}>0$ such that
$$\Big|D_{\alpha,t}^{\mathcal{L},m}(x,y)\Big|\leq \frac{C_{N}t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha }}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii)} Let $0<\delta<\min\{2\alpha,\delta_{0},1\}$. For every $N>0$, there exists a constant $C_{N}>0$ such that, for all $|h|<\sqrt{t^{1/\alpha}}$,
$$\Big|D^{\mathcal{L},m}_{\alpha,t}(x+h,y)-D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big|\leq C_{N}\Big(\frac{h}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha }}
\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (iii)} There exists a constant $C_{N}>0$ such that
$$\Big|\int_{\mathbb{R}^{n}}D^{\mathcal{L},m}_{\alpha,t}(x,y)dy\Big|\leq \frac{C_{N}(\sqrt{t^{1/\alpha}}/\rho(x))^{\delta}}{(1+\sqrt{t^{1/\alpha}}/\rho(x))^{N}}.$$
\end{proposition}
%For $t>0$, define
%$$\widetilde{D}^{L}_{\alpha,t}(x,y)=t^{1/2\alpha}\nabla_{x}K^{L}_{\alpha,t}(x,y).$$
\begin{proposition}\label{pro3.13}{\rm (\cite[Propositions 3.6, \ 3.9\ and\ 3.10]{Li2})} Suppose that and $V\in B_{q}$ for some $q>n$.
\item{\rm (i)} Let $\alpha\in(0,1)$. For every $N$, there is a constant $C_{N}>0$ such that
$$\Big|\widetilde{D}_{\alpha,t}^{\mathcal{L}}(x,y)\Big|\leq \frac{C_{N}t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha }}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii)} Let $\alpha\in(0,1)$ and $\delta_{1}=1-n/q$. For every $N>0$, there exists a constant $C_{N}>0$ such that for all $|h|<\sqrt{t^{1/\alpha}}$,
$$\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x+h,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big|\leq C_{N}\Big(\frac{h}{\sqrt{t^{1/\alpha}}}\Big)^{\delta_{1}}\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha }}
\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
\item{\rm (iii)} Let $\alpha\in (0, 1/2-n/2q)$. There exists a constant $C_{N}>0$ such that
$$\Big|\int_{\mathbb{R}^{n}}\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)dy\Big|\leq C_{N}\min\Bigg\{\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{1+2\alpha},\
\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{-N}\Bigg\}.$$
\end{proposition}
In order to establish the $BMO^{\gamma}_{L}$-boundedness of operators via $T1$ type theorem, we need the following propositions.
\begin{proposition}\label{pro1}
There exists a constant $C>0$ such that
\begin{equation*}
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big|\leq
\left\{\begin{aligned}
&C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\leq |x-y|;\\
&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\geq |x-y|.
\end{aligned}\right.
\end{equation*}
\end{proposition}
\begin{proof}
By the subordinative formula (\ref{p-1}) and Lemma {\ref{lem1}}, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& \int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big|K^{\mathcal{L}}_{s}(x,y)-K_{s}(x-y)\Big|ds \\
&\leq& C\int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-n/2}e^{-|x-y|^{2}/s}ds.
\end{eqnarray*}
On the one hand, letting $s=t^{1/\alpha}u$, we can get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& C\int^{\infty}_{0}\frac{t}{s^{1+\alpha}}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-n/2}e^{-|x-y|^{2}/s}ds \\
&\leq&C\int^{\infty}_{0}\frac{t}{(t^{1/\alpha}u)^{1+\alpha}}\Big(\frac{\sqrt{t^{1/\alpha}u}}{\rho(x)}\Big)^{\delta_{0}}(t^{1/\alpha}u)^{-n/2}
e^{-c|x-y|^{2}/(t^{1/\alpha}u)}t^{1/\alpha}du \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-{n}/{(2\alpha)}}\int^{\infty}_{0}u^{-1-\alpha-n/2+\delta_{0}/2}e^{-|x-y|^{2}/(t^{1/\alpha}u)}
du.
\end{eqnarray*}
Applying the change of variables: ${|x-y|^{2}}/{(t^{1/\alpha}u)}=r^{2}$, we deduce that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/(2\alpha)}\int^{\infty}_{0}
\Big(\frac{|x-y|^{2}}{t^{1/\alpha}r^{2}}\Big)^{-1-\alpha+\delta_{0}/2-n/2}e^{-r^{2}}\frac{|x-y|^{2}}{t^{1/\alpha}r^{3}}dr \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}|x-y|^{-2\alpha+\delta_{0}-n}t^{1-\delta_{0}/(2\alpha)}\int^{\infty}_{0}
e^{-r^{2}}r^{2\alpha
-\delta_{0}+n+1}dr \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t^{1-\delta_{0}/(2\alpha)}}{|x-y|^{2\alpha+n-\delta_{0}}}.
\end{eqnarray*}
On the other hand, taking $\tau=s/t^{1/\alpha}$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& C\int^{\infty}_{0}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-n/2}t^{-1/\alpha}\eta^{\alpha}_{1}(s/t^{1/\alpha})ds \\
&\leq&C\int^{\infty}_{0}\Big(\frac{\sqrt{t^{1/\alpha}\tau}}{\rho(x)}\Big)^{\delta_{0}}(t^{1/\alpha}\tau)^{-n/2}\eta^{\alpha}_{1}(\tau)d\tau \\
&\leq&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/2\alpha}\int^{\infty}_{0}\eta^{\alpha}_{1}(\tau)\tau^{\delta_{0}/2-n/2}d\tau \\
&\leq& C \Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/2\alpha}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}\leq |x-y|$, then
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{|x-y|^{2\alpha+n}} \\
&\leq& C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{2\alpha+n}}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}>|x-y|$, we can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{t^{n/(2\alpha)+1}} \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{2\alpha+n}}.
\end{eqnarray*}
\end{proof}
Let $\widetilde{D}_{\alpha,t}(\cdot)=t^{1/(2\alpha)}\nabla_{x}e^{-t(-\Delta)^{\alpha}}(\cdot).$ Similar to the proof of Proposition \ref{pro1}, by (\ref{p-1}) and Lemma \ref{lem1-1}, we have
\begin{proposition}\label{pro1-1}
There exists a constant $C>0$ such that
\begin{equation*}
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big|\leq
\left\{\begin{aligned}
&C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\leq |x-y|;\\
&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\geq |x-y|.
\end{aligned}\right.
\end{equation*}
\end{proposition}
\begin{proof}
By the subordinative formula (\ref{p-1}) and Lemma {\ref{lem1-1}}, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& \sqrt{t^{1/\alpha}}\int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big|\nabla_{x}K^{\mathcal{L}}_{s}(x,y)-\nabla_{x}K_{s}(x-y)\Big|ds \\
&\leq& C\sqrt{t^{1/\alpha}}\int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-(n+1)/2}e^{-c|x-y|^{2}/s}ds.
\end{eqnarray*}
On the one hand, letting $s=t^{1/\alpha}u$, we have
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& C\sqrt{t^{1/\alpha}}\int^{\infty}_{0}\frac{t}{s^{1+\alpha}}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-(n+1)/2}e^{-c|x-y|^{2}/s}ds \\
&\leq&C\sqrt{t^{1/\alpha}}\int^{\infty}_{0}\frac{t}{(t^{1/\alpha}u)^{1+\alpha}}\Big(\frac{\sqrt{t^{1/\alpha}u}}{\rho(x)}\Big)^{\delta_{0}}
(t^{1/\alpha}u)^{-(n+1)/2}
e^{-c|x-y|^{2}/(t^{1/\alpha}u)}t^{1/\alpha}du \\
&\leq& C\sqrt{t^{1/\alpha}}\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-(n+1)/(2\alpha)}\int^{\infty}_{0}u^{-1-\alpha-(n+1)/2+\delta_{0}/2}
e^{-c|x-y|^{2}/(t^{1/\alpha}u)}
du.
\end{eqnarray*}
By the change of variables ${|x-y|^{2}}/{(t^{1/\alpha}u)}=r^{2}$, we can get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/2\alpha}\int^{\infty}_{0}
\Big(\frac{|x-y|^{2}}{t^{1/\alpha}r^{2}}\Big)^{-1-\alpha+\delta_{0}/2-(n+1)/2}e^{-r^{2}}\frac{|x-y|^{2}}{t^{1/\alpha}r^{3}}dr \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}|x-y|^{-2\alpha+\delta_{0}-n-1}t^{1-\delta_{0}/(2\alpha)+1/(2\alpha)}
\int^{\infty}_{0}e^{-r^{2}}r^{2\alpha
-\delta_{0}+n}dr \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t^{1+1/(2\alpha)-\delta_{0}/(2\alpha)}}{|x-y|^{2\alpha+n+1-\delta_{0}}}.
\end{eqnarray*}
On the other hand, the change of variable $\tau=s/t^{1/\alpha}$ gives
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& C\sqrt{t^{1/\alpha}}\int^{\infty}_{0}\Big(\frac{\sqrt{s}}{\rho(x)}\Big)^{\delta_{0}}s^{-(n+1)/2}t^{-1/\alpha}\eta^{\alpha}_{1}(s/t^{1/\alpha})ds \\
&\leq&C\sqrt{t^{1/\alpha}}\int^{\infty}_{0}\Big(\frac{\sqrt{t^{1/\alpha}\tau}}{\rho(x)}\Big)^{\delta_{0}}(t^{1/\alpha}\tau)^{-(n+1)/2}
\eta^{\alpha}_{1}(\tau)d\tau \\
&\leq&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/(2\alpha)}\int^{\infty}_{0}\eta^{\alpha}_{1}(\tau)\tau^{\delta_{0}/2-(n+1)/2}d\tau \\
&\leq& C \Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}t^{-n/(2\alpha)}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}\leq |x-y|$, then
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{|x-y|^{2\alpha+n}} \\
&\leq& C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{2\alpha+n}}.
\end{eqnarray*}
If $t^{1/2\alpha}>|x-y|$, we can also get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big| &\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{t^{n/(2\alpha)+1}} \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{2\alpha+n}}.
\end{eqnarray*}
\end{proof}
Let $D^{m}_{\alpha,t}(\cdot)=t^{m}\partial^{m}_{t}e^{-t(-\Delta)^{\alpha}}(\cdot).$ We have
\begin{proposition}\label{pro1-1-1}
There exists a constant $C>0$ such that
\begin{equation*}
\Big|D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{m}_{\alpha,t}(x-y)\Big|\leq
\left\{ \begin{aligned}
&C\Big(\frac{|x-y|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\leq |x-y|;\\
&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(|x-y|+\sqrt{t^{1/\alpha}})^{n+2\alpha}},\ \ \ \sqrt{t^{1/\alpha}}\geq |x-y|.
\end{aligned}\right.
\end{equation*}
\end{proposition}
\begin{proof}
The proposition can be proved by Proposition \ref{pro2.5} and (\ref{p-1}). Since the argument is similar to that of Proposition \ref{pro1}, the details is omitted.
\end{proof}
\begin{proposition}\label{pro2}
For every $0<\delta<\min\{2\alpha,\delta_{0}\}$ and $|y-z|<\min\{C\rho(y),|x-y|/4\}$, there exists a constant $C>0$ such that
$$\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}.$$
\end{proposition}
\begin{proof}
By the subordinative formula (\ref{p-1}), we can use Lemma \ref{lem2} to deduce
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\\
&&\leq \int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big|\Big(K^{\mathcal{L}}_{s}(x,y)-K_{s}(x-y)\Big)-\Big(K^{\mathcal{L}}_{s}(x,z)-K_{s}(x-z)\Big)\Big|ds \\
&&\leq C\int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}s^{-n/2}e^{-|x-y|^{2}/s}ds.
\end{eqnarray*}
On the one hand, taking $s=t^{1/\alpha}u$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\\ &&\leq
\int^{\infty}_{0}\frac{Ct}{s^{1+\alpha}}\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}s^{-n/2}e^{-|x-y|^{2}/s}ds \\
&&\leq C\int^{\infty}_{0}\frac{t}{(t^{1/\alpha}u)^{1+\alpha}}\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}(t^{1/\alpha}u)^{-n/2}
e^{-|x-y|^{2}/(t^{1/\alpha}u)}t^{1/\alpha}du \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}t^{-n/(2\alpha)}\int^{\infty}_{0}e^{-|x-y|^{2}/(t^{1/\alpha}u)}u^{-1-\alpha-n/2}du.
\end{eqnarray*}
Let $\displaystyle \frac{|x-y|^{2}}{(t^{1/\alpha}u)}=r$. We can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\\ &&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}t^{-n/(2\alpha)}\int^{\infty}_{0}e^{-r}\Big(\frac{|x-y|^{2}}{t^{1/\alpha}r}\Big)^{-1-\alpha-n/2}
\frac{|x-y|^{2}}{t^{1/\alpha}r^{2}}dr \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}\frac{t}{|x-y|^{2\alpha+n}}.
\end{eqnarray*}
On the other hand, letting $\displaystyle \frac{s}{t^{1/\alpha}}=\tau$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big| \\ &&\leq \int^{\infty}_{0}\frac{1}{t^{1/\alpha}}\eta^{\alpha}_{1}(s/t^{1/\alpha})\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}s^{-n/2}ds \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}\int^{\infty}_{0}\eta^{\alpha}_{1}(\tau)(t^{1/\alpha}\tau)^{-n/2}d\tau \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}t^{-n/(2\alpha)}.
\end{eqnarray*}
{\it Case 1: $\sqrt{t^{1/\alpha}}\leq|x-y|$.} We obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}.
\end{eqnarray*}
{\it Case 2: $\sqrt{t^{1/\alpha}}>|x-y|$.} We can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(x,y)-K_{\alpha,t}(x-y)\Big)-\Big(K^{\mathcal{L}}_{\alpha,t}(x,z)-K_{\alpha,t}(x-z)\Big)\Big|\\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}
\frac{t}{t^{n/(2\alpha)+1}}\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}.
\end{eqnarray*}
\end{proof}
\begin{proposition}\label{pro2-2-2}
For every $0<\delta<\min\{2\alpha,\delta_{0}\}$ and $|y-z|<\min\{C\rho(y),|x-y|/4\}$, there exists a constant $C>0$ such that
$$\Big|\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{m}_{\alpha,t}(x-y)\Big)-\Big(D^{\mathcal{L},m}_{\alpha,t}(x,z)-D^{m}_{\alpha,t}(x-z)\Big)\Big|\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}.$$
\end{proposition}
\begin{proof}
The proof is similar to that of Proposition \ref{pro2}, so we omit the details.
\end{proof}
\begin{proposition}\label{pro2-2}
Suppose that $V\in B_{q}$ for some $q>n$. For $0<\delta'<\delta_{1}:=1-n/q$, and $|y-z|<\min\{C\rho(y),|x-y|/4\}$, there exists a constant $C>0$ such that
$$\Big|\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big)-
\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)-\widetilde{D}_{\alpha,t}(x-z)\Big)\Big|\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}.$$
\end{proposition}
\begin{proof}
By the subordinative formula (\ref{p-1}), we can use Lemma \ref{lem2-2} to get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big)-
\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)-\widetilde{D}_{\alpha,t}(x-z)\Big)\Big|\\
&&\leq t^{1/(2\alpha)}\int^{\infty}_{0}\eta^{\alpha}_{t}(s)
\Big|\Big(\nabla_{x}K^{\mathcal{L}}_{s}(x,y)-\nabla_{x}K_{s}(x-y)\Big)-\nabla_{x}\Big(K^{\mathcal{L}}_{s}(x,z)-K_{s}(x-z)\Big)\Big|ds \\
&&\leq Ct^{1/(2\alpha)}\int^{\infty}_{0}\eta^{\alpha}_{t}(s)\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}s^{-(n+1)/2}e^{-c|x-y|^{2}/s}ds.
\end{eqnarray*}
On the one hand, taking $s=t^{1/\alpha}u$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big)-
\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)-\widetilde{D}_{\alpha,t}(x-z)\Big)\Big|\\ &&\leq
Ct^{1/(2\alpha)}\int^{\infty}_{0}\frac{t}{s^{1+\alpha}}\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}s^{-(n+1)/2}e^{-c|x-y|^{2}/s}ds \\
&&\leq Ct^{1/(2\alpha)}\int^{\infty}_{0}\frac{t}{(t^{1/\alpha}u)^{1+\alpha}}\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}(t^{1/\alpha}u)^{-(n+1)/2}
e^{-c|x-y|^{2}/(t^{1/\alpha}u)}t^{1/\alpha}du \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}t^{-n/(2\alpha)}\int^{\infty}_{0}e^{-c|x-y|^{2}/(t^{1/\alpha}u)}u^{-1-\alpha-(n+1)/2}du.
\end{eqnarray*}
Let ${|x-y|^{2}}/{(t^{1/\alpha}u)}=r$. It holds
\begin{eqnarray}\label{eq-3.1}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big)-
\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)-\widetilde{D}_{\alpha,t}(x-z)\Big)\Big|\\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}t^{-n/(2\alpha)}\int^{\infty}_{0}e^{-r}\Big(\frac{|x-y|^{2}}{t^{1/\alpha}r}\Big)^{-1-\alpha-(n+1)/2}
\frac{|x-y|^{2}}{t^{1/\alpha}r^{2}}dr \nonumber\\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}\frac{t^{1+1/(2\alpha)}}{|x-y|^{2\alpha+n+1}}.\nonumber
\end{eqnarray}
On the other hand, letting $s/t^{1/\alpha}=\tau$, we obtain
\begin{eqnarray}\label{eq-3.2}
% \nonumber to remove numbering (before each equation)
&&\Big|\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}_{\alpha,t}(x-y)\Big)-
\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)-\widetilde{D}_{\alpha,t}(x-z)\Big)\Big| \\
&&\leq Ct^{1/(2\alpha)}\int^{\infty}_{0}\frac{1}{t^{1/\alpha}}\eta^{\alpha}_{1}(s/t^{1/\alpha})\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}s^{-(n+1)/2}ds\nonumber \\
&&\leq Ct^{1/(2\alpha)}\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}\int^{\infty}_{0}\eta^{\alpha}_{1}(\tau)(t^{1/\alpha}\tau)^{-(n+1)/2}d\tau\nonumber \\
&&\leq C\Big(\frac{|y-z|}{\rho(x)}\Big)^{\delta'}t^{-n/(2\alpha)}.\nonumber
\end{eqnarray}
It is easy to see that the desired estimate follows from (\ref{eq-3.1})\ and \ (\ref{eq-3.2}). The proof of Proposition \ref{pro2-2} is finished.
\end{proof}
\section{$BMO^{\gamma}_{\mathcal{L}}$-boundedness via $T1$ theorem}\label{sec-4}
\subsection{Maximal operators for the fractional heat semigroup}\label{sec-4-1}
\begin{definition}\label{defc-1}
Let $0<\gamma\leq 1$. The Campanato type space $BMO^{\gamma}_{\mathcal{L},L^{\infty}((0,\infty),dt)}(\mathbb{R}^{n})$ is defined as the set of all locally integrable functions $f$ satisfying
$$\|f\|_{BMO^{\gamma}_{\mathcal{L},L^{\infty}((0,\infty),dt)}}:=\sup_{B\subset\mathbb{R}^{n}}
\Bigg\{\frac{1}{|B|^{1+\gamma/n}}\int_{B}\|f(x,t)-f(B,V)\|_{L^{\infty}((0,\infty),dt)}dx\Bigg\}<\infty.$$
\end{definition}
To prove that $M^{\alpha}_{\mathcal{L}}$ is bounded from $BMO_{\mathcal{L}}^{\gamma}(\mathbb R^{n})$ into itself, we give a vector-valued interpretation of the operator and apply Lemma \ref{lem4.1}. Indeed, it is clear that $M^{\alpha}_{\mathcal{L}}f=\|e^{-t\mathcal{L}^{\alpha}}f\|_{L^{\infty}((0,\infty),dt)}$. Hence, it is enough to show that the operator $\Lambda(f):=\{e^{-t\mathcal{L}^{\alpha}}f\}_{t>0}$ is bounded from $BMO^{\gamma}_{\mathcal{L}}$ into $BMO_{\mathcal{L},L^{\infty}((0,\infty),dt)}^{\gamma}$.
By the spectral theorem, $\Lambda$ is bounded from $L^{2}(\mathbb{R}^{n})$ into $L^{2}_{L^{\infty}((0,\infty),dt)}(\mathbb{R}^{n})$. The desired result can be then deduced from the following theorem.
\begin{theorem}\label{the3}
Assume that the potential $V\in B_{q}$ with $q>n/2$. Let $x,y,z\in\mathbb{R}^{n}$.
\item{\rm (i)} For any $N>0$, there exists a constant $C_{N}$ such that
$$\Big\|K^{\mathcal{L}}_{\alpha,t}(x,y)\Big\|_{L^{\infty}((0,\infty),dt)}\leq \frac{C_{N}}{|x-y|^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii)} For $|x-y|>2|y-z|$ and any $0<\delta<\min\{2\alpha,\delta_{0}\}$, there exists a constant $C>0$ such that
\begin{multline}\label{t-1}
\Big\|K^{\mathcal{L}}_{\alpha,t}(x,y)-K^{\mathcal{L}}_{\alpha,t}(x,z)\Big\|_{L^{\infty}((0,\infty),dt)}+
\Big\|K^{\mathcal{L}}_{\alpha,t}(y,x)-K^{\mathcal{L}}_{\alpha,t}(z,x)\Big\|_{L^{\infty}((0,\infty),dt)}\\
\quad\leq \frac{C|y-z|^{\delta}}{|x-y|^{n+\delta}}.
\end{multline}
\item{\rm (iii)} There exists a constant $C$ such that for every ball $B=B(x,r)$ with $0<r\leq \rho(x)/2$,
$$\log\Big(\frac{\rho(x)}{r}\Big)\frac{1}{|B|}\int_{B}\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-(e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{\infty}((0,\infty),dt)}dy\leq C,$$
and, if $\gamma<\min\{2\alpha,1,\delta_{0}\}$ then
$$\Big(\frac{\rho(x)}{r}\Big)^{\gamma}\frac{1}{|B|}\int_{B}\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-(e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{\infty}((0,\infty),dt)}dy\leq C.$$
\end{theorem}
\begin{proof}
For (i), from (i) of Proposition \ref{pro2.51}, we can get
$$\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)\Big|\leq C_{N} \min\Bigg\{\frac{t^{1+N/\alpha}}{|x-y|^{n+2\alpha+2N}},\ t^{-n/(2\alpha)}\Bigg\}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
If $\sqrt{t^{1/\alpha}}>|x-y|$, then
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)\Big| &\leq& \frac{C_{N}}{(\sqrt{t^{1/\alpha}})^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N} \\
&\leq&\frac{C}{|x-y|^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}\leq |x-y|$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)\Big| &\leq& \frac{C_{N}t^{1+N/\alpha}}{|x-y|^{n+2\alpha+2N}}\Big(\frac{\sqrt{t^{1/\alpha}}}{|x-y|}\Big)^{-N}
\Big(\frac{|x-y|}{\sqrt{t^{1/\alpha}}}+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N} \\
&\leq& \frac{C_{N}(\sqrt{t^{1/\alpha}})^{2\alpha+N}}{|x-y|^{n+2\alpha+N}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N} \\
&\leq& \frac{C_{N}}{|x-y|^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N}.
\end{eqnarray*}
For (ii), from (ii) of Proposition \ref{pro2.51}, we obtain
$$\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K^{\mathcal{L}}_{\alpha,t}(x,z)\Big|\leq C_{N}\min\Bigg\{\frac{t^{1+N/\alpha}|y-z|^{\delta}}{|x-y|^{2\alpha+n+2N+\delta}},\ t^{-n/(2\alpha)}\Bigg\}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}.$$
If $\sqrt{t^{1/\alpha}}>|x-y|$, then
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K^{\mathcal{L}}_{\alpha,t}(x,z)\Big| &\leq& \frac{C_{N}}{|x-y|^{n}}\Big(\frac{|y-z|}{|x-y|}\Big)^{\delta}
\leq \frac{C|y-z|^{\delta}}{|x-y|^{n+\delta}}.
\end{eqnarray*}
If $|x-y|>\sqrt{t^{1/\alpha}}$, we can also get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big|K^{\mathcal{L}}_{\alpha,t}(x,y)-K^{\mathcal{L}}_{\alpha,t}(x,z)\Big| &\leq& \frac{C_{N}|x-y|^{2\alpha+2N}}{|x-y|^{2\alpha+n+2N+\delta}}|y-z|^{\delta}
\leq \frac{C_{N}|y-z|^{\delta}}{|x-y|^{n+\delta}}.
\end{eqnarray*}
The symmetry of the kernel $K^{L}_{\alpha,t}(\cdot,\cdot)$ gives the conclusion of (ii).
For (iii), letting $B=B(x,r)$ with $0<r\leq \rho(x)/2$, the triangle inequality gives
$$\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-(e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{\infty}((0,\infty),dt)}\leq \frac{1}{|B|}\int_{B}\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|_{L^{\infty}((0,\infty),dt)}dz.$$
We estimate $\|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\|_{L^{\infty}((0,\infty),dt)}$. Because $y,z\in B$, $\rho(y)\sim \rho(z)\sim\rho(x).$
By Proposition \ref{pro1}, we split
$|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)|\leq S_{1}+S_{2}$,
where
\begin{equation*}
\left\{ \begin{aligned}
S_{1}:= \int_{\mathbb{R}^{n}}\Big|K^{\mathcal{L}}_{\alpha,t}(y,w)-K_{\alpha,t}(y,w)\Big|dw;\\
S_{2}:=\int_{\mathbb{R}^{n}}\Big|K^{\mathcal{L}}_{\alpha,t}(z,w)-K_{\alpha,t}(z,w)\Big|dw.
\end{aligned}\right.
\end{equation*}
For $S_{1}$, if $|y-w|\leq\sqrt{t^{1/\alpha}}$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
S_{1} \leq\int_{\mathbb{R}^{n}}\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(\sqrt{t^{1/\alpha}}+|y-w|)^{n+2\alpha}}dw
\leq \Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}.
\end{eqnarray*}
If $|y-w|>\sqrt{t^{1/\alpha}}$, we can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
S_{1}&\leq&\int_{\mathbb{R}^{n}}\Big(\frac{|y-w|}{\rho(x)}\Big)^{\delta_{0}}\frac{t}{(\sqrt{t^{1/\alpha}}+|y-w|)^{n+2\alpha}}dw \\
&\leq&\frac{1}{\rho(x)^{\delta_{0}}}\int^{\infty}_{0}\frac{|y-w|^{\delta_{0}+n-1}t}{(|y-w|+\sqrt{t^{1/\alpha}})^{n+2\alpha}}dw \\
&\leq&\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\int^{\infty}_{0}u^{\delta_{0}+n-1}(1+u)^{-n-2\alpha}du \\
&\leq& \Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}},
\end{eqnarray*}
where we have chosen $\delta_{0}<2\alpha$ since $\delta_{0}=2-n/q$, $q>n/2$.
The proof of the term $S_{2}$ is similar to that of the term $S_{1}$, so we omit it. Then we can get
$$\Big|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big|\leq C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}},$$
which shows that if $\sqrt{t^{1/\alpha}}\leq 2r$,
$$\Big|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big|\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta_{0}}.$$
If $\sqrt{t^{1/\alpha}}>2r$, then $|y-z|\leq 2r<\sqrt{t^{1/\alpha}}$. Hence, Proposition \ref{pro2.51} implies that for $0<\delta<\delta_{0}$,
\begin{multline}\label{2}
% \nonumber to remove numbering (before each equation)
\Big|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big| \leq \int_{\mathbb{R}^{n}}\Big|K^{\mathcal{L}}_{\alpha,t}(y,w)-K^{\mathcal{L}}_{\alpha,t}(z,w)\Big|dw \\
\leq \int_{\mathbb{R}^{n}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}\frac{t}{(\sqrt{t^{1/\alpha}}+|y-w|)^{n+2\alpha}}dw \leq C\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}\leq C \Big(\frac{r}{\sqrt{t^{1/\alpha}}}\Big)^{\delta}.
\end{multline}
Therefore, if $\sqrt{t^{1/\alpha}}>\rho(x)$, (\ref{2}) gives
$$|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)|\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}.$$
When $2r<\sqrt{t^{1/\alpha}}<\rho(x)$, we have
$ |e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)| = I+II+III,$
where
\begin{equation*}
\left\{ \begin{aligned}
I&:=\int_{|w-y|>c\rho(y)>4|y-z|}\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(y,w)-K_{\alpha,t}(y,w)\Big)-\Big(
K^{\mathcal{L}}_{\alpha,t}(z,w)-K_{\alpha,t}(z,w)\Big)\Big|dw;\\
II&:=\int_{4|y-z|<|w-y|<c\rho(y)}\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(y,w)-K_{\alpha,t}(y,w)\Big)-\Big(
K^{\mathcal{L}}_{\alpha,t}(z,w)-K_{\alpha,t}(z,w)\Big)\Big|dw;\\
III&:= \int_{|w-y|<4|y-z|}\Big|\Big(K^{\mathcal{L}}_{\alpha,t}(y,w)-K_{\alpha,t}(y,w)\Big)-\Big(
K^{\mathcal{L}}_{\alpha,t}(z,w)-K_{\alpha,t}(z,w)\Big)\Big|dw.
\end{aligned}\right.
\end{equation*}
Noticing that the estimate (\ref{t-1}) is valid for the classical fractional heat kernel. For $I$, by (\ref{t-1}), we can get
$$I\leq C\int_{|w-y|>c\rho(y)>4|y-z|}\frac{|y-z|^{\delta}}{|w-y|^{n+\delta}}dw\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}.$$
For $II$, we apply Proposition \ref{pro2} and the fact that $\rho(w)\sim \rho(y)$ in the region of integration.
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
II \leq C|y-z|^{\delta}\int_{4|y-z|<|w-y|<c\rho(y)}\frac{tdw}{\rho(w)^{\delta}(\sqrt{t^{1/\alpha}}+|w-y|)^{n+2\alpha}}
\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}.
\end{eqnarray*}
For $III$, since $|y-z|\leq 2r<\sqrt{t^{1/\alpha}},$ we have $|w-y|<C\sqrt{t^{1/\alpha}}$. For $n-\delta_{0}>0$, by Proposition \ref{pro1}, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
III &\leq&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\Bigg(\int_{|w-y|<4|y-z|}\frac{tdw}{(\sqrt{t^{1/\alpha}}+|w-y|)^{n+2\alpha}}+
\int_{|w-z|\leq 5|y-z|}\frac{tdw}{(\sqrt{t^{1/\alpha}}+|w-z|)^{n+2\alpha}}\Bigg) \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\int^{5|y-z|/\sqrt{t^{1/\alpha}}}_{0}u^{n-1}du
\leq C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}\Big)^{\delta_{0}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{n} \\
&\leq&\frac{Cr^{n}}{\rho(x)^{\delta_{0}}(\sqrt{t^{1/\alpha}})^{n-\delta_{0}}}\leq \frac{Cr^{n}}{\rho(x)^{\delta_{0}}r^{n-\delta_{0}}}=C\Big(\frac{r}{\rho(x)}\Big)^{\delta_{0}}.
\end{eqnarray*}
Thus, when $2r<\sqrt{t^{1/\alpha}}<\rho(x)$,
$$\Big|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big|\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}.$$
Combining the above estimates, we can get
\begin{equation}\label{4.8}
\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|_{L^{\infty}((0,\infty),dt)}\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}.
\end{equation}
Therefore, it holds
$$\log\Big(\frac{\rho(x)}{r}\Big)\frac{1}{|B|}\int_{B}\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-(e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{\infty}((0,\infty),dt)}dy\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta}\log\Big(\frac{\rho(x)}{r}\Big)\leq C,$$
which is the first conclusion of (iii).
For the second estimate of (iii), take $\delta\in(\gamma,\ \min\{2\alpha,\ 2-n/q\})$. by (\ref{4.8}), we have
$$\Big(\frac{\rho(x)}{r}\Big)^{\gamma}\frac{1}{|B|}\int_{B}\Big\|e^{-t\mathcal{L}^{\alpha}}1(y)-(e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{\infty}((0,\infty),dt)}dy\leq C\Big(\frac{r}{\rho(x)}\Big)^{\delta-\gamma}\leq C.$$
\end{proof}
\subsection{Boundedness of the Littlewood-Paley $g$-function $g^{\mathcal{L}}_{\alpha}$}\label{sec-4-2}
Similar to Section \ref{sec-4-1}, we introduce the following function space:
\begin{definition}\label{defc-2}
Let $0<\gamma\leq 1$. The Campanato type space $BMO^{\gamma}_{ \mathcal{L},L^{2}((0,\infty),dt/t)}(\mathbb{R}^{n})$ is defined as the set of all locally integrable functions $f$ satisfying
$$\|f\|_{BMO^{\gamma}_{\mathcal{L},L^{2}((0,\infty),dt)}}:=\sup_{B\subset\mathbb{R}^{n}}
\Bigg\{\frac{1}{|B|^{1+\gamma/n}}\int_{B}\|f(x,t)-f(B,V)\|_{L^{2}((0,\infty),dt/t)}dx\Bigg\}<\infty.$$
\end{definition}
The functional calculus and the spectral theorem imply that $g^{\mathcal{L}}_{\alpha}$ is an isometry on $L^{2}(\mathbb{R}^{n})$. As before, to get the boundedness of $g^{\mathcal{L}}_{\alpha}$ on $BMO^{\gamma}_{\mathcal{L}}(\mathbb R^{n})$, it is sufficient to prove the following result.
\begin{theorem}\label{the3.1}
Assume that the potential $V\in B_{q}$ with $q>n/2$. Let $x,y,z\in\mathbb{R}^{n}$ and $N>0$.
\item{\rm (i) } For any $N>0$, there exists a constant $C_{N}$ such that
$$\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}\leq \frac{C_{N}}{|x-y|^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N}.$$
\item{\rm (ii) } If $|x-y|>2|y-z|$ and $0<\delta<\min\{2\alpha,\delta_{0},1\}$, for any constant $N>0$, there exists a constant $C_{N}$ such that
\begin{eqnarray}\label{t-2}
% \nonumber to remove numbering (before each equation)
&&\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}+
\Big\|D^{\mathcal{L},m}_{\alpha,t}(y,x)-D^{\mathcal{L},m}_{\alpha,t}(z,x)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}\\
&&\quad \leq \frac{C_{N}|y-z|^{\delta}}{|x-y|^{n+\delta}}.\nonumber
\end{eqnarray}
\item{\rm (iii)} There exists a constant $C$ such that for every ball $B=B(x_{0},r)$ with $0<r\leq \rho(x)/2$,
$$\log\Big(\frac{\rho(x)}{r}\Big)\frac{1}{|B|}\int_{B}\Big\|t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(y)-
(t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}dy\leq C,$$
and, if $\gamma<\min\{2\alpha,\delta_{0},1\}$, then
$$\Big(\frac{\rho(x)}{r}\Big)^{\gamma}\frac{1}{|B|}\int_{B}\Big\|t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(y)-(t^{m}\partial_{t}^{m}
e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}dy\leq C.$$
\end{theorem}
\begin{proof}
For (i), from Proposition \ref{pro3.12}, we have
$$\Big|D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big|\leq C_{N}\min\Bigg\{\frac{t^{1+N/\alpha}}{|x-y|^{n+2\alpha+2N}},\ t^{-n/(2\alpha)} \Bigg\}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
If $\sqrt{t^{1/\alpha}}\leq |x-y|$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
&&\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}^{2}\\
&&\leq C_{N}\int^{|x-y|^{2\alpha}}_{0}\frac{t^{2+2N/\alpha}}{|x-y|^{2n+4\alpha+4N}}
\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-2N}\frac{dt}{t} \\
&&\leq C_{N}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\int^{|x-y|^{2\alpha}}_{0}
\frac{(\sqrt{t^{1/\alpha}})^{4\alpha+4N}}{|x-y|^{2n+4\alpha+4N}}\Big(\frac{\sqrt{t^{1/\alpha}}}{|x-y|}\Big)^{-2N}\frac{dt}{t} \\
&&\leq \Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\frac{C_{N}}{|x-y|^{2n}}\int^{|x-y|^{2\alpha}}_{0}
\Big(\frac{\sqrt{t^{1/\alpha}}}{|x-y|}\Big)^{4\alpha+2N}\frac{dt}{t}.
\end{eqnarray*}
Let ${\sqrt{t^{1/\alpha}}}/{|x-y|}=u$. We can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}^{2} &\leq&\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\frac{C_{N}}{|x-y|^{2n}}\int^{1}_{0}u^{4\alpha+2N-1}du \\
&\leq&\frac{C_{N}}{|x-y|^{2n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}\geq |x-y|$, we can get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq&C_{N}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\int^{\infty}_{|x-y|^{2\alpha}}t^{-n/\alpha-1}dt \\
&\leq& \frac{C_{N}}{|x-y|^{2n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}.
\end{eqnarray*}
For (ii), By Proposition \ref{pro3.12}, we have
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq& \int^{\infty}_{0}\frac{C_{N}t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{2n+4\alpha}}\Big(
\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}dt \\
&\leq& C_{N}|y-z|^{2\delta}\int^{\infty}_{0}\frac{(\sqrt{t^{1/\alpha}})^{2\alpha-2\delta}}{(\sqrt{t^{1/\alpha}}+|x-y|)^{2n+4\alpha}}dt.
\end{eqnarray*}
Let ${\sqrt{t^{1/\alpha}}}/{|x-y|}=u$. We obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq&C_{N}|y-z|^{2\delta}|x-y|^{-2n-2\delta}\int^{\infty}_{0}\frac{u^{4\alpha-2\delta-1}}{(1+u)^{2n+4\alpha}}du \\
&\leq& \frac{C_{N}|y-z|^{2\delta}}{|x-y|^{2n+2\delta}}.
\end{eqnarray*}
The symmetry of the kernel $D^{L,m}_{\alpha,t}(\cdot,\cdot)$ gives the conclusion of (ii).
For (iii), let us fix $y,z\in B=B(x_{0},r)$, $0<r\leq \rho(x_{0})/2$. Similar to Theorem \ref{the3}, we must handle $$\Big\|t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}.$$ We can write
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})}
&=& M_{1}+M_{2}+M_{3},
\end{eqnarray*}
where
\begin{equation*}
\left\{ \begin{aligned}
M_{1}&:=\int^{(2r)^{2\alpha}}_{0}\Big|
\int_{\mathbb{R}^{n}}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t};\\
M_{2}&:=\int^{\rho(x_{0})^{2\alpha}}_{(2r)^{2\alpha}}\Big|
\int_{\mathbb{R}^{n}}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t};\\
M_{3}&:=\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big|
\int_{\mathbb{R}^{n}}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t}.
\end{aligned}\right.
\end{equation*}
Since $y,z\in B\subset B(x_{0},\rho(x_{0}))$, it follows that $\rho(y)\sim\rho(x_{0})\sim\rho(z)$. By Proposition \ref{pro3.12} (iii),
$$M_{1}\leq C\int^{(2r)^{2\alpha}}_{0}\frac{(\sqrt{t^{1/\alpha}}/\rho(x_{0}))^{2\delta}}{(1+\sqrt{t^{1/\alpha}}/\rho(x_{0}))^{2N}}\frac{dt}{t}
\leq C\int^{(2r)^{2\alpha}}_{0}\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x_{0})}\Big)^{2\delta}\frac{dt}{t}=C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}.$$
Also, by Proposition \ref{pro3.12}(ii),
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
M_{3} &\leq& C\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}\Big|\int_{\mathbb{R}^{n}}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}dx\Big|^{2}\frac{dt}{t} \\
&\leq& C\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}\frac{dt}{t}\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}.
\end{eqnarray*}
It remains to estimate the term $M_{2}$. In this case, $|y-z|\leq 2r\leq \sqrt{t^{1/\alpha}}\leq \rho(x_{0})$. Then we can use the methods in Theorem \ref{the3},
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
M_{2}
=\int^{\rho(x_{0})^{2\alpha}}_{(2r)^{2\alpha}}\Big|M_{2,1}+M_{2,2}+M_{2,3}\Big|^{2}\frac{dt}{t},
\end{eqnarray*}
where
\begin{equation*}
\left\{ \begin{aligned}
M_{2,1}&:=\int_{|x-y|>c\rho(y)>4|y-z|}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{m}_{\alpha,t}(x-y)\Big)-\Big(D^{\mathcal{L},m}_{\alpha,t}(x,z)-D^{m}_{\alpha,t}(x-z)\Big)dx;\\
M_{2,2}&:=\int_{4|y-z|<|x-y|<c\rho(y)}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{m}_{\alpha,t}(x-y)\Big)-\Big(D^{\mathcal{L},m}_{\alpha,t}(x,z)-D^{m}_{\alpha,t}(x-z)\Big)dx;\\
M_{2,3}&:=\int_{|x-y|<4|y-z|}\Big(D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{m}_{\alpha,t}(x-y)\Big)-\Big(D^{\mathcal{L},m}_{\alpha,t}(x,z)-D^{m}_{\alpha,t}(x-z)\Big)dx.
\end{aligned}\right.
\end{equation*}
For $M_{2,1}$, similar to prove (\ref{t-2}), we also can get
$$\Big|D^{\mathcal{L},m}_{\alpha,t}(x,y)-D^{\mathcal{L},m}_{\alpha,t}(x,z)\Big|\leq \frac{C|y-z|^{\delta}}{|x-y|^{n+\delta}},$$
which is valid to $D^{m}_{\alpha,t}(\cdot)$. So we obtain
$$|M_{2,1}|\leq C\int_{|x-y|>c\rho(y)>4|y-z|}\frac{|y-z|^{\delta}}{|x-y|^{n+\delta}}dx\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{\delta}.$$
For $M_{2,2}$, by Proposition \ref{pro2-2-2}
and the fact that $\rho(x)\sim \rho(y)$ in the region of integration.
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
|M_{2,2}| &\leq&C|y-z|^{\delta}\int_{4|y-z|<|x-y|<c\rho(y)}\frac{t}{\rho(x)^{\delta}(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}dx \leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{\delta}.
\end{eqnarray*}
For $M_{2,3}$, since $|y-z|\leq 2r<\sqrt{t^{1/\alpha}},$ we have $|x-y|<C\sqrt{t^{1/\alpha}}$. For $n-\delta_{0}>0$, by Proposition \ref{pro1-1-1},
we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
| M_{2,3}| &\leq&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x_{0})}\Big)^{\delta_{0}}\Bigg(\int_{|x-y|<4|y-z|}\frac{tdx}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}+
\int_{|x-z|\leq 5|y-z|}\frac{tdx}{(\sqrt{t^{1/\alpha}}+|x-z|)^{n+2\alpha}}\Bigg) \\
&\leq& C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x_{0})}\Big)^{\delta_{0}}\int^{5|y-z|/\sqrt{t^{1/\alpha}}}_{0}u^{n-1}du \\
&\leq&C\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x_{0})}\Big)^{\delta_{0}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{n} \\
&\leq&\frac{Cr^{n}}{\rho(x_{0})^{\delta_{0}}(\sqrt{t^{1/\alpha}})^{n-\delta_{0}}}\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{\delta_{0}}.
\end{eqnarray*}
The estimates for $M_{2,i}, i=1,2,3$, imply that
$$M_{2}\leq \int^{\rho(x_{0})^{2\alpha}}_{(2r)^{2\alpha}}\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}\frac{dt}{t}=C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}
\log\Big(\frac{\rho(x_{0})}{r}\Big).$$
Finally, we can get
$$\Big\|t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{m}\partial_{t}^{m}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{\delta}\Big(\log(\frac{\rho(x_{0})}{r})\Big)^{1/2}.$$
Thus (iii) readily follows.
\end{proof}
\subsection{Boundedness of Littlewood-Paley $g$-function $\widetilde{g}^{\mathcal{L}}_{\alpha}$}\label{sec-4-2}
By the $L^{2}$-boundedness of Riesz transforms $\nabla_{x}\mathcal{L}^{-1/2}$, we can see that
$$\|\widetilde{g}^{\mathcal{L}}_{\alpha}f\|^{2}_{L^{2}}\leq C\int^{\infty}_{0}\Big(\int_{\mathbb{R}^{n}}|t^{1/2\alpha}\mathcal{L}^{1/2}e^{-t\mathcal{L}^{\alpha}}f(x)|^{2}dx\Big)\frac{dt}{t}.$$
Then by the spectral theorem, we know that $\widetilde{g}^{\mathcal{L}}_{\alpha}$ is bounded from $L^{2}(\mathbb{R}^{n})$ into $L^{2}(\mathbb{R}^{n})$.
\begin{theorem}\label{the3.2}
Assume that the potential $V\in B_{q}$ with $q>n$. Let $x,y,z\in\mathbb{R}^{n}$. Then
\item{\rm (i) } For any $N>0$, there exists a constant $C_{N}$ such that
$$\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}\leq \frac{C_{N}}{|x-y|^{n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-N};$$
\item{\rm (ii) } Let $|x-y|>2|y-z|$ and $0<\delta<\min\{2\alpha,\delta_{1},1\}$. For any $N>0$, there exists a constant $C_{N}$ such that $$\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}+
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(y,x)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(z,x)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}\leq
\frac{C_{N}|y-z|^{\delta}}{|x-y|^{n+\delta}};$$
\item{\rm (iii)} There exists a constant $C$ such that for every ball $B=B(x_{0},r)$ with $0<r\leq \rho(x)/2$,
$$\log\Big(\frac{\rho(x)}{r}\Big)\frac{1}{|B|}\int_{B}\Big\|t^{1/(2\alpha)}\nabla_{x}
e^{-t\mathcal{L}^{\alpha}}1(y)-(t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}dy\leq C,$$
and, if $\gamma<\min\{2\alpha,\delta_{1},1\}$ then
$$\Big(\frac{\rho(x)}{r}\Big)^{\gamma}\frac{1}{|B|}\int_{B}\Big\|t^{1/(2\alpha)}\nabla_{x}
e^{-t\mathcal{L}^{\alpha}}1(y)-(t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1)_{B}\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}dy\leq C.$$
\end{theorem}
\begin{proof}
For (i), from Proposition \ref{pro3.13}, we have
$$\Big|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big|\leq C_{N}\min\Bigg\{\frac{t^{1+N/\alpha+1/(2\alpha)}}{|x-y|^{n+2\alpha+2N+1}},\ t^{-n/(2\alpha)} \Bigg\}\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-N}.$$
If $\sqrt{t^{1/\alpha}}\leq |x-y|$, we obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}^{2}&\leq& C_{N}\int^{|x-y|^{2\alpha}}_{0}\frac{t^{2+2N/\alpha+1/\alpha}}{|x-y|^{2n+4\alpha+4N+2}}
\Big(1+\frac{\sqrt{t^{1/\alpha}}}{\rho(x)}+\frac{\sqrt{t^{1/\alpha}}}{\rho(y)}\Big)^{-2N}\frac{dt}{t} \\
&\leq&C_{N}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\int^{|x-y|^{2\alpha}}_{0}
\frac{(\sqrt{t^{1/\alpha}})^{4\alpha+4N+2}}{|x-y|^{2n+4\alpha+4N+2}}\Big(\frac{\sqrt{t^{1/\alpha}}}{|x-y|}\Big)^{-2N}\frac{dt}{t} \\
&\leq& \Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\frac{C_{N}}{|x-y|^{2n}}\int^{|x-y|^{2\alpha}}_{0}
\Big(\frac{\sqrt{t^{1/\alpha}}}{|x-y|}\Big)^{4\alpha+2N+2}\frac{dt}{t}.
\end{eqnarray*}
Let $\sqrt{t^{1/\alpha}}/|x-y|=u$. We can see that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}^{2} &\leq&\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\frac{C_{N}}{|x-y|^{2n}}\int^{1}_{0}u^{4\alpha+2N+1}du \\
&\leq&\frac{C_{N}}{|x-y|^{2n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}.
\end{eqnarray*}
If $\sqrt{t^{1/\alpha}}\geq |x-y|$, we can get
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq&C_{N}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N}\int^{\infty}_{|x-y|^{2\alpha}}t^{-n/\alpha-1}dt \\
&\leq& \frac{C_{N}}{|x-y|^{2n}}\Big(1+\frac{|x-y|}{\rho(x)}+\frac{|x-y|}{\rho(y)}\Big)^{-2N},
\end{eqnarray*}
which proves (i).
For (ii), By Proposition \ref{pro3.13}, we have
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq& \int^{\infty}_{0}\frac{C_{N}t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{2n+4\alpha}}\Big(
\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}dt \\
&\leq& C_{N}|y-z|^{2\delta}\int^{\infty}_{0}\frac{(\sqrt{t^{1/\alpha}})^{2\alpha-2\delta}}{(\sqrt{t^{1/\alpha}}+|x-y|)^{2n+4\alpha}}dt.
\end{eqnarray*}
Let ${\sqrt{t^{1/\alpha}}}/{|x-y|}=u$. We obtain
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})} &\leq&C_{N}|y-z|^{2\delta}|x-y|^{-2n-2\delta}\int^{\infty}_{0}\frac{u^{4\alpha-2\delta-1}}{(1+u)^{2n+4\alpha}}du \\
&\leq& \frac{C_{N}|y-z|^{2\delta}}{|x-y|^{2n+2\delta}}.
\end{eqnarray*}
The symmetry of the kernel $D^{\mathcal{L}}_{\alpha,t}(\cdot,\cdot)$ gives the desired conclusion of (ii).
For (iii), fix $y,z\in B=B(x_{0},r)$ with $0<r\leq \rho(x_{0})/2$. Similar to Theorem \ref{the3.1}, we need to deal with the term $$\Big\|t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|_{L^{2}((0,\infty),\frac{dt}{t})}$$ first. Write
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
\Big\|t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})}
% &=&\int^{\infty}_{0}\Big|\int_{\mathbb{R}^{n}}(\widetilde{D}^{L}_{\alpha,t}(x,y)-\widetilde{D}^{L}_{\alpha,t}(x,z))dx\Big|^{2}\frac{dt}{t} \\
&=& G_{1}+G_{2}+G_{3},
\end{eqnarray*}
where
\begin{equation*}
\left\{ \begin{aligned}
G_{1}&:=\int^{(2r)^{2\alpha}}_{0}\Big|\int_{\mathbb{R}^{n}}\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t};\\
G_{2}&:=\int^{\rho(x_{0})^{2\alpha}}_{(2r)^{2\alpha}}\Big|\int_{\mathbb{R}^{n}}\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t};\\
G_{3}&:=\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big|\int_{\mathbb{R}^{n}}\Big(\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,y)-\widetilde{D}^{\mathcal{L}}_{\alpha,t}(x,z)\Big)dx\Big|^{2}\frac{dt}{t}. \end{aligned}\right.
\end{equation*}
Since $y,z\in B\subset B(x_{0},\rho(x_{0}))$, then $\rho(y)\sim\rho(x_{0})\sim\rho(z)$. It follows from Proposition \ref{pro3.13} (iii) that
$$G_{1}\leq C\int^{(2r)^{2\alpha}}_{0}(\sqrt{t^{1/\alpha}}/\rho(x_{0}))^{1+2\alpha}\frac{dt}{t}
\leq C\int^{(2s)^{2\alpha}}_{0}\Big(\frac{\sqrt{t^{1/\alpha}}}{\rho(x_{0})}\Big)^{2\delta}\frac{dt}{t}=C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}.$$
Also, we apply Proposition \ref{pro3.13} (ii) to deduce that
\begin{eqnarray*}
% \nonumber to remove numbering (before each equation)
G_{3} &\leq& C\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}\Big|\int_{\mathbb{R}^{n}}
\frac{t}{(\sqrt{t^{1/\alpha}}+|x-y|)^{n+2\alpha}}dx\Big|^{2}\frac{dt}{t} \\
&\leq& C\int^{\infty}_{\rho(x_{0})^{2\alpha}}\Big(\frac{|y-z|}{\sqrt{t^{1/\alpha}}}\Big)^{2\delta}\frac{dt}{t}\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}.
\end{eqnarray*}
Then for $G_{2}$, following the procedure of the treatment for $M_{2}$ in Theorem \ref{the3.1}, we obtain
$$G_{2}\leq \int^{\rho(x_{0})^{2\alpha}}_{(2r)^{2\alpha}}\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}\frac{dt}{t}=C\Big(\frac{r}{\rho(x_{0})}\Big)^{2\delta}
\log\Big(\frac{\rho(x_{0})}{r}\Big).$$
From the above estimates, we can get
$$\Big\|t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(y)-t^{1/(2\alpha)}\nabla_{x}e^{-t\mathcal{L}^{\alpha}}1(z)\Big\|^{2}_{L^{2}((0,\infty),\frac{dt}{t})}
\leq C\Big(\frac{r}{\rho(x_{0})}\Big)^{\delta}\Big(\log(\frac{\rho(x_{0})}{r})\Big)^{1/2}.$$
Thus (iii) readily follows.
\end{proof}
%\begin{acknowledgements}
%If you'd like to thank anyone, place your comments here
%and remove the percent signs.
%\end{acknowledgements}
% Authors must disclose all relationships or interests that
% could have direct or potential influence or impart bias on
% the work:
%
% \section*{Conflict of interest}
%
% The authors declare that they have no conflict of interest.
% BibTeX users please use one of
%\bibliographystyle{spbasic} % basic style, author-year citations
%\bibliographystyle{spmpsci} % mathematics and physical sciences
%\bibliographystyle{spphys} % APS-like style for physics
%\bibliography{} % name your BibTeX data base
% Non-BibTeX users please use
\begin{thebibliography}{}
\bibitem{bet} Betancor, J., Crescimbeni, R., Fari\~{n}a, J., Stinga, P., Torrea, J.: A $T1$ criterion for Hermite-Calder\'{o}n-Zygmund operators on the $BMO_{H}(\mathbb{R}^{n})$ space and applications. Ann. Sc. Norm. Sup. Pisa Cl.
Sci. {\bf 12}, 157-187 (2013)
\bibitem{BDS} Bennett, C., DeVore R.A., Sharpley, R.: {Weak-$L^{\infty}$ and BMO}, Ann. Math. \textbf{113}, 601-611 (1981)
\bibitem{BDY} Bui, H., Duong, X., Yan, L.: Calder\'on reproducing formulas and new Besov spaces associated with operators. Adv. Math. \textbf{229}, 2449-2502 (2012)
\bibitem{duo} Duong, X., Yan, L., Zhang, C.: On characterization of Poisson integrals of Schr\"{o}dinger operators with $BMO$ traces. J. Funct. Anal. {\bf 266}, 2053-2085 (2014)
\bibitem{DGMTZ} Dziuba\'{n}ski, J., Garrig\'os, G., Mart\'inez, T., Torrea, J., Zienkiewicz, J.: BMO spaces related to Schr\"{o}dinger operators with potentials satisfying a reverse H\"{o}lder inequality. {Math. Z.} \textbf{249}, 329-356 (2005)
\bibitem{DZ3} Dziuba\'{n}ski, J., Zienkiewicz, J.: Hardy space $ H^ 1$ associated to Schr\"{o}dinger operator with
potential satisfying reverse H\"{o}lder inequality. Rev. Mat. Iberoam. {\bf 15}, 279-296 (1999)
\bibitem{DZ2} Dziuba\'{n}ski, J., Zienkiewicz, J.: $H^{p}$ Spaces for Schr\"odinger operators. Fourier Analysis and Related Topics, Banach Center Publications, {\bf 56}, 45-53 (2002)
\bibitem{DZ1} Dziuba\'{n}ski, J., Zienkiewicz, J.:$H_p$ spaces associated with Schr\"odinger operators with potentials from reverse H\"older classes. Colloq. Math. {\bf 98} 5--36 (2003)
\bibitem{gri} Grigor'yan, A.: Heat kernels and function theory on metric measure spaces. Contemp. Math. {\bf 338}, 143-172 (2002)
\bibitem{Kurata}Kurata, K.: An estimate on the heat kernel of magnetic Schr?dinger operators and uniformly elliptic operators with non-negative potentials. J. London Math. Soc. {\bf 62}, 885--903 (2000)
\bibitem{Li2} Li, P., Wang, Z., Qian, T., Zhang, C.: Regularity of fractional heat semigroup associated with Schr\"{o}dinger operators. preprint available at arXiv:2012.07234.
\bibitem{LL} Lin, C., Liu, H.: $BMO_{L}(\mathbb H^{n})$ spaces and Carleson measures for Schr\"odinger operators. Adv. Math. \textbf{228}, 1631-1688 (2011)
\bibitem{ma1} Ma, T., Stinga, P., Torrea, J., Zhang, C.: Regularity properties of Schr\"{o}dinger operators. J. Math. Anal. Appl. {\bf 338},
817-837 (2012)
\bibitem{ma2} Ma, T., Stinga, P., Torrea, J., Zhang, C.: Regularity estimates in H\"{o}lder spaces for Schr\"{o}dinger operators via $T1$ theorem. Ann. Mat. Pur. Appl. {\bf 193}, 561-589 (2014)
\bibitem{she} Shen, Z.: $L^{p}$ estimates for Schr\"{o}dinger operators with certain potentials. Ann. Inst. Fourier. {\bf 45}, 513-546 (1995)
\bibitem{15} Stein, E.: Harmonic Analysis: Real Variable Methods, Orthogonality and Oscillatory Integrals.
Princeton Math. Serises {\bf 43}, Princeton Univ. Press, (1993).
\bibitem{wyx}
Wang, Y., Liu, Y., Sun, C., Li, P.: Carleson measure characterizations of the Campanato type space associated with Schr\"{o}dinger operators on stratified Lie groups. Forum Math. {\bf 32}, 1337-1373 (2020)
\bibitem{jia} Yang, Da., Yang, Do., Zhou, Y.: Localized Campanato-type spaces related to admissible functions on $RD$-spaces and
applications to Schr\"{o}odinger operators. Nagoya Math. J. {\bf 198}, 77-119 (2010)
\bibitem{yang} Yang, Da., Yang, Do., Zhou, Y.: Localized $BMO$ and $BLO$ spaces on $RD$-spaces and applications to Schr\"{o}dinger operators. Commu. Pure Appl. Anal. {\bf 9}, 779-812 (2010)
\bibitem{Zhang} Zhang, C., Torrea, J.: Boundedness of differential transforms for heat semigroups generated by Schr\"{o}dinger operators. To appear in Canad. J. Math.
\end{thebibliography}
\end{document}
% end of file template.tex