1: \begin{abstract}

2: In this paper, we consider $d$-dimensional mixing random fields $\bigl(X^{\scalebox{0.6}{$(n)$}}_{i}\bigr)_{i\in T_{n}}$ and study the convergence of the empirical average $W_n:=\sigma_n^{-1} \sum_{i\in T_n}X^{\scalebox{0.6}{$(n)$}}_{i}. $ Under $\alpha$-mixing and moment conditions, we obtain smooth Edgeworth expansions for $W_n$ of any order $k\ge 1$ with better controlled remainder terms. We exploit this to obtain rates for the convergence of $W_n$ to its normal limit in the Wasserstein-$p$ distance for any $p\geq 1$. The bounds depend on the moments of the random variables $\bigl(X^{\scalebox{0.6}{$(n)$}}_{i}\bigr)_{i\in T_{n}}$ and the rate of polynomial decay of the mixing coefficients. Finally, we apply those results to obtain tail inequalities and non-uniform Berry--Esseen bounds with polynomial decay. 

3: \end{abstract}

4: