1: \begin{abstract}

2: \setlength{\baselineskip}{1.8em}

3: Nonparametric density estimators are studied for $d$-dimensional, strongly spatial mixing data which is defined on a general $N$-dimensional lattice structure. We consider linear and nonlinear hard thresholded wavelet estimators which are derived from a $d$-dimensional multiresolution analysis. We give sufficient criteria for the consistency of these estimators and derive rates of convergence in $L^{p'}$ for $p'\in [1,\infty)$. For this reason, we study density functions which are elements of a $d$-dimensional Besov space $B^s_{p,q}(\R^d)$. We also verify the analytic correctness of our results in numerical simulations.\medskip\\

4: \noindent {\bf Keywords:} Besov spaces; Density estimation; Hard thresholding; Rate of convergence; Spatial lattice processes; Strong spatial mixing; Wavelets

5: 

6: \noindent {\bf MSC 2010:} Primary: 62G07; 62H11; 65T60; Secondary: 65C40; 60G60

7: \end{abstract}