Renormalization of polygon exchange maps arising from corner percolation (Q1941066)

The author studies dynamical properties of a two-dimensional parameter family of polygon exchange maps. The considered family consists of rectangle exchange maps, where the polygons are rectangles with horizontal and vertical sides. To define these maps, \(\Lambda\) denotes a planar lattice generated by the vectors \((\frac{1}{2}, \frac{1}{2})\) and \((-\frac{1}{2}, \frac{1}{2}),\) \(Y\) denotes the torus \(\mathbb{R}^2/\Lambda\), the union of the two squares \(A_1= [0, 1/2) \times [0, 1/2)\) and \(A_{-1}=[0, 1/2) \times [ 1/2, 1)\) is a fundamental domain of the action of \(\Lambda\) by translations on \(\mathbb{R}^2\), and \(N\) is the finite set \(\{(1,0), (-1,0), (0,1), (0,-1)\}\). For parameter values \(\alpha\) and \(\beta\) in the interval \([0,1/2],\) \(\psi_{\alpha, \beta}: Y \times N \rightarrow Y \times N\) is given by \[ \psi_{\alpha, \beta} ((x,y), \mathbf{v}) =((x+bs\alpha, y +a s \beta)\, \mod \Lambda ,(bs,as)), \] with \(s \in\{\pm 1\}\) and \(\mathbf{v}=(a,b)\). This family exhibits very interesting dynamical behavior, we list the following results from the paper: {\parindent=6mm \begin{itemize}\item[1.] If \(\alpha\) and \(\beta \) are irrational, \(\psi_{\alpha, \beta}\) has periodic points of arbitrarily large periods; \item[2.] Every periodic point has an open neighborhood of points which are periodic and of the same period; \item[3.] Let \(\lambda\) be the normalized Lebesgue measure on \(Y \times N\) and \(M(\alpha, \beta)\) be the \(\lambda\)-measure of the set of all periodic points. {\parindent=12mm \begin{itemize}\item[(a)] Then the following holds: \(\psi_{\alpha, \beta}\) is periodic almost everywhere, i.e., \(M(\alpha, \beta)=1\) for almost every parameter vector \((\alpha, \beta) \in [0,1/2]\times [0,1/2]\). \item[(b)] For every \(\epsilon >0\) there are irrational parameters \((\alpha, \beta)\) such that \(M(\alpha, \beta)< \epsilon\).\item[(c)] There is a dense set of irrational parameters \((\alpha, \beta)\) such that \(M(\alpha, \beta)\neq 1.\) \end{itemize}}\end{itemize}} The techniques used in the proofs are broad and deep. These include a renormalization method of rectangle exchange maps, the ``arithmetic graph'' which is a combinatorial tool developed by the author related to Tóth's percolation model, and a new interpretation of the space of Truchet tilings as a dynamical system, just to name a few. In the introduction, Subsection 1.6, the author gives a very useful summary of the main points of each one of the remaining eight sections in the paper. There are several questions left unanswered, as for example: What is the Hausdorff dimension of \(\{(\alpha, \beta): \, M(\alpha, \beta)\neq 1\}?\)