Bers and Hénon, Painlevé and Schrödinger (Q731744)

This nice paper shows how to apply techniques of holomorphic dynamics of several complex variables in an effective way to study the dynamics of mapping classes, even in the real case. Let \(S^2_4\) be the four-punctured sphere. The set of representations of the fundamental group of \(S^2_4\) in \(SL(2,\mathbb{C})\) modulo conjugations can be identified with a quartic hypersurface \(S\) of \(\mathbb{C}^7\), the character variety of \(S^2_4\). The mapping class group MCG of \(\pi_1(S^2_4)\) (that is, the group of outer automorphisms of the fundamental group of \(S^2_4\)) acts polynomially on \(S\). Furthermore, there is a finite index subgroup \(\Gamma^*_2\) of MCG acting on \(S\) so to preserve a family of cubic surfaces \(S_{(A,B,C,D)}\) foliating \(S\), where \((A,B,C,D)\in\mathbb{C}^4\). The group \(\Gamma^*_2\) can be identified with a finite index subgroup of \(PGL(2,\mathbb{Z})\), and \(S_{(A,B,C,D)}\) can be identified with the cubic of \(\mathbb{C}^3\) of equation \(x^2+y^2+z^2+xyz=Ax+By+Cz+D\); and it is known that the group~\(\mathcal{A}\) of automorphisms of \(S_{(A,B,C,D)}\) induced by \(\Gamma^*_2\) has finite index in the group of all polynomial automorphisms of \(S_{(A,B,C,D)}\). A similar construction can be carried out with the once-punctured torus \(T_1\); in this case the character variety is \(\mathbb{C}^3\), and it is foliated by the cubic surfaces \(S_D=S_{(0,0,0,D)}\). The aim of this paper is to study the dynamics of the elements of \(\mathcal{A}\) on \(S_{(A,B,C,D)}\), using the techniques developed by Bedford, Smillie and others to study the dynamics of polynomial automorphisms of the complex plane. The results are quite complete and detailed, with interesting applications. For instance, it allows the author to answer a question raised by Dumas and Goldman giving an example of a representation of the group \(G=\langle\alpha, \beta\mid[\alpha,\beta]^4\rangle\) in \(SL(2,\mathbb{C})\) (corresponding to a point in~\(S_2\)) whose orbit under the action of \(\Gamma^*_2\) contains both (the image in \(S_2\) of) a discrete faithful representation of \(G\) and the (image in \(S_2\) of the) whole set of \(SU(2)\)-representations of \(G\). The techniques developed by Bedford and Smillie can also be used to study real Hénon maps; accordingly, they can be (non-trivially) adapted to study the dynamics of real mapping classes. When \(D\in\mathbb{R}\), let \(S_D(\mathbb{R})\) denote the real part of \(S_D\). Then (among several other things) the author shows that if \(f\in\Gamma^*_2\) is hyperbolic with spectral radius \(\lambda(f)>1\), then the topological entropy of the action of \(f\) on~\(S_D(\mathbb{R})\) (still denoted by \(f\)) is bounded above by \(\log\lambda(f)\). Moreover, the following properties are equivalent: (i) the topological entropy of \(f\) on \(S_D(\mathbb{R})\) is equal to \(\log\lambda(f)\); (ii) all periodic points of \(f\) on \(S_D\) are contained in \(S_D(\mathbb{R})\); (iii) the topological entropy of \(f\) on \(S_D(\mathbb{R})\) is positive and the dynamics of \(f\) on the subset \(K(f,\mathbb{R})\) of \(S_D(\mathbb{R})\) of points with bounded forward and backward orbits is uniformly hyperbolic; (iv) \(S_D(\mathbb{R})\) is connected; (v) \(D\geq 4\). Finally, the author also uses these techniques to study the Hausdorff dimension of the spectrum of discrete Schrödinger operators, and to give examples of Painlevé VI equations with rich monodromy, answering a question of Iwasaki and Uehara.