Pseudo-Anosov homeomorphisms on translation surfaces in hyperelliptic components have large entropy (Q412048)

Let \(S\) be a closed orientable surface of genus \(g\). The associated mapping class group \textbf{Mod}\(_{g}\) is the group of homotopy classes of orientation-preserving homeomorphisms of \(S\). Following the Thurston-Nielsen classification, mapping classes are either reducible, periodic, or pseudo-Anosov, the pseudo-Anosovs being by far the richest class. Associated to each pseudo-Anosov mapping class \(f\) is a dilatation factor \(\delta(f)\), which records the exponential growth rate of lengths of curves under iteration of \(f\), in any fixed metric. While \(\delta(f) \) is known to be an algebraic integer [\textit{W. P. Thurston}, Bull. Am. Math. Soc., New Ser. 19, No. 2, 417--431 (1988; Zbl 0674.57008)], and the set \(\{ \delta(f) \mid f \in\) \textbf{Mod}\(_{g}\) is pseudo-Anosov \(\}\) discrete as a subset of \(\mathbb{R}\) [\textit{P. Arnoux} and \textit{J.-C. Yoccoz}, C. R. Acad. Sci., Paris, Sér. I 292, 75--78 (1981; Zbl 0478.58023)] and [\textit{N. V. Ivanov}, J. Sov. Math. 52, No. 1, 2822--2824 (1990; Zbl 0707.55006)], further characterization of this set remains an open question. Of particular interest is the least dilatation for a fixed genus, which we refer to by \(\delta_g\). While many results are known for upper bounds, establishing lower bounds seems somewhat more difficult. One particular result, due to \textit{R. C. Penner} [Proc. Am. Math. Soc. 113, No. 2, 443--450 (1991; Zbl 0726.57013)], is that \(log(\delta_g) \geq \frac{log(2)}{12g-12}\). The paper under review is concerned with a restricted version of this minimization problem, as follows. A pseudo-Anosov homeomorphism on a compact surface \(S\) uniquely defines a pair \((X,q)\), where \(X\) is a Riemann surface homeomorphic to \(S\), and \(q\) a non-zero holomorphic differential. A ``stratum'' of the unit cotangent bundle of the moduli space \(\mathcal{M}_g\) of compact genus-\(g\) Riemann surfaces is then the moduli space of quadratic differentials with prescribed multiplicities of the zeros of the differentials. \textit{B. Farb} [Proc. Symp. Pure Math. 74, 11--55 (2006; Zbl 1191.57015)] proposed the restriction of the minimization problem to a particular stratum. The paper under review considers a somewhat more restricted setting. Firstly, the authors are concerned only with translation surfaces, i.e., those pairs \((X,q)\) where \(q\) is the global square of an abelian differential \(\omega \in H^1(X,\mathbb{C})\). Moreover, they restrict to the hyperelliptic components of strata, where hyperellipticity means here that each element \((X,\omega)\) of the component is not only such that \(X\) is hyperelliptic, but also such that \(\omega\) has at most two zeros, which, in the case that there are two, are permuted by the hyperelliptic involution. The main result of the paper is then that the dilatation of any pseudo-Anosov homeomorphism on a translation surface that belongs to a hyperelliptic component is uniformly bounded from below by \(\sqrt{2}\). The method of proof relies on the Rauzy-Veech induction. Moreover, in an appendix, the authors show via examples that this bound is sharp, more precisely that the least dilatation of a pseudo-Anosov homeomorphism on a genus-\(g\) surface in a hyperelliptic component belongs to the interval \((\sqrt{2},\sqrt{2} + 2^{1-g})\). A second appendix shows, again via examples, the necessity of the assumptions concerning the zeros of \(\omega\) in the definition of hyperelliptic components.