Rank one orbit closures in \(\mathcal{H}^{\mathrm{hyp}}({g}-1,{g}-1)\) (Q2009021)

The moduli space \(\Omega M_g\) of pairs \((X, \omega)\), where \(X\) is a compact genus \(g\) Riemann surface and \(\omega\) a holomorphic \(1\)-form is stratified by the combinatorial data of the zeros of \(\omega\). Each integer partition of \(2g-2\) determines a stratum, and the strata \(\mathcal H(g-1, g-1)\) and \(\mathcal H(2g-2)\) are important families. Inside these strata are connected components consisting of hyperelliptic surfaces, surfaces which admit an involution whose quotient yields a sphere. There is an \(\mathrm{GL}(2, \mathbb R)\) action on \(\Omega M_g\) which comes from viewing \((X, \omega)\) as yielding an atlas of charts to \(\mathbb C\) (away from zeros of \(\omega\)) whose transition maps are translations, and the group acts by \(\mathbb R\)-linear postcomposition with charts. The seminal work of Eskin-Mirzakhani-Mohammadi [\textit{A. Eskin} et al., Ann. Math. (2) 182, No. 2, 673--721 (2015; Zbl 1357.37040)] showed that all \(\mathrm{GL}(2, \mathbb R)\)-orbit closures are swept out by linear equations in appropriate period coordinates. The main result of this paper is that in these particular families of hyperelliptic strata, each orbit is either closed, dense in the stratum, or contained in a locus of branched covers. Key tools include the study of the isoperiodic foliation and its interaction with the unipotent upper-triangular part of the \(\mathrm{GL}(2, \mathbb R)\)-action, the horocycle flow.