The Picard group of the moduli space of \(r\)-spin Riemann surfaces (Q436181): Difference between revisions

Fix a genus \(g\) smooth surface \(\Sigma_g\). Teichmüller space, denoted \(\mathcal{T}(\Sigma_g)\), is the moduli space of complex structures on \(\Sigma_g\). By the Uniformization Theorem, if \(g\geq 2\) it is the space of hyperbolic structures on \(\Sigma_g\); in this case it is homeomorphic \(\mathbb{R}^{6g-6}\). The mapping class group, denoted \(\mathrm{Mod}(\Sigma_g)\), acts properly discontinuously on \(\mathcal{T}(\Sigma_g)\) with finite stabilizers, and so the quotient \(\mathcal{M}_g\), the moduli space of Riemann surfaces of genus \(g\), is an aspherical orbifold. The first part of this paper describes two analogous constructions for possible moduli spaces of \textit{r-Spin Riemann surfaces of fixed genus}. The first is denoted by \(\tilde{\mathcal{M}}_g(r,\zeta)\). It is constructed as follows: \(\mathrm{Spin}^r(\Sigma_g)\) consists of pairs \(\zeta=(L,\phi)\) where \(L\) is complex line bundle over \(\Sigma_g\) and \(\phi\) is a vector bundle isomorphism between \(L^{\otimes r}\) and \(T\Sigma_g\). Then \(\tilde{\mathcal{M}}_g(r,\zeta)\) is the quotient of \(\mathcal{T}(\Sigma_g)\) by the subgroup of \(\mathrm{Mod}(\Sigma_g)\) that preserves \(\zeta\in \mathrm{Spin}^r(\Sigma_g)\). The projection from \(\tilde{\mathcal{M}}_g(r,\zeta)\) to \(\mathcal{M}_g\) is a finite covering map since the subgroup fixing \(\zeta\) is of finite index. The authors give a competing definition for this moduli space using a different but related quotient of Teichmüller space; it is denoted \(\mathcal{M}_g(r,\zeta)\). In particular, the action is given by a central extension of the subgroup fixing \(\zeta\) with kernel \(\mathbb{Z}/r\mathbb{Z}\). There is then a natural mapping \(\mathcal{M}_g(r,\zeta)\to \tilde{\mathcal{M}}_g(r,\zeta)\). Both these moduli spaces are complex orbifolds. The author takes the position that \(\mathcal{M}_g(r,\zeta)\) is the moduli space of \(r\)-Spin Riemann surfaces because of its universal modular description, and calls \(\tilde{\mathcal{M}}_g(r,\zeta)\) the \textit{moduli space of \(r\)-theta characteristics} because of the \(r=2\) case. The first described result of this paper gives a description of the rational cohomology of \(\mathcal{M}_g(r,\zeta)\) in a range of degrees (which increases in \(g\)), kin to the description by Madsen and Weiss for \(\mathcal{M}_g\) which proved the Mumford Conjecture. The Picard group for a moduli space \(X/\!/G\) is the group (under tensor product) of isomorphism classes of \(G\)-equivariant line bundles on \(X\). As the author indicates, ``the main result of this paper is the computation of algebraic Picard group of \(\mathcal{M}_g(r,\zeta)\) and a determination of a presentation for it, analogous to that for \(\mathcal{M}_g\).'' This is achieved, in part, by explicit computation of the low-dimensional integral (co)homology of \(\mathcal{M}_g(r,\zeta)\). The author also describes relationships of his work to previous related work.