Anti-holomorphic multiplication and a real algebraic modular variety (Q1880409)

Let \(A\) be a principally-polarized abelian variety over the complex numbers \(\mathbb C\) and let \(\mathcal O\) be the ring of integers of an imaginary quadratic field. The authors say that \(A\) has anti-holomorphic multiplication by \(\mathcal O\) if there is an action of \(\mathcal O\) on \(A\) by real endomorphisms such that the purely imaginary elements of \(\mathcal O\) act anti-holomorphically. In this case, the dimension of \(A\) must be even. Fix integers \(n>0\) and \(M>2\), and let \(X\) be the usual Siegel moduli space of principally-polarized \(2n\)-dimensional abelian varieties with level-\(M\) structure. Let \(\mathcal O\) be the ring of integers of an imaginary quadratic field, and assume that \(2\) ramifies in \(\mathcal O\) if \(M\) is odd. The authors show that there is an anti-holomorphic involution \(\tau\) on \(X\) such that \(X_{\mathbb R}\) is the coarse moduli space of principally-polarized \(2n\)-dimensional abelian varieties with level-\(M\) structure and with anti-holomorphic multiplication by \(\mathcal O\). They show that \(X_{\mathbb R}\) has finitely many components, all isomorphic to one another, and that each component can be written as an arithmetic quotient \(W\) of quaternionic Siegel space \(\text{Sp}(2n,{\mathbb C}) \setminus {U}(n,{\mathbb H})\). Thus, although the arithmetic quotient \(W\) does not have any obvious algebraic structure, a finite union of copies of \(W\) \textit{does} have a natural algebraic structure. Furthermore, the authors show that variety \(X\) and the involution \(\tau\) can be defined over the rational numbers.