Algebraic surfaces of general type with \(p_g=q=1\) and genus 2 Albanese fibrations (Q2311334)

This work is about algebraic surfaces of general type with \(p_g=q=1\) admitting a genus \(2\) Albanese fibration. According to Xiao we have \(2 \leq K^2 \leq 6\) for such surfaces. Thanks to the work of Catanese, Horikawa, Ciliberto et al., there is a good understanding if \(K^2 \leq 4\). For \(K^2=5\) however, few is known, apart from some examples constructed by Catanese and by Ishida. In the case \(K^2=6\) there are no examples at all. The author of the paper analyzes Catanese's examples, which were constructed as bidouble covers of the projective plane. More precisely he proves that these surfaces form a 3-dimensional irreducible and connected component of \(\mathcal M^{5,2}_{1,1}\), the Gieseker moduli space of surfaces of general type with \(p_g=q=1\), \(K^2=5\) and genus \(2\) Albanese fibration. Let \(S\) be a surface of general type with \(p_g=q=1\) and let \(\alpha\) be the Albanese map. For such surfaces Catanese and Ciliberto have shown that the number of direct summands of \(\alpha_{\ast}\omega_S\) is invariant under deformation. The author proves that the analog statement for \(\alpha_{\ast}\omega_S^{\otimes 2}\) fails, and thus gives negative answer to Pignatelli's question.