Classification of quadruple Galois canonical covers. II. (Q2370246)

This paper is the second and last part of the classification of quadruple Galois canonical covers of surfaces of minimal degree. In other words, in these two papers the authors classify minimal surfaces of general type whose canonical map is Galois of degree \(4\) and whose canonical image is a surface of minimal degree. Recall that \textit{E. Horikawa} [Ann. Math. (2) 104, 357--387 (1976; Zbl 0339.14024); Invent. Math. 47, 209--248 (1978; Zbl 0409.14005)] and \textit{K. Konno} [Math. Ann. 290, No.~1, 77--107 (1991; Zbl 0711.14021)] have already classified these covers when the degree is 2 or 3, that explains why the authors look at the degree \(4\) case. In the first part, unpublished at the time when this review is written [Classification of quadruple Galois canonical covers I, \texttt{math.AG/0302045}], they do the classification under the further assumption that the canonical image is smooth. This paper deals with the singular case. The result (in sharp contrast with the result in the smooth case) is that there are very few possibilities. More precisely there are exactly 4 families, two with Galois group \({\mathbb Z}_2 \times {\mathbb Z}_2\) and two with Galois group \({\mathbb Z}_4\). All surfaces are regular with \(p_g=4\) and their canonical image is a quadric cone in \({\mathbb P}^3\). Note that in the smooth case they could construct examples with arbitrarily high \(p_g\) and arbitrarily high irregularity. In particular, looking at the resulting (together with the above mentioned results of Horikawa and Konno) complete classification of Galois canonical cover of degree \(d\leq 4\) of a surface of minimal degree, one notes that in the singular case \(p_g\leq 8-d\): this observation leads them naturally to conjecture that there are no other Galois canonical covers of singular surface of minimal degree.