On the Shafarevich conjecture for surfaces of general type over function fields (Q1977009)

Let \(B\) be a projective algebraic curve over an algebraically closed field, and let \(S\subset B\) a finite set of points. The so-called Shafarevich conjecture for families of curves, proved in the 1970s by \textit{A. Parshin} [Sov. Math. Dokl. 9, 1419--1422 (1968; Zbl 0176.50903)] and \textit{A. Arakelov} [Math. USSR, Izv. 5, 1277--1302 (1971; Zbl 0248.14004)], states the following: (i) There are only finitely many isomorphism classes of smooth and non-isotrivial families of curves over \(B- S\). (ii) If \(2\cdot\text{genus}(B)- 2+ \text{card}(S)\leq 0\), then there are no such families at all. The paper under review is devoted to analogous questions for smooth families of higher-dimensional algebraic manifolds over \(B- S\) which are currently subject to intensive study by various authors. Due to \textit{G. Faltings's} work on Arakelov's theorem for abelian varieties [Invent. Math. 73, 337--347 (1983; Zbl 0588.14025)], statement (i) seems to be too much to hope for in the higher-dimensional case, and so the authors propose to split it up into two separate questions regarding boundedness and rigidity. Then, in this note, they prove their formulated boundedness conjecture for surfaces of general type, and also for canonically polarized manifolds in the special case of \(S=\emptyset\). The method of proof is based on particular ampleness conditions and cohomological vanishing theorems for suitable sheaves on the manifolds under consideration. Some earlier partial results in this context, obtained by several authors, are reproved within the authors' general framework.