On the nondensity of integral points (Q541155)

In this paper, the author continues his investigation of the Lang-Vojta conjecture and modifications. In [``Generalizations of Siegel's and Picard's theorems'', Ann. Math. (2) 170, No. 2, 609--655 (2009; Zbl 1250.11067)], \textit{A. Levin} made the following conjecture: ``Let \(k\) be a number field, \(S\) be a finite set of places of \(k\) and \(X/k\) be a projective variety of dimension \(d\geq 1\). Let \(D_1,\dots,D_r\) be big (for example, ample) effective divisors on \(X\) intersecting properly, with \(r\geq d+2\). Let \(Y=X\setminus D_1\cup\dots\cup D_r\). Then the set of \(S\)-integers of \(Y\) is not Zariski-dense in \(Y\).'' \textit{P. Corvaja, A. Levin} and \textit{U. Zannier} [``Integral points on threefolds and other varieties'', Tohoku Math. J. (2) 61, No. 4, 589--601 (2009; Zbl 1250.11066)] proved the conjecture for \(d^2-d+1\) ample divisors on a smooth variety of dimension \(d\geq 3\). Here, the author improves on this result by showing that the conjecture is true for a variety of dimension \(d\geq 2\) and \(2d\) ample effective divisors intersecting properly. For the proof he uses the method introduced by Corvaja and Zannier for curves and surfaces (with the essential use of the subspace theorem), but adds a new ingredient, namely, a convexity theorem for global sections of certain coherent sheaves.