Nef vector bundles on a projective space or a hyperquadric with the first Chern class small (Q2154692)

A vector bundle \(\mathcal{E}\) on a projective variety \(X\) is said to be \textit{numerically effective} (\textit{nef}) if the relative very ample line bundle \(\mathcal{O}_{\mathbb{P}(\mathcal{E})}(1)\) on \(\mathbb{P}(\mathcal{E})\) is nef. \textit{T. Peternell} et al. [Lect. Notes Math. 1507, 145--156 (1992; Zbl 0781.14006)] give a complete classification of nef vector bundles with \(c_1(\mathcal{E})= 0, 1, 2\) on the projective space \(\mathbb{P}^n\) and with \(c_1(\mathcal{E})=0, 1\) on a smooth hyperquadric \(\mathbb{Q}^n\) \((n \geq 3)\), by using Mori theory and analyzing the contraction morphism of an extrema ray of the Fano manifold \(\mathbb{P}(\mathcal{E})\). In the paper under review, the author gives a new proof of the these results, which depends only on the cohomological study of \(\mathcal{E}\) with respect to a full strong exceptional sequence of vector bundles. The auther also classify nef vector bundles containing a line bundle of next-to-maximal degree on a projective space. Futhermore, the author gives an upper bound of the second Chern class of a nef vector bundle having a global section whose zero locus contains a divisor on a surface. This upper bound is so sharp that the cases (18), (19), and (20) in [\textit{M. Ohno}, Rend. Circ. Mat. Palermo (2) 69, No. 2, 425--458 (2020; Zbl 1451.14129), Theorem 1], which were uncertain to exist, can be ruled out.