arXiv subjects
Publications and source records attributed to John Kopper.
We use hypersurfaces containing unexpected linear spaces to construct interesting vector bundles on complete intersection surfaces in projective space. We discover examples of moduli spaces of rank 2 stable bundles on surfaces of Picard rank one with arbitrarily many connected components.
We study ample stable vector bundles on minimal rational surfaces. We give a complete classification of those moduli spaces for which the general stable bundle is both ample and globally generated. We also prove that if $V$ is any stable bundle, then a large enough direct sum $V^{\oplus n}$ has ample deformations unless there is an obvious numerical reason why it cannot. Previous work in this area has mostly focused on rank two bundles and relied primarily on classical constructions such as the Serre construction. In contrast, we use recent advances in moduli of vector bundles to obtain strong results for vector bundles of any rank.
Computing the cohomology of the tensor product of two vector bundles is central in the study of their moduli spaces and in applications to representation theory, combinatorics and physics. These computations play a fundamental role in the construction of Brill-Noether loci, birational geometry and $S$-duality. Using recent advances in the Minimal Model Program for moduli spaces of sheaves on $\mathbb{P}^2$, we compute the cohomology of the tensor product of general semistable bundles on $\mathbb{P}^2$. This solves a natural higher rank generalization of the polynomial interpolation problem. More precisely, let $v$ and $w$ be two Chern characters of stable bundles on $\mathbb{P}^2$ and assume that $w$ is sufficiently divisible depending on $v$. Let $V \in M(v)$ and $W \in M(w)$ be two general stable bundles. We fully compute the cohomology of $V \otimes W$. In particular, we show that if $W$ is exceptional, then $V \otimes W$ has at most one nonzero cohomology group determined by the slope and the Euler characteristic, generalizing foundational results of Dr\'{e}zet, G\"{o}ttsche and Hirschowitz. We characterize the invariants of effective Brill-Noether divisors on $M(v)$. We also characterize when $V\otimes W$ is globally generated. Our computation is canonical given the birational geometry of the moduli space, suggesting a roadmap for tackling analogous problems on other surfaces.
We describe the locus of stable bundles on a smooth genus $g$ curve that fail to be globally generated. For each rank $r$ and degree $d$ with $rg<d<r(2g-1)$, we exhibit a component of the expected dimension. We show moreover that no component has larger dimension and give an explicit description of those families of smaller dimension than expected. For large enough degrees, we show that the locus is irreducible.
We compute the nef cone of the Hilbert scheme of points on a general rational elliptic surface. As a consequence of our computation, we show that the Morrison-Kawamata cone conjecture holds for these nef cones.
Using Bridgeland stability conditions we give sufficient criteria for a stable vector bundle on a surface to remain stable when restricted to a curve. We give a stronger criterion when the vector bundle is a general vector bundle on the plane. As an application, we compute the cohomology of such bundles for curves that lie in the plane or on Hirzebruch surfaces.
In this paper we study the pseudoeffective cones of blow-ups of Grassmannians at sets of points. For small numbers of points, the cones are often spanned by proper transforms of Schubert classes. In some special cases, we provide sharp bounds for when the Schubert classes fail to span and we describe the resulting geometry.