English

Equations for point configurations to lie on a rational normal curve

Algebraic Geometry 2019-08-06 v4

Abstract

The parameter space of nn ordered points in projective dd-space that lie on a rational normal curve admits a natural compactification by taking the Zariski closure in (Pd)n(\mathbb{P}^d)^n. The resulting variety was used to study the birational geometry of the moduli space M0,n\overline{\mathrm{M}}_{0,n} of nn-tuples of points in P1\mathbb{P}^1. In this paper we turn to a more classical question, first asked independently by both Speyer and Sturmfels: what are the defining equations? For conics, namely d=2d=2, we find scheme-theoretic equations revealing a determinantal structure and use this to prove some geometric properties; moreover, determining which subsets of these equations suffice set-theoretically is equivalent to a well-studied combinatorial problem. For twisted cubics, d=3d=3, we use the Gale transform to produce equations defining the union of two irreducible components, the compactified configuration space we want and the locus of degenerate point configurations, and we explain the challenges involved in eliminating this extra component. For d4d \ge 4 we conjecture a similar situation and prove partial results in this direction.

Keywords

Cite

@article{arxiv.1711.06286,
  title  = {Equations for point configurations to lie on a rational normal curve},
  author = {Alessio Caminata and Noah Giansiracusa and Han-Bom Moon and Luca Schaffler},
  journal= {arXiv preprint arXiv:1711.06286},
  year   = {2019}
}

Comments

28 pages. Minor correction. We removed the erroneous Lemma 4.7 in the previous version, but the remaining results are valid