The Cox ring of a complexity-one horospherical variety (Q512272): Difference between revisions

Horospherical varieties are an important special type of spherical varieties. As the latter, they are defined as (partial) $G$-compactifications $X$ of homogeneous spaces $G/H$ for a reductive group $G$; but the subgroup $H\subseteq G$ is supposed to have a very special feature -- it contains a maximal unipotent subgroup. Equivalently, the normalizer $P=N_G(H)$ is parabolic, making $G/H$ into a torus bundle over the generalized flag variety $G/P$. Horospherical varieties are direct generalizations of toric varieties, and both gadgets give rise to another generalization: The fact that $G$ had an open orbit $G/H\subset X$ can be replaced by asking $G/H$ to be of codimension one in $X$ -- this is what is meant with ``complexity-one''. ``Complexity one toric varieties'' can be described by combinatorial tools like so-called polyhedral divisors and divisorial fans. In the spirit of Luna-Vust theory, this theory extends to complexity-one horospherical varieties -- there one needs a colored version for both combinatorial notions. The paper of Langlois and Terpereau use this description to provide an explicit description (by generators and relations) of the Cox (or ``homogeneous coordinate'') ring of complexity-one horospherical varieties. This is a generalization of a similar project of Hausen (and coauthors) for the toric complexity one case.