Relative Bruhat decomposition of wonderful compactification
Fei Chen, Shang Li
TL;DR
The paper develops a comprehensive relative Bruhat theory for the wonderful compactification of a reductive group over a general base field. It constructs a precise orbit stratification of \\overline{G}(k) into P(k)×P^−(k)-orbits, refined from G(k)×G(k)-orbits, and provides a geometric refinement and an explicit relative Bruhat order describing when one orbit lies in the closure of another. The results extend Brion–Springer-type theorems from algebraically closed fields to arbitrary base fields via Galois descent, while also addressing topological closure properties under non-discrete topologies. Together, these findings illuminate the asymptotic (infinity) behavior of Bruhat decompositions in the setting of wonderful compactifications and unify geometric and combinatorial descriptions across split and non-split groups. The framework has potential implications for representation theory, invariant theory, and the study of compactifications of reductive groups over general fields.
Abstract
In the seminal paper of Borel and Tits about reductive groups, they show some fundamental results about Bruhat cells with respect to a minimal parabolic subgroup, e.g., relative Bruhat decomposition and its geometrization, relative Bruhat order and the relation of Zariski closure and topological closure. In this paper, we show analogous results for Bruhat cells of wonderful group compactification in the sense of De Concini and Procesi. Our results can be viewed as the version at infinity of those of Borel and Tits. Our main focus is general base field. When the base field is algebraically closed, most of our results are proved by Brion and Springer.
