Dominant regions in noncrystallographic hyperplane arrangements (Q2371290)

Let \(\Phi \) be a root system in an \(n\)-dimensional Euclidean space \(V\), \(\Delta =\{ \alpha _1, \alpha _2,\dots,\alpha _n\}\) a set of simple roots and \(\Phi ^+=\{ \beta \in \Phi \mid \beta =\sum_{i=1}^nc_i\alpha _i\) with \(\;c_i\in \mathbb{R}_{\geq 0}\}\) the resulting positive root. A region in \(V\) is a connected component of the set \(V\backslash \bigcup_{(\beta ,c)\in \mathcal{F}}\{ v\in V\mid (v,\beta )=c\}\), where \(\mathcal{F}=\Phi ^+\times \{-1,0,1\}\). A dominant region is a region contained in the dominant chamber \(C=\{ v\in V\mid (v,\alpha _i)>0\) for \(1\leq i\leq n\}\). Define on \(\Phi ^+\) the partial order \(\beta \leq \gamma \) if \(\gamma -\beta =\sum_{i=1}^nc_i\alpha _i\) with \(c_i\in \mathbb{R}_{\geq 0}\). Each increasing set in \((\Phi ^+,\leq )\) is generated by its minimal elements, which form an antichain. In the case of a crystallographic root system, dominant regions are in bijection with antichains. The authors show that such a bijection does not hold in the case of certain noncrystallographic root systems. Particularly, they present a list of 16 antichains in the \(H_4\) root order which correspond to empty regions.