Torus quotients of homogeneous spaces --- minimal dimensional Schubert varieties admitting semi-stable points (Q1043863)

Let \(G\) be a simply-connected, semi-simple algebraic group over an algebraic closed field, let \(T\) be a maximal torus of \(G\) and let \(B\) be a Borel subgroup of \(G\) containing \(T\). The authors are interested to determine what Schubert varieties of a projective homogeneous variety \(G/P\) contain a semistable point with respect to the action of \(T\) and to a fixed ample line bundle \(\mathcal{L}\). In this work there are two main results. In the first part of this article the authors restrict themselves to the case where \(G\) is a simple group of type \(B\), \(C\) or \(D\) and \(P\) is a maximal parabolic subgroup. With these hypotheses they classify the minimal elements \(w\in W/W_{P}\) such that \(X(w)_{T}^{ss}(\mathcal{L})\neq\emptyset\). Here \(X(w)=\overline{BwP/P}\) is the Schubert variety associated to \(w\). The authors affirm that, when \(G\) exceptional or \(P\) non-maximal, the same problem is more complicated. In the second part of this work, the authors classify the Coxeter elements \(\tau\) of \(W\) such that there is a non-trivial line bundle \(\mathcal{L}\) on \(G/B\) with \(X(\tau)_{T}^{ss}(\mathcal{L})\neq\emptyset\). An element of \(W\) is a Coxeter element if it can be written as a product of distinct simple reflections. The authors are interested to such elements for the following reason: a Schubert variety \(X(w)\) contains a (rank \(G\))-dimensional \(T\)-orbit if and only if \(w\geq \tau\) for some Coxeter element \(\tau\). In this part, they do not make special assumption on \(G\). Let \(\chi\) be the \(B\)-character associated to a fixed globally generated line bundle \(\mathcal{L}\). Supposing that \(\chi\) belongs to the root lattice, the authors prove that \(X(w)_{T}^{ss}(\mathcal{L})\neq\emptyset\) if and only if \(w\chi\leq0\). This fact allows them to prove their main theorems in a combinatorial way. Remark that in the Proposition 3 (and only in that Proposition) the authors use a definition of \(\omega\geq0\) different from the usual one. They say that a weight \(\omega\) is greater or equal to 0 if it can written as a positive linear combination of simple roots. In particular, they do not assume that the coefficients are integral.