Chabauty limits of parahoric subgroups of \(\mathrm{SL}(n,\mathbb{Q}_p)\) (Q2238905)

For any locally compact group \(G\), the set \(S(G)\) of closed subgroups is compact under a natural topology called the Chabauty topology. Consequently, when \(X\) is a topological space on which a locally compact group \(G\) acts continuously, the stabilizers \(G_x\) are closed subgroups, and so, the closure of \(\{G_x\}_{x \in X}\) under the Chabauty topology gives a natural compactification of \(X\). Examples of such \(X\) are symmetric spaces and Bruhat-Tits buildings (where the stabilizers are certain compact subgroups). Guivarch and Remy had proved in 2006 some general structure theorems and compactification theorems. In particular, when \(\mathcal{G}\) is a semisimple algebraic group over a nonarchimedean local field and \(G= \mathcal{G}(F)\), they obtained compactifications of the Bruhat-Tits building \(X\) of \(G\). One of the main results they proved is that after taking stabilizers, the so-called fundamental sequences of points in the building \(X\) lead to convergent sequences (with respect to the Chabauty topology) of parahoric subgroups. In the present paper, the authors give beautiful geometric arguments for the above result for the case of \(G=\mathrm{SL}(n, \mathbb{Q}_p)\). The Guivarch-Remy theorem is more general, but depends on more complicated, measure-theoretic methods and the Furstenberg boundary. The precise statement of the theorem proved in the present paper is: Let \(G = \mathrm{SL}(n,\mathbb{Q}_p)\), and let \(X\) be its associated Bruhat-Tits building. Let \(x\) be a fixed vertex of \(X\). For a sequence \(\{x_n\}\) of points in \(X\), the sequence \(\{G_{x_n} \}\) of stabilizers admits a convergent subsequence with respect to the Chabauty topology on \(S(G)\) such that the corresponding limit is \(G_x\)-conjugate to a specific subgroup of one of the parabolic subgroups of \(\mathrm{SL}(n, \mathbb{Q}_p)\), which preserves a facet in the spherical building \(\partial X\) at infinity, and has a Levi decomposition that can be given explicitly. As the authors point out, their proof will not carry over to general semisimple groups because the Cartan (KAK) decomposition and Iwasawa decomposition that they use is valid only for special vertices of \(X\) (and all vertices are special when \(G= \mathrm{SL}(n, \mathbb{Q}_p)\)). To give an example of the geometric methods employed by the authors, we recall a key lemma proved and used by them which asserts: Let \(\Delta\) be a locally finite, affine building and let \(G\) be a closed, non-compact subgroup of \(\mathrm{Aut}(\Delta)\). Suppose, a sequence \(\{g_n\}\) in \(G\) converges to an isometry \(g\) in \(G\). If \(g\) is elliptic (resp., hyperbolic), then there exists \(N >0\) such that for all \(n \geq N\), \(g_n\) is elliptic (resp., hyperbolic). In particular, the translation length of \(g_n\) is that of \(g\) for every \(n \geq N\).