Geometrically connected components of Lubin-Tate deformation spaces with level structures (Q959031): Difference between revisions

Let \(F\) be a non-archimedean local field, with residue field \(\mathbb{F}_q\) and ring of integers \(\mathfrak{o}\). Let \(\mathbb{X}\) be a one-dimensional formal \(\mathfrak{o}\)-module of \(F\)-height \(n\) over \(\mathbb{F}_q^{\text{ sep}}\). \textit{V.~G.~Drinfeld} [Math. USSR, Sb. 23, 561--592 (1974; Zbl 0321.14014)] introduced the notion of a level-\(m\)-structure on deformations of \(\mathbb{X}\), and proved that the functor of deformations of \(\mathbb{X}\) equipped with level-\(m\)-structure is representable by an affine formal scheme \(\mathcal{M}_m=\text{Spf}\,R_m\), where \(R_m\) is a regular, complete, local, Noetherian \(\hat{\mathfrak{o}}^{\text{nr}}\)-algebra. Let \(M_m=\mathcal{M}_m^{\text{ rig}}\) be the generic fibre of \(\mathcal{M}_m\), as a rigid-analytic space in the sense of \textit{A.~J.~de~Jong} [Publ. Math., Inst. Hautes Étud. Sci. 82, 5--96 (1995; Zbl 0864.14009)]. According to the non-abelian Lubin-Tate theory conjectured by [\textit{H.~Carayol}, Automorphic forms, Shimura varieties, and L-functions. Vol. II, Proc. Conf., Ann Arbor/MI (USA) 1988, Perspect. Math. 11, 15--39 (1990; Zbl 0704.11049)] and V.~G.~Drinfeld, and proved by \textit{P.~Boyer} [Invent. Math. 138, No. 3, 573--629 (1999; Zbl 1161.11408)] and [\textit{M.~Harris} and \textit{R.~Taylor}, The geometry and cohomology of some simple Shimura varieties (Annals of Mathematics Studies 151. Princeton, NJ: Princeton University Press) (2001; Zbl 1036.11027)], the limit \(\varinjlim_m H^{n-1}(M_m\times_{\hat{F}^{\text{nr}}}\mathbb{C}_\pi,\overline{\mathbb{Q}}_l)\) (\(\mathbb{C}_\pi=\) completion of algebraic closure of \({\hat{F}^{\text{nr}}}\)) realises both the local Jacquet-Langlands and Langlands correspondences. The aforementioned proofs of this result depend on global techniques such as Shimura or Drinfeld varieties, and there is now considerable interest in providing purely local proofs. The paper at hand raises a slightly different question, namely determining the Galois action on \(H^r(M_m\times_{\hat{F}^{\text{nr}}}\mathbb{C}_{\pi,}\overline{\mathbb{Q}}_{l})\) for \(r\neq n-1\); the paper treats the most basic (though by no means non-trivial) case, namely \(r=0\), in which case one must identify the Galois action on the set of geometrically connected components of \(M_m\). The paper begins with a review of formal \(\mathfrak{o}\)-modules, their deformations, and the existence of the deformation space \(\mathcal{M}_m\). Secondly, elements of Lubin-Tate theory are reviewed and used to describe the Galois action on the determinant of the Tate module of the universal deformation \(X^{\text{univ}}\) of \(\mathbb{X}\). It is then shown that \(M_m\) has \((q-1)q^{m-1}\) geometrically connected components, that they can be naturally identified with \((\mathfrak{o}/\mathfrak{p}^m)^{\times}\) (\(\mathfrak{p}=\) maximal ideal of \(\mathfrak{o}\)), and that they are defined over the \(m^{\text{th}}\) Lubin-Tate extension of \(\hat{F}^{\text{nr}}\). Finally, the Galois group of \(R_m/R_0\) is known to be naturally isomorphic to \(\text{GL}_n(\mathfrak{o}/\mathfrak{p}^m)\) by other work of the author [Adv. Math. 217, No. 3, 889--951 (2008; Zbl 1140.22017)] and the resulting Galois action on the set of components is explicitly given. The paper is very clearly written.