On two definitions of wave-front sets for $p$-adic groups

Abstract

The wave-front set for an irreducible admissible representation of a $p$-adic reductive group is the set of maximal nilpotent orbits which appear in the local character expansion. By Mœglin-Waldspurger, they are also the maximal nilpotent orbits whose associated degenerate Whittaker models are non-zero. However, in the literature there are two versions commonly used, one defining maximality using analytic closure and the other using Zariski closure. We show that these two definitions are non-equivalent for $G=S{p}_4$.

9 pages. Comments welcome! (v2: journal version. Mild fix and update around Proof of Lemma 2.1, Prop. 3.1 and other stuff)