Stable orbital integrals for classical Lie algebras and smooth integral models

TL;DR

This paper introduces a new stratification-based approach to describe stable orbital integrals for classical Lie algebras over non-Archimedean fields, providing explicit formulas, bounds, and conjectures on optimality.

Contribution

It offers a novel stratification and smoothing method for stable orbital integrals in classical Lie algebras, with explicit formulas and bounds, extending understanding across various types.

Findings

Closed formulas for certain low-rank cases

Lower bounds for all ranks n

Conjectures on optimality of bounds

Abstract

A main goal of this paper is to introduce a new description of the stable orbital integral for a regular semisimple element and for the unit element of the Hecke algebra in the case of $\mathfrak{gl}_{n,F}$, $\mathfrak{u}_{n,F}$, and $\mathfrak{sp}_{2n,F}$, by assigning a certain stratification and then smoothening each stratum, where $F$ is a non-Archimedean local field of any characteristic. As applications, we will provide a closed formula for the stable orbital integral for $\mathfrak{gl}_{2,F}$, $\mathfrak{gl}_{3,F}$, and $\mathfrak{u}_{2,F}$. We will also provide a lower bound for the stable orbital integral for $\mathfrak{gl}_{n,F}$, $\mathfrak{u}_{n,F}$, and $\mathfrak{sp}_{2n,F}$ with all $n$. Finally we will propose conjectures that our lower bounds are optimal in a sense of the second leading term for $\mathfrak{gl}_{n,F}$ and the first leading term for $\mathfrak{u}_{n,F}$ and $\mathfrak{sp}_{2n,F}$. There is a restriction about the factorization of the characteristic polynomial arising from the parabolic descent when we work with $\mathfrak{u}_{n,F}$ and $\mathfrak{sp}_{2n,F}$, whereas this assumption does not appear in $\mathfrak{gl}_{n,F}$ case.