On a t-exactness property of the Harish-Chandra transform

Roman Gonin,Andrei Ionov,Kostiantyn Tolmachov

2024-12-10

Abstract:Using hyperbolic localization, we identify the nearby cycles along the Vinberg degeneration with the composition of Radon and Harish-Chandra functors, both considered for the category of character sheaves. This provides a new, simple proof of the exactness of this composition, extending previously known results to arbitrary monodromy and more general sheaf-theoretic set-ups.

Representation Theory,Algebraic Geometry

What problem does this paper attempt to address?

The problem that this paper attempts to solve is related to the t - exactness property of the Harish - Chandra transform. Specifically, by introducing and using hyperbolic localization, the author equates the nearby cycles along Vinberg degeneration with the composition of the Radon and Harish - Chandra functors, thus providing a new and simple proof for the exactness of this composition and extending the previously known results to be applicable to arbitrary monodromy and a more general sheaf - theoretic framework.

### Main Problem Analysis

1. **t - exactness of the Harish - Chandra transform**
- The Harish - Chandra transform \( \text{hc}! \) was introduced by Lusztig for studying character sheaves. This transform is defined as the pull - push functor along the horocycle correspondence.
- A key question is: when composed with the Radon transform \( R_* \), does \( R_* \circ \text{hc}! \) have t - exactness and Verdier self - duality under certain conditions?

2. **Extension of Results**
- Previous literature has already proven t - exactness in specific cases (such as in the D - module setting).
- The goal of this paper is to generalize these results to arbitrary monodromy and a broader sheaf - theoretic framework.

3. **Innovation in Methodology**
- Use the hyperbolic localization technique to connect the nearby cycle functor with the Harish - Chandra transform.
- In this way, the author is able to simplify the proof process and provide a more general framework for understanding the properties of these transforms.

### Formula Representation

- **Harish - Chandra Transform**
\[
\text{hc}! := \text{Av}_U^! \circ \text{For}_G
\]
where \( \text{Av}_U^! \) represents the! - pushforward functor from \( H' \backslash X \) to \( H \backslash X \), and \( \text{For}_G \) represents the forgetful functor from \( H \backslash X \) to \( H' \backslash X \).

- **Radon Transform**
\[
R_* := \text{Av}_{U^-}^* \circ \text{For}_U[2 \dim U]
\]
where \( U^- \) is the unipotent radical part of the opposite Borel subgroup \( B^- \), and \( \text{Av}_{U^-}^* \) represents the * - pushforward functor from \( (U \backslash G / U) / \text{Ad}T \) to \( (U^- \backslash G / U) / \text{Ad}T \).

- **Main Theorem**
\[
\text{Theorem 1.2.1: } R_* \circ \text{hc}!, \text{ when restricted to the category } D_b^{\mathcal{C}_L}(G), \text{ is t - exact with respect to the perverse t - structure and commutes with the Verdier duality.}
\]

By this method, the author not only verifies the t - exactness of the Harish - Chandra transform after composition.