Automorphism groups of toroidal horospherical varieties

This paper establishes a structure theorem for the connected automorphism groups of smooth complete toroidal horospherical varieties by characterizing extendable Demazure roots, thereby providing a criterion for their reductivity and proving the K-unstability of certain $\mathbb{P}^1$-bundles over rational homogeneous spaces.

The Gist

Imagine you are an architect designing a magnificent, multi-story building. In the world of mathematics, this building is called a variety, and the "people" who can walk around inside it, rotate it, or stretch it without breaking it are called automorphisms.

The group of all these possible movements is the Automorphism Group.

This paper is about understanding the "personality" of this group for a very specific type of building: a Toroidal Horospherical Variety. That's a mouthful, so let's break it down with some analogies.

1. The Building: A "Toric Bundle" over a "Homogeneous Space"

Think of your building as a stack of identical apartments (the fibers) built on top of a perfectly symmetrical city square (the base).

• The City Square (Base): This is a "Rational Homogeneous Space." It's like a perfect sphere or a cube where every point looks exactly like every other point. The symmetries here are very rigid and "pure" (mathematically, they are semisimple).
• The Apartments (Fibers): These are "Toric Varieties." Think of them as a standard apartment complex where the rooms are arranged in a grid. These have a lot of flexibility; you can slide things around in specific directions. Their symmetries are a mix of rigid rotations and "sliding" movements (mathematically, they have unipotent parts).

The Whole Building: You take the city square and build a tower of these apartments on every single point. The result is a Toroidal Horospherical Variety. It combines the rigid perfection of the square with the sliding flexibility of the apartments.

2. The Problem: Is the Building "Reductive"?

Mathematicians love to categorize groups of movements as either Reductive or Non-Reductive.

• Reductive (The "Stable" Group): Imagine a group of dancers who only do perfect, balanced spins and jumps. They don't slide or drift. If a building's symmetry group is reductive, it's considered "stable" and well-behaved. In the real world, this is crucial for a concept called K-stability, which determines if a shape can support a perfect, balanced physical structure (like a constant curvature metric).
• Non-Reductive (The "Sliding" Group): Imagine a group of dancers who, in addition to spinning, also slide uncontrollably in one direction. This "sliding" is called a unipotent action. If your building has too much "sliding" potential, the group is non-reductive, and the building is considered "unstable" in the K-stability sense.

The Big Question: For our mixed building (City Square + Apartments), when is the group of movements purely "spinning" (Reductive), and when does it start "sliding" (Non-Reductive)?

3. The Discovery: The "Demazure Roots" as Keys

The authors found a way to answer this by looking at keys hidden in the blueprints.

In the world of the "Apartments" (Toric varieties), there are special keys called Demazure Roots.

• Some keys unlock Spin Doors (Semisimple roots).
• Some keys unlock Sliding Doors (Unipotent roots).

The authors realized that for the whole building, a key from the apartment blueprint only works on the whole structure if it fits a specific lock on the City Square. They called these valid keys $B^+$-roots.

The Golden Rule (The Main Theorem):
The group of movements for the whole building is Reductive (stable) IF AND ONLY IF there are NO "Sliding Door" keys (Unipotent roots) that fit the locks of the whole building.

If even one sliding door key works, the whole group becomes "non-reductive," and the building is unstable.

4. The Application: The "Unstable" P1-Bundles

The authors didn't just stop at theory; they used their rule to build new examples of "unstable" buildings.

They looked at a specific type of building: a P1-bundle. Imagine a city square where every point has a tiny line segment (like a pencil) attached to it, pointing up or down.

• They asked: "If I attach these pencils in a certain way, will the whole structure be K-unstable?"
• Using their "No Sliding Doors" rule, they proved that if you attach the pencils using a specific type of "non-trivial" line (a nef line bundle), the structure will have sliding doors.
• Conclusion: These specific pencil-towers are K-unstable. They cannot support a perfect, balanced physical metric.

Why Does This Matter?

In the real world (and in theoretical physics), K-stability is the difference between a shape that can exist as a perfect, balanced universe and one that collapses or distorts.

• Before this paper: We knew how to check stability for pure city squares (Homogeneous spaces) and pure apartment grids (Toric varieties).
• After this paper: We have a clear, easy-to-use checklist for the complex hybrid buildings in between. We can now look at the "blueprint" (the line bundles), check if any "sliding keys" exist, and instantly know if the structure is stable or not.

Summary in One Sentence

The authors figured out that for a specific type of complex geometric building, you can tell if it's structurally "stable" by checking if any of its internal "sliding mechanisms" (unipotent roots) are allowed to operate; if they are, the building is unstable, and they provided a simple rule to check this for many new types of buildings.

Technical Summary

Here is a detailed technical summary of the paper "Automorphism groups of toroidal horospherical varieties" by Lorenzo Barban, DongSeon Hwang, and Minseong Kwon.

1. Problem Statement

The paper addresses the structural characterization of the connected automorphism group, denoted $\text{Aut}^0(X)$, for a specific class of algebraic varieties: smooth complete toroidal horospherical varieties.

• Context: The automorphism group of an algebraic variety encodes significant geometric information. A central question in the field is determining when $\text{Aut}^0(X)$ is reductive. This is motivated by the Matsushima–Lichnerowicz theorem, which links the reductivity of $\text{Aut}^0(X)$ to the existence of constant scalar curvature Kähler metrics (and Kähler–Einstein metrics for Fano varieties).
• Gap in Knowledge:
  • For rational homogeneous spaces, $\text{Aut}^0(X)$ is semisimple (hence reductive).
  • For toric varieties, $\text{Aut}^0(X)$ is generally not reductive; its structure is well-understood via Demazure roots (specifically, it is reductive iff there are no unipotent Demazure roots).
  • Horospherical varieties sit between these two classes (generalizing both). While partial results exist for specific cases (e.g., low Picard number), a complete characterization of the reductivity and the Levi decomposition of $\text{Aut}^0(X)$ for general smooth complete toroidal horospherical varieties was missing.

2. Methodology

The authors employ a global approach combining the geometry of the base space, the fiber, and the combinatorial data of the variety.

• Geometric Setup: A toroidal horospherical variety $X$ is viewed as a toric bundle $\Phi: X \to G/P$, where:
  • $G/P$ is a rational homogeneous space (the base).
  • The fiber $F$ is a complete toric variety with respect to an algebraic torus $S$.
  • $X$ is constructed via parabolic induction: $X \cong G \times_P F$.
• Combinatorial Tools:
  • Demazure Roots: The authors utilize the theory of Demazure roots for the toric fiber $F$.
  • Color Map: They introduce a color map $\epsilon_+: \mathcal{D}_{B^+} \to N_{G/H}$, which relates the $B^+$-invariant divisors (colors) of the open orbit to the lattice of the torus $S$.
  • Generalized Roots: They define $B^+$-roots for $X$ by filtering the Demazure roots of the fiber $F$ based on their compatibility with the color map. Specifically, a root $m$ of the fiber is a $B^+$-root of $X$ if $\langle m, \epsilon_+(D) \rangle \geq 0$ for all colors $D$.
• Extension of Actions: The core technical mechanism involves determining when a $\mathbb{G}_a$-action (associated with a Demazure root) on the fiber $F$ extends to a $B^+$-normalized action on the total space $X$. This is analyzed using $L$-normalized actions (where $L$ is a Levi subgroup).

3. Key Contributions and Results

A. Structure Theorem for $\text{Aut}^0(X)$

The paper establishes a precise decomposition of the Lie algebra of the automorphism group.

• Dimension Formula: The dimension of $\text{Aut}^0(X)$ is given by:
  $$\dim \text{Aut}^0(X) = \dim \text{Aut}^0(G/P) + \dim F + |S^+_G(X)| + \sum_{m \in U^+_G(X)} \dim V(m)$$
  Where:
  • $S^+_G(X)$: Set of semisimple $B^+$-roots.
  • $U^+_G(X)$: Set of unipotent $B^+$-roots.
  • $V(m)$: The irreducible $G$-module of highest weight $m$.
• Levi Decomposition:
  • The unipotent radical $R_u(\text{Aut}^0(X))$ is generated by the $\mathbb{G}_a$-subgroups associated with $m \in U^+_G(X)$ and their $G$-conjugates. As a $G$-module, its Lie algebra is the direct sum of irreducible modules $V(m)$ for $m \in U^+_G(X)$.
  • A Levi subgroup is generated by $G$, the group of $G$-equivariant automorphisms $\text{Aut}_G(X)$, and the $\mathbb{G}_a$-subgroups associated with $m \in S^+_G(X)$.

B. Criterion for Reductivity

The paper provides a necessary and sufficient condition for the reductivity of $\text{Aut}^0(X)$:

• Theorem: $\text{Aut}^0(X)$ is reductive if and only if $U^+_G(X) = \emptyset$.
• In other words, the automorphism group is reductive if and only if every $B^+$-root of $X$ is semisimple (i.e., no unipotent roots exist).

C. Application to Projective Bundles

The authors apply their general theory to totally decomposable projective bundles over rational homogeneous spaces.

• Let $Y = G/P$ be a rational homogeneous space and $X = \mathbb{P}_Y(\bigoplus_{i=1}^k L_i)$ be a projective bundle of line bundles.
• Reductivity Criterion: $\text{Aut}^0(X)$ is reductive if and only if for any distinct $i, j$ such that $L_i \not\simeq L_j$, the line bundle $L_i \otimes L_j^\vee$ is not nef.
• This provides an effective, concrete check for reductivity based on the positivity properties of the defining line bundles.

D. K-Unstability of Fano Varieties

Using the Matsushima obstruction (non-reductive automorphism groups imply K-instability for Fano varieties), the paper constructs new examples of K-unstable Fano varieties.

• Corollary: If $Y$ is a rational homogeneous space and $L$ is a non-trivial nef line bundle such that $K_Y^\vee \otimes L^\vee$ is ample, then $X = \mathbb{P}_Y(\mathcal{O}_Y \oplus L^\vee)$ is a smooth K-unstable Fano variety.
• This result is significant because it produces K-unstable examples even in cases where previous methods (requiring high Fano index) failed, specifically covering cases with Fano index 1.

4. Significance

• Unification: The work unifies the understanding of automorphism groups for toric varieties and rational homogeneous spaces under the broader framework of horospherical varieties.
• Explicit Characterization: It moves beyond abstract existence theorems to provide explicit combinatorial criteria (via Demazure roots and color maps) for calculating dimensions and determining reductivity.
• K-Stability Implications: By linking the non-reductivity of automorphism groups to K-unstability, the paper offers a powerful tool for generating counterexamples in the study of Kähler–Einstein metrics and K-stability, a hot topic in complex geometry and the Yau-Tian-Donaldson conjecture.
• Generality: The results apply to a broad class of varieties (smooth complete toroidal horospherical varieties) and offer a direct approach that complements more general (but less explicit) results on arbitrary spherical varieties.

In summary, the paper successfully characterizes the automorphism groups of toroidal horospherical varieties by identifying the specific Demazure roots that extend to the total space, thereby providing a definitive criterion for reductivity and enabling the construction of new K-unstable Fano varieties.