🔢 mathematics

Preservation of stability under the Fourier-Mukai transform whose kernel is the Poincare line bundle

This paper investigates the preservation of Gieseker stability for sheaves on arbitrary abelian surfaces under the Fourier-Mukai transform defined by the Poincaré line bundle, and applies these findings to provide insights into the weak Brill-Noether property.

Original authors: Kota Yoshioka

Published 2026-06-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kota Yoshioka

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a complex, multi-layered cake (representing a mathematical object called a "sheaf" on a special type of surface called an "abelian surface"). Mathematicians want to know if this cake is "stable"—meaning it holds its shape and doesn't crumble or mix unpredictably when you look at it from different angles.

This paper is about a specific magic trick called the Fourier-Mukai transform. Think of this transform as a magical mirror or a translation device. When you put your cake in front of it, the mirror doesn't just reflect the cake; it completely reassembles it into a new shape on a different table (the "dual surface").

The big problem the author, Kota Yoshio, is solving is this: Does the new cake in the mirror still hold its shape? In mathematical terms, does "stability" survive the trip through the mirror?

Here is a breakdown of the paper's journey, using simple analogies:

1. The Setup: The Cake and the Mirror

The Cake (The Sheaf): Mathematicians study these objects to understand the geometry of surfaces. They classify them based on how "stable" they are. A stable cake is one that is perfectly balanced.
The Mirror (The Transform): The Fourier-Mukai transform is a powerful tool that swaps information between two surfaces. It's like translating a book from English to French. Usually, when you translate a book, the sentence structure changes, and sometimes the meaning gets a bit fuzzy.
The Goal: The author wants to know: If I start with a perfectly stable cake, will the translated version in the mirror also be a perfectly stable cake?

2. The Problem: When the Mirror Gets Foggy

In the past, mathematicians knew this "translation" worked perfectly if the surface was very simple (like a plain, single-colored wall). But if the surface is more complex (has more "colors" or directions, known as a higher Picard number), the translation often fails. The stable cake might turn into a messy pile of crumbs in the mirror.

The author asks: Can we fix the mirror or choose the right angle to look at it so the cake stays stable?

3. The Solution: Finding the Right Angle

The paper proves that yes, you can keep the cake stable, but you have to be careful about how you look at it.

The "Polarization" (The Angle): Imagine the surface has a "gravity" or a "wind" blowing in a specific direction (mathematically called an "ample divisor"). The stability of the cake depends on which way the wind is blowing.
The Discovery: The author shows that for almost any complex cake, there is a specific wind direction (a specific polarization) where, if you run the cake through the mirror, it comes out as a stable, intact object on the other side.
- If the cake has a certain "positive weight" (mathematically $a > 0$ ), the mirror shows you the cake flipped upside down (dual), but it's still a solid cake.
- If the cake has a "negative weight" ( $a \le 0$ ), the mirror shows you the cake shifted by one step (like moving it from the table to a shelf), but it's still a solid cake.

4. The "Weak Brill-Noether" Property: The One-Group Rule

The paper also touches on a concept called the "Weak Brill-Noether property."

The Analogy: Imagine your cake has layers of filling. Sometimes, when you look at it through the mirror, you see filling in two different layers at once (cohomology groups). This is messy.
The Result: The author proves that if you choose the right wind direction (polarization), the mirror will show you a cake where the filling is concentrated in only one layer. It's a "clean" translation. This is a very desirable property for mathematicians because it makes the object much easier to study.

5. The Exceptions: When the Cake is Special

The author is honest about the limits. There are a few specific, weird types of cakes (like those made on a surface that is actually just two circles glued together, or cakes with very specific, rigid structures) where the mirror cannot be tuned to keep them stable. In these rare cases, the cake will always turn into a messy pile in the mirror. The paper identifies exactly which cakes these are so mathematicians know when to stop trying to force the trick.

Summary

In everyday terms, this paper is a user manual for a magical mirror.

Old belief: "This mirror breaks fragile objects if the room is too complex."
New discovery: "Actually, if you turn the lights on just right (choose the right polarization), the mirror preserves the object's shape perfectly, even in complex rooms. The only time it fails is if the object itself is built in a very specific, rigid way."

This helps mathematicians move objects between different geometric worlds without losing their essential properties, which is a huge step forward in understanding the shape of the universe in algebraic geometry.

Technical Summary: Preservation of Stability under the Fourier-Mukai Transform

Problem Statement
The paper addresses the behavior of Gieseker stability for coherent sheaves on an abelian surface $X$ under the Fourier-Mukai transform $\Phi_{P^\vee}: D(X) \to D(\hat{X})$ , where the kernel is the dual of the Poincaré line bundle $P$ . While the Fourier-Mukai transform induces an equivalence of derived categories, it does not generally preserve the stability of sheaves. Previous work (specifically [16]) established stability preservation for abelian surfaces with Picard number $\rho(X)=1$ . This paper extends the investigation to the general case where the Picard number $\rho(X) \geq 1$ , aiming to identify conditions under which the transform of a general stable sheaf remains stable (possibly after a shift or dualization) with respect to a suitable ample divisor on the dual surface $\hat{X}$ .

A secondary focus is the weak Brill-Noether property, which asserts the existence of a sheaf in a moduli space with at most one non-zero cohomology group. The author notes that this property fails in general when the moduli stack is reducible and the polarization is near the boundary of the ample cone. The paper seeks to establish conditions where this property holds for a fixed ample divisor.

Methodology
The primary tool employed is the theory of Bridgeland stability conditions on the derived category $D(X)$ . The author analyzes the "wall-crossing" behavior of stability conditions within the space $\text{Stab}(X)$ .

Geometric Stability Conditions: The study focuses on the line $L = \{(0, tH) \mid t > 0\}$ in the space of stability parameters, corresponding to geometric stability conditions $\sigma_{(0, tH)}$ .
Totally Semi-Stable Walls: The author investigates "totally semi-stable walls," defined as walls where the moduli spaces of stable objects on either side are disjoint. These walls correspond to decompositions of the Mukai vector $v$ into isotropic vectors.
Wall-Crossing Analysis: By tracking the behavior of a general stable object $E$ as the parameter $t$ varies along the line $L$ , the paper determines when $E$ transitions from being a torsion-free sheaf to a complex with cohomology in degree $-1$ (or vice versa).
Fourier-Mukai Compatibility: The paper utilizes the known isometry of the Mukai lattice induced by $\Phi$ and specific lemmas (e.g., Lemma 1.2) regarding how $\Phi$ acts on semi-stable sheaves with isotropic Mukai vectors. The core argument involves showing that for a general stable sheaf $E$ , its transform $\Phi(E)$ (or $\Phi(E)[1]$ ) remains a coherent sheaf and is stable with respect to a specific polarization $L'$ on $\hat{X}$ .

Key Contributions and Results

1. Main Theorem on Stability Preservation (Theorem 0.1 / Theorem 3.12)
The paper establishes conditions under which the Fourier-Mukai transform preserves stability for a general sheaf $E$ in the moduli stack $M_L(v)$ , where $v=(r, \xi, a)$ is a Mukai vector with $r>0$ and $(\xi \cdot H) > 0$ . Let $\ell = \langle v^2 \rangle / 2$ .

Case $a > 0$ : If $X$ is not a product of elliptic curves or $\xi$ is not primitive, there exist ample divisors $L$ and $L'$ such that $\Phi(E)^\vee$ is $L'$ -stable for a general $E \in M_L(v)$ .
Case $a \leq 0$ : Excluding specific exceptional cases involving elliptic curves (e.g., $v = (\ell, kC, -1)$ ), there exist ample divisors $L$ and $L'$ such that $\Phi(E)[1]$ is $L'$ -stable for a general $E \in M_L(v)$ .

2. Weak Brill-Noether Property (Proposition 0.2 / Proposition 3.16)
The author proves that for any Mukai vector $v=(r, \xi, a)$ with $r>0$ , $(\xi \cdot H) > 0$ , and $\langle v^2 \rangle \geq 0$ , there exists an ample divisor $L$ such that the weak Brill-Noether property holds for $M_L(v)$ . Specifically, there exists a sheaf $E \in M_L(v)$ having at most one non-zero cohomology group. This result is derived directly from the stability preservation results, which ensure that the transform lands in the heart of the derived category (i.e., is a sheaf) for a general object.

3. Irreducibility and Polarization Independence (Corollary 0.3)
Under the assumption that the moduli stack $M(v)$ is irreducible (which holds if $v$ is primitive and certain numerical conditions are met, e.g., $\langle v^2 \rangle \geq 2r, 2a$ ), the paper shows that the stability preservation holds with respect to the naturally associated polarization $\hat{H}$ on the dual surface, without needing to adjust the polarization on $X$ .

4. Specific Case: Ample Divisors of the Form $dH$ (Theorem 0.4 / Theorem 3.24)
When the Mukai vector is of the form $v=(r, dH, a)$:

If $d \geq 0$ , the weak Brill-Noether property holds.
If $d < 0$ , the property holds unless $v$ takes a specific exceptional form related to negative integers and elliptic curves.
This section also provides a characterization of when an $aCM$ (arithmetically Cohen-Macaulay) sheaf exists for a given Mukai vector (Corollary 3.25).

Significance and Claims
The paper claims to generalize previous results on the preservation of stability from the case of Picard number 1 to arbitrary abelian surfaces. By utilizing Bridgeland stability and analyzing wall-crossing along a specific line in the stability manifold, the author provides a constructive method to find polarizations that preserve stability under the Fourier-Mukai transform.

The author explicitly states that these results resolve the weak Brill-Noether property for fixed ample divisors, contrasting with earlier findings (e.g., in [6]) where the property fails for certain reducible moduli spaces or polarizations near the boundary of the ample cone. The work refines the understanding of the geometry of moduli spaces of sheaves on abelian surfaces and their behavior under derived equivalences.

Limitations and Exceptions
The paper carefully notes exceptional cases where the general theorems do not apply, primarily involving:

Products of elliptic curves.
Primitive vectors $\xi$ in specific configurations.
Mukai vectors of the form $(\ell, kC, -1)$ or $(1, kC, -\ell)$ where $C$ is an elliptic curve and $k \geq \ell + 1$ .
In these exceptional cases, the transform often results in a coherent sheaf with torsion rather than a stable vector bundle, or the moduli space structure prevents the standard stability preservation.