🔢 mathematics

Relative Gieseker's problem on $F$ -divided bundles

This paper proves that for a proper surjective morphism of varieties with geometrically connected fibers in positive characteristic, the induced homomorphism of $F$ -divided fundamental groups is faithfully flat, utilizing a new descent theorem for $F$ -divided bundles to generalize and strengthen recent results by Sun and Zhang as well as earlier work by Esnault, Mehta, and others.

Original authors: Adrian Langer

Published 2026-01-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Adrian Langer

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 are trying to understand the shape of a complex, multi-layered object (like a giant, intricate sculpture) by looking at it through a series of different lenses. In the world of mathematics, specifically in a field called algebraic geometry, these "sculptures" are shapes called varieties, and the "lenses" are mathematical tools used to measure their hidden structures.

This paper, written by Adrian Langer, is about solving a puzzle regarding how these shapes behave when you map one onto another, specifically in a mathematical world where numbers work differently than in our everyday life (a world called "positive characteristic").

Here is a breakdown of the paper's main ideas using simple analogies:

1. The Two Types of "Maps"

The author is comparing two different ways of looking at these shapes:

The Etale Lens (The "Rough Sketch"): This is like looking at a sculpture from far away. You can see the big bumps and holes, but you miss the fine details. Mathematicians call this the étale fundamental group.
The F-Divided Lens (The "High-Definition Scan"): This is a much more powerful, high-resolution tool. It looks at the shape not just as it is, but as it changes when you apply a specific mathematical "zoom" or "twist" (called the Frobenius map) over and over again. This is the F-divided fundamental group.

The Problem: If you have a map (a function) that takes Shape A to Shape B, and the "Rough Sketch" of Shape A covers all of Shape B perfectly, does the "High-Definition Scan" of Shape A also cover Shape B perfectly?

2. The Main Discovery: The "Faithful" Connection

The paper proves that yes, it does.

The Analogy: Imagine Shape A is a large, connected forest, and Shape B is a smaller clearing. You have a path (the map $f$ ) leading from the forest to the clearing.

If the path is surjective (it reaches every part of the clearing) and the forest is connected (you can walk from any tree to any other tree without leaving the forest), then the "High-Definition Scan" of the forest is perfectly aligned with the "High-Definition Scan" of the clearing.
The author calls this relationship "faithfully flat." In plain English, it means the detailed structure of the big forest is faithfully preserved and transferred to the clearing. Nothing gets lost in the translation when you zoom in.

3. The Secret Weapon: The "Descent" Theorem

To prove this, the author had to solve a tricky problem: How do you take a complex object that exists in the forest (Shape A) and prove it actually comes from the clearing (Shape B)?

Usually, if you have a pattern in the forest that looks the same on every single tree, you might guess it came from the clearing. But in this specific mathematical world, patterns can be tricky.

The author introduces a new tool, an analogue of a theorem by Bhatt and Scholze.

The Analogy: Imagine you have a magical blanket (an F-divided bundle) draped over the forest. If you look at the blanket on every single tree in the forest, and it looks like a plain, boring sheet of cloth (trivial) on each one, then the author proves that the entire blanket must have been woven in the clearing to begin with.
This is a "Descent" theorem: It allows you to "descend" or move a complex object from the big space down to the smaller space if it looks simple everywhere locally.

4. Why This Matters (In Math Terms)

Before this paper, mathematicians knew this result worked if the shapes were perfectly smooth (like a polished marble statue). But real-world mathematical shapes are often "normal" but not perfectly smooth (they might have sharp corners or singularities).

The Breakthrough: The author proves this works even if the shapes are a bit rough (normal varieties), as long as the path between them is connected.
The "Isomorphism" Result: The paper also proves that if the "Rough Sketch" of the map is a perfect one-to-one match (an isomorphism), then the "High-Definition Scan" is also a perfect one-to-one match.

5. The "Gieseker" Connection

The title mentions "Gieseker's problem." This refers to a question posed by a mathematician named Gieseker decades ago.

The Context: Gieseker asked: "If we look at these shapes through the high-definition lens, do they behave nicely?"
The Answer: This paper says, "Yes, they do, and here is exactly how they behave when you map one shape to another." It strengthens previous answers by other mathematicians (Sun, Zhang, Esnault, Mehta, etc.) by removing some of the strict "smoothness" requirements that were previously thought necessary.

Summary

Think of this paper as a master key. It shows that if you have a connected path between two mathematical shapes, and that path covers the destination completely, then the most detailed, high-tech way of measuring those shapes (the F-divided fundamental group) will also match up perfectly. The author achieved this by inventing a new way to prove that if something looks simple everywhere on a large shape, it must have originated from the smaller shape it was mapped from.

Technical Summary: Relative Gieseker's Problem on F-divided Bundles

Problem Statement
The paper addresses the behavior of fundamental groups under morphisms of varieties in positive characteristic, specifically focusing on the relationship between the étale fundamental group ( $\pi_1^{\text{ét}}$ ) and the $F$ -divided fundamental group ( $\pi_1^{\text{F-div}}$ ). In characteristic zero, Grothendieck established that for a morphism $f: X \to Y$ of proper connected schemes, the surjectivity (or isomorphism) of the map on étale fundamental groups implies the faithful flatness (or isomorphism) of the map on stratified fundamental groups.

In positive characteristic, stratified sheaves are equivalent to $F$ -divided bundles (sequences of vector bundles related by Frobenius pullbacks). The central problem, known as Gieseker's problem, asks whether the analogous statements hold for $\pi_1^{\text{F-div}}$ . Specifically, if $f: X \to Y$ is a surjective morphism of normal projective varieties over an algebraically closed field of positive characteristic, does the surjectivity of $f_*: \pi_1^{\text{ét}}(X) \to \pi_1^{\text{ét}}(Y)$ imply that $f_*: \pi_1^{\text{F-div}}(X) \to \pi_1^{\text{F-div}}(Y)$ is faithfully flat? Furthermore, does the isomorphism of étale fundamental groups imply an isomorphism of $F$ -divided fundamental groups?

Previous results by Esnault and Mehta [EM10] and Biswas, Kumar, and Parameswaran [BKP25] addressed absolute cases or specific finite morphisms, often requiring smoothness assumptions. Sun and Zhang [SZ25] recently generalized these to normal varieties but relied on complex proofs involving the smoothness of the target $Y$ . This paper aims to generalize and strengthen these results, removing unnecessary smoothness assumptions where possible and providing a unified framework.

Methodology
The proof strategy relies on Tannakian duality. Since the category of $F$ -divided bundles, $\text{Vect}^{\text{perf}}(X)$ , is a Tannakian category, the properties of the induced homomorphism on fundamental groups are equivalent to properties of the pullback functor $f^*: \text{Vect}^{\text{perf}}(Y) \to \text{Vect}^{\text{perf}}(X)$ . Specifically:

Faithful Flatness: Requires $f^*$ to be fully faithful and its essential image to be closed under subobjects.
Isomorphism: Requires $f^*$ to be fully faithful and essentially surjective.

The paper employs the following key technical tools:

Descent Theorem for $F$ -divided Bundles: The authors prove an analogue of the descent theorem by Bhatt and Scholze [BS17] specifically for $F$ -divided bundles. This is the core technical innovation. It establishes that an $F$ -divided bundle on $X$ descends to $Y$ if and only if it is trivial on all geometric fibers of $f$ , provided $f$ has geometrically connected fibers. Unlike the Bhatt-Scholze result, which allows descent after a finite purely inseparable map, this result provides a single descent map for the entire sequence of bundles.
Stein Factorization: The morphism $f$ is decomposed into a part with geometrically connected fibers and a finite universal homeomorphism. The latter induces an equivalence of categories (Lemma 1.3), reducing the problem to the case where $f$ has geometrically connected fibers.
Genuinely Ramified Morphisms: For the isomorphism case, the authors utilize the theory of "genuinely ramified" morphisms (introduced in [PS10] and [BKP25]). They adapt arguments regarding slope stability and the behavior of reflexive sheaves under pullback to the setting of normal varieties.
Representation Schemes and Moduli: To handle the essential surjectivity (the isomorphism case), the paper utilizes X. Sun's generalization of Simpson's representation schemes to positive characteristic [Sun19]. This involves constructing moduli spaces of framed, Gieseker semistable vector bundles with numerically vanishing Chern classes.
Hrushovski's Theorem: The proof of essential surjectivity relies on Hrushovski's theorem [Hru04] regarding periodic points of rational maps. By spreading out the situation to a family over a finitely generated base and analyzing the action of the Frobenius morphism on the representation scheme, the authors show that stable bundles with specific Frobenius properties are dense. This allows them to construct the necessary descent data.

Key Contributions and Results

Descent Theorem (Theorem 0.4 / Theorem 2.2):
The paper proves that for a proper surjective map $f: X \to Y$ of connected Noetherian $F$ -finite schemes with geometrically connected fibers, an $F$ -divided bundle on $X$ descends to $Y$ if and only if it is trivial on every geometric fiber. This result is independent of the Bhatt-Scholze theorem and is crucial for the subsequent applications.
Faithful Flatness (Theorem 0.3 and Theorem 3.8):
- Theorem 0.3: If $f: X \to Y$ is a proper surjective morphism with geometrically connected fibers, then $f_*: \pi_1^{\text{F-div}}(X) \to \pi_1^{\text{F-div}}(Y)$ is faithfully flat. This result holds without assuming $Y$ is smooth or $X$ is normal, provided the fibers are geometrically connected.
- Theorem 3.8: For normal projective varieties where $Y$ is smooth and $f_*: \pi_1^{\text{ét}}(X) \to \pi_1^{\text{ét}}(Y)$ is surjective, the induced map on $F$ -divided fundamental groups is faithfully flat. This generalizes previous results by removing the requirement that $f$ be finite.
Isomorphism (Theorem 0.2 and Theorem 3.11):
- Theorem 0.2: If $X$ and $Y$ are normal projective, $Y$ is smooth, and $f_*: \pi_1^{\text{ét}}(X) \to \pi_1^{\text{ét}}(Y)$ is an isomorphism, then $f_*: \pi_1^{\text{F-div}}(X) \to \pi_1^{\text{F-div}}(Y)$ is an isomorphism.
- Theorem 3.11: Under the additional assumption that $f_*\mathcal{O}_X = \mathcal{O}_Y$ (which implies geometrically connected fibers), the isomorphism of étale fundamental groups implies the isomorphism of $F$ -divided fundamental groups.
- The paper notes that the smoothness of $Y$ is currently essential for the isomorphism part due to limitations in understanding the behavior of étale fundamental groups under spreading out for general normal varieties and the reliance on purity theorems in existing proofs of genuinely ramified morphisms.
Generalization of Previous Work:
The results generalize and strengthen the recent work of Sun and Zhang [SZ25], which in turn generalized Esnault-Mehta [EM10] and Biswas-Kumar-Parameswaran [BKP25]. The paper provides a more direct proof for the faithful flatness case using the new descent theorem, avoiding the more involved arguments of [SZ25].

Significance and Claims
The paper claims to provide a definitive answer to the relative Gieseker problem for $F$ -divided bundles in the context of normal projective varieties, subject to the smoothness of the target for the isomorphism case. The primary significance lies in:

The Descent Theorem: The authors present the analogue of the Bhatt-Scholze descent theorem for $F$ -divided bundles as a result of independent interest, noting that it overcomes limitations of applying the original theorem directly to the entire sequence of bundles.
Removal of Smoothness Assumptions: For the faithful flatness result, the paper successfully removes the assumption that $f$ must be a finite morphism or that $X$ must be smooth, relying instead on the geometric connectedness of fibers.
Unification: The work unifies the treatment of stratified sheaves in characteristic zero and $F$ -divided bundles in positive characteristic, confirming that the structural relationship between étale and "differential" (stratified/ $F$ -divided) fundamental groups persists in positive characteristic under appropriate conditions.

The authors explicitly state that the smoothness of $Y$ in the isomorphism theorem is a current limitation of the proof techniques (specifically the reliance on [BKP25] and [SZ25] for the genuinely ramified case) rather than a known necessity of the result itself. They also note that the proof of the second part of Theorem 0.2 depends on Hrushovski's theorem and the boundedness of families of semistable sheaves, hidden within Sun's representation schemes.