🔢 mathematics

Simply branched covers of curves and wild conductor exponents

Using deformation theory, this paper demonstrates that covers of smooth projective curves over $p$ -adic fields can be perturbed into simply branched covers and applies this result to derive a formula for the wild conductor exponent of a curve at a prime $p>d$ based on the ramification data of a degree $d$ cover to $\mathbb{P}^1$ .

Original authors: Harry Spencer

Published 2026-07-28

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Harry Spencer

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

The Shape of Numbers and the Art of Smoothing

Imagine you are a cartographer trying to map a mysterious, winding river that flows through a landscape of pure numbers. In the world of mathematics, specifically a field called algebraic geometry, these "rivers" are actually curves—smooth, continuous shapes defined by equations. Just as a river might twist, turn, or even crash against a rock, these mathematical curves can have points where they fold over themselves or get tangled. These tangles are called "ramification points."

Now, imagine you want to study the hidden properties of this river, like how it behaves when the water gets very cold or very turbulent. Mathematicians have a special tool for this called the "conductor," which acts like a weather report for the curve, telling us how "wild" or chaotic its behavior is at specific points. However, calculating this weather report is incredibly difficult if the river is tangled up in a messy knot. It turns out that if the river flows smoothly, with only simple, gentle bends (mathematicians call this "simple branching"), the weather report is easy to read. The big question has always been: Can we take any messy, tangled river and gently nudge it until it flows smoothly, without changing its fundamental nature? If we could, we could use the easy-to-read weather reports of the smooth version to understand the messy original. This is exactly the puzzle Harry Spencer tackles in his paper.

Smoothing the Tangles: A Mathematical Massage

In this paper, Harry Spencer acts like a mathematical masseur, but instead of working on muscles, he is working on the geometry of curves over "p-adic fields." Think of a p-adic field as a very strange, zoomed-in way of looking at numbers, where distance is measured by how many times a number can be divided by a specific prime number (like 5 or 7) rather than by how far apart they are on a number line. In this weird, zoomed-in world, Spencer proves a powerful new fact: You can always take a messy, tangled cover of curves and "perturb" it—give it a tiny, precise nudge—to turn it into a "simply branched" cover.

To visualize this, imagine a complex origami crane made of paper that has been crumpled and folded in a chaotic way. Spencer shows that if you are working in this specific p-adic universe, you can apply a microscopic, almost invisible pressure to the paper. This pressure doesn't rip the paper or change the fact that it's a crane; it just smooths out the crumples until the folds are perfectly simple and neat. Once the crane is smooth, it becomes much easier to study.

The paper's main finding is that this "smoothing" isn't just a lucky guess; it's a guaranteed result. Spencer uses a technique called "deformation theory," which is like a time-lapse camera that shows how a shape changes as you slowly adjust its settings. He proves that there is a whole family of shapes surrounding your original messy curve, and within that family, you can always find a version that is perfectly simple. Even better, he shows that these simple versions are so common that if you look at any tiny neighborhood around your original curve, you will almost certainly find a simple one right next to it.

The Payoff: Reading the Wild Weather

Why does this matter? The paper connects this smoothing trick to a specific calculation called the "wild conductor exponent." This number is a key part of a grand formula that mathematicians use to predict the behavior of these curves, similar to how a meteorologist uses data to predict a storm. Previously, there was a formula to calculate this number, but it only worked if the curve was already perfectly smooth (simply branched). If the curve was messy, the formula was useless, and calculating the number was a nightmare.

Spencer's work removes that roadblock. He proves that because you can always nudge a messy curve into a smooth one without changing its core identity, you can now use the easy formula on the smooth version to find the answer for the messy original. He provides a specific recipe: take any cover of a curve (like a map from a complex shape to a simple line), look at where it gets tangled, and plug those details into his new formula. However, there are two important rules for this recipe to work: First, the prime number p involved must be larger than the degree of the cover (the "size" of the map). Second, you must arrange the map so that the point at infinity is not a branch point (a condition that can always be arranged by rotating the view). If these conditions are met, the formula works perfectly.

The paper also tackles a specific question: Can we smooth the curve without changing the curve itself? Usually, when you nudge a shape, you might end up with a slightly different shape. Spencer shows that if the curve is embedded in a space in a certain "very ample" way (meaning the map defining the curve is sufficiently rich and flexible), then you can indeed find a smooth version that sits on the exact same curve. It's like being able to untangle a knot in a rope without ever cutting the rope or tying a new one; you just slide the loops around until they fall into place. If the curve isn't "very ample," you might still be able to smooth the map, but you may need to change the curve itself to do so.

A Real-World Example from the Paper

To prove his theory works, Spencer looks at a specific example involving a curve of genus 12 (a fancy way of saying it has 12 "holes" or handles, like a pretzel with many loops). This curve lives in a high-dimensional space and is defined by some complicated equations. When projected onto a line, it creates a map that is messy and tangled in a specific way. Using his new method, he calculates the "wild conductor exponent" for this curve.

He finds that the answer is 18. Before this paper, calculating this number for such a messy curve would have been nearly impossible using existing tools. But by applying his "smoothing" logic, he could bypass the mess and get the answer directly. He also explains why this curve is special: it has a unique way of mapping to a line, which means there's no other "smooth" version of this specific map to confuse the results. This confirms that his method is robust and can handle real, complicated mathematical objects.

In short, Spencer has given mathematicians a new, reliable tool to untangle the knots in their number-theoretic maps. By proving that every messy cover can be gently nudged into a simple one, he has opened the door to calculating important invariants that were previously out of reach, turning a chaotic landscape into a navigable one.

Technical Summary: Simply Branched Covers of Curves and Wild Conductor Exponents

Problem Statement
The paper addresses the arithmetic geometry of covers of curves over $p$ -adic fields. Specifically, it seeks to establish an arithmetic analogue of the classical principle that simple branching is a generic property of covers. Over the complex numbers, any cover of curves can be perturbed to a simply branched one (where every ramification point has index 2 and no fiber contains more than one ramification point). The author aims to prove that for a cover $\pi: C \to D$ of smooth, projective, geometrically connected curves over a $p$ -adic field $K$ , one can find a $p$ -adically close cover $\pi': C' \to D$ that is simply branched.

A primary motivation for this result is the computation of wild conductor exponents ( $n_{wild}$ ) of curves. Recent work by the author ([18]) provided a formula for $n_{wild}$ based on the branch locus of a simply branched cover. However, this formula was previously restricted to cases where the cover was already simply branched or required ad hoc perturbations for specific plane curves. The goal is to remove the "simply branched" hypothesis from these conductor formulas, allowing the computation of $n_{wild}$ for arbitrary covers by perturbing them to the simply branched case.

Methodology
The paper employs deformation theory combined with $p$ -adic approximation techniques. The core strategy, inspired by Schröer [16], involves the following steps:

Iterative Lifting: The author lifts the given cover $\pi: C \to D$ $π : C \to D$ to a formal deformation over $K[[t]]$ $K [[t]]$ . This is achieved by iteratively lifting the cover through the rings $K[t]/(t^{n+1})$ $K [t] / (t^{n + 1})$ .
- The obstruction to lifting lies in $\text{Ext}^2(L^\bullet_{C/D}, \mathcal{O}_C)$ , which vanishes for covers of smooth curves (Proposition 2.3).
- The set of first-order deformations is a torsor under $\text{Ext}^1(L^\bullet_{C/D}, \mathcal{O}_C)$ .
Separation of Ramification: To ensure the generic fiber of the deformation is simply branched, the author constructs a specific first-order deformation. By analyzing the local structure of ramification points (Lemma 2.5), it is shown that one can choose a deformation parameter such that:
- Ramification points of index $e \ge 3$ split into $e-1$ distinct points of index 2 (Lemma 2.10).
- Ramification points lying in the same fiber are separated into distinct fibers.
- This is achieved by selecting a deformation class in a specific Zariski-open subset $U$ of the deformation space (Lemma 2.9).
Algebraisation and Descent: Using Grothendieck's algebraisation theorems [6], the formal deformation over $K[[t]]$ is algebraised to a morphism over a finite type scheme $T$ .
$p$ -adic Approximation: To ensure the existence of a $K$ -rational point in the simply branched locus that is $p$ -adically close to the original cover, the author applies Greenberg's Approximation Theorem [5]. This allows the descent of the formal solution to a rational point in a $p$ -adic neighborhood of the origin in the parameter space $T$ .

Key Results

Theorem 1.1 (Theorem 2.7): Let $K/\mathbb{Q}_p$ be a finite extension and $\pi: C \to D$ a degree $d$ cover of smooth, projective, geometrically connected curves. There exists a smooth, geometrically integral $K$ -scheme $T$ with a marked point $0$, a smooth proper family of curves $q_T: \mathcal{C} \to T$ , and a finite flat morphism $\Pi_T: \mathcal{C} \to D \times_K T$ such that the fiber at $0$ is $\pi$ . Crucially, the set of points $t \in T(K)$ where the fiber $\Pi_{T,t}$ is simply branched is Zariski-open, and every $p$ -adic neighborhood of $0$ contains a $K$ -rational point of this simply branched locus.
Proposition 1.2 (Proposition 3.2): If the target is $\mathbb{P}^1_K$ and the line bundle $L = \pi^*\mathcal{O}_{\mathbb{P}^1}(1)$ is very ample, then the source curve $C$ itself can be preserved. Specifically, every $p$ -adic neighborhood of the pencil defining $\pi$ in the Grassmannian $\text{Gr}(2, H^0(C, L))$ contains a base-point free pencil defining a simply branched cover $C \to \mathbb{P}^1_K$ .
Corollary 1.3: For a curve $C$ over $K$ and a cover $\pi: C \to \mathbb{P}^1_K$ where $\infty$ is not a branch point, if $p > \deg(\pi)$ , the wild conductor exponent is given by:
$n_{wild}(C) = w_K(\pi) = \sum_{b \in B} \sum_{r \in \pi^{-1}(b)} (e_r - 1) \cdot (v_K(\Delta_{K(b)/K}) - [K(b):K] + f_{K(b)/K})$
This formula holds for any cover $\pi$ , not just simply branched ones, because $\pi$ can be perturbed to a simply branched cover $\pi'$ with the same conductor exponent (due to local constancy of conductors in smooth families) and the same value of $w_K$ .

Significance and Claims
The paper claims to provide a rigorous arithmetic foundation for the "genericity" of simple branching over $p$ -adic fields. Its significance lies in:

Generalization: It extends previous deformation-theoretic results (which often assumed algebraically closed fields or characteristic 0) to non-algebraically closed $p$ -adic fields.
Application to Conductors: It removes the restrictive "simply branched" hypothesis from the author's previous formula for wild conductor exponents ([18]). This allows the computation of $n_{wild}$ for arbitrary covers, provided the characteristic $p$ is sufficiently large relative to the degree of the cover.
Fixed Source Perturbations: It establishes conditions (very ampleness of the pullback line bundle) under which the source curve $C$ need not be deformed; only the map to $\mathbb{P}^1$ needs to be perturbed.

The author notes that while the proof strategy uses techniques familiar to experts (deformation theory, Grothendieck algebraisation), the extension to non-algebraically closed fields and the specific application of $p$ -adic approximation to this problem appear to be new contributions to the literature. The work does not claim to solve the conductor problem for all primes, but specifically for primes $p > \deg(\pi)$ (or $p > \gamma_{rat}$ in the context of rational gonality), where the perturbation argument holds.

The Shape of Numbers and the Art of Smoothing

Smoothing the Tangles: A Mathematical Massage

The Payoff: Reading the Wild Weather

A Real-World Example from the Paper

Technical Summary: Simply Branched Covers of Curves and Wild Conductor Exponents

More like this