🔢 mathematics

Rational points on smooth surfaces in $\mathbb{P}^3$ over finite fields

This paper improves an existing bound on the number of rational points on smooth surfaces in $\mathbb{P}^3$ over finite fields and computes the exact counts for specific families of surfaces that achieve or nearly achieve this bound.

Original authors: Yves Aubry, José Felipe Voloch

Published 2026-05-12

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Yves Aubry, José Felipe Voloch

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 a detective trying to count the number of "treasure spots" (rational points) hidden on a giant, smooth, 3D sculpture (a surface) floating in a mathematical universe. This universe is built not on infinite numbers like ours, but on a finite grid called a finite field. Think of this grid as a giant, flat checkerboard where the numbers wrap around like a clock; once you reach a certain number (the prime number $p$ ), you start over at zero.

The paper by Yves Aubry and José Felipe Voloch is about finding the maximum possible number of treasure spots that can exist on these specific sculptures, and then building examples of sculptures that actually hit (or get very close to) that maximum.

Here is a breakdown of their work using simple analogies:

1. The Goal: Counting the Dots

Mathematicians have known for a long time that there are limits to how many dots can exist on these shapes. It's like knowing a bucket can only hold a certain amount of water.

The Old Bucket: A famous mathematician named Deligne gave a "bucket size" (a bound) for how many dots could fit.
The Better Bucket: The second author of this paper previously found a slightly smaller bucket size for certain types of sculptures.
The New Bucket: This paper introduces an even smaller, more precise bucket. They figured out that if you know how many "straight lines" are drawn on the sculpture, you can predict the maximum number of dots even more accurately.

2. The "Line" Factor

Imagine the sculpture is made of smooth clay. Sometimes, you can find straight lines carved into it.

The authors discovered that the more straight lines ( $m$ ) the sculpture has, the fewer "extra" dots you can have elsewhere.
They created a new formula (a mathematical recipe) that subtracts points based on how many lines are present. It's like saying, "If you use up your space carving 10 lines, you have less room left for scattered dots."
The Result: Their new formula is tighter (smaller) than previous ones, meaning it gives a more accurate "ceiling" for the number of dots.

3. The "Fingerprint" Test (Surfaces $S_1$ and $S_2$ )

How do they know where the dots are? They use a clever trick involving "fingerprints."

They take the equation that defines the sculpture and create two new, slightly different equations (called $S_1$ and $S_2$ ).
Where the original sculpture ( $S$ ) intersects with these two new "shadow" sculptures, they find the dots.
The Magic: Most dots on the original sculpture show up as "heavy" intersections (multiplicity 6) in this shadow world. However, the straight lines show up as "light" intersections (multiplicity 1).
By counting these intersections, they can separate the "line dots" from the "scattered dots" and get a precise count.

4. The "Perfect" Examples

The authors didn't just make up a formula; they built specific sculptures to test it.

The Fermat Sculpture: They looked at a classic shape defined by $x^d + y^d - z^d - w^d = 0$ . They proved that for certain prime numbers, these shapes are packed with straight lines, and their formula works perfectly to count the dots.
The "No-Line" Sculpture: They also built a strange sculpture using a special 5th-root-of-unity pattern. They proved this one has zero straight lines.
- Why is this cool? Because it has no lines, the "line subtraction" part of their formula disappears, and they could calculate the exact number of dots purely based on the shape's geometry.
- They found that for most prime numbers, the dots on this shape follow a very clean, predictable pattern. However, for a few "tricky" primes (like 11, 41, and 61), the pattern breaks slightly because the math behaves differently, creating "extra" dots.

5. The "Sum of Roots" Puzzle

To prove the "No-Line" sculpture really has no lines, they had to solve a puzzle about adding up numbers.

Imagine you have a bag of 5 special numbers (roots of unity).
They proved that for most prime numbers, the only way to add up to zero using these numbers is to use all 5 distinct ones.
However, for the "tricky" primes (11, 41, 61), you can add up fewer numbers or repeat them to get zero. This explains why those specific sculptures have a few extra dots that don't fit the standard pattern.

Summary

In short, Aubry and Voloch have:

Refined the limit: They gave a better, tighter rule for the maximum number of dots on these 3D shapes.
Used lines as a tool: They showed that counting the straight lines on the shape helps you calculate the dot count more precisely.
Built test cases: They created specific mathematical sculptures where they could count the dots exactly, proving their new rule works and showing exactly how the dots are arranged.

They didn't claim this helps with medicine or engineering; they simply wanted to solve a beautiful, abstract puzzle about how numbers and shapes interact in a finite world.

Technical Summary: Rational Points on Smooth Surfaces in $\mathbb{P}^3$ over Finite Fields

Problem Statement
The paper addresses the problem of bounding and computing the number of rational points, denoted $\sharp S(\mathbb{F}_p)$ , on a smooth surface $S$ of degree $d$ defined over a finite field $\mathbb{F}_p$ (where $p$ is prime). Specifically, the authors focus on the case where $2 < d < p$ . While Deligne's proof of the Weil conjectures provides a general bound, and previous elementary methods have established other limits, the authors aim to refine a specific bound previously derived by the second author (Voloch) in [10]. This previous bound, $\sharp S(\mathbb{F}_p) \le \frac{d(d+p-1)(d+2p-2)}{6} + m(p+1)$ , depends on $m$ , the number of $\mathbb{F}_p$ -lines contained in $S$ . The paper seeks to improve this bound and identify families of surfaces that achieve or approach it.

Methodology
The core methodology involves the intersection theory of algebraic surfaces within projective space $\mathbb{P}^3$ .

Construction of Auxiliary Surfaces: Given a surface $S$ $S$ defined by a homogeneous polynomial $f$ $f$ , the authors construct two auxiliary surfaces, $S_1$ $S_{1}$ and $S_2$ $S_{2}$ .
- $S_1$ is the zero locus of $\sum_{i=0}^3 (\partial f / \partial x_i) x_i^p$ .
- $S_2$ is the zero locus of $\sum_{i,j=0}^3 (\partial^2 f / \partial x_i \partial x_j) x_i^p x_j^p$ .
Intersection Scheme: The authors analyze the scheme $X := S \cap S_1 \cap S_2$ . Previous work established that the one-dimensional components of $X$ are the lines contained in $S$ . The authors focus on the zero-dimensional part of $X$ , denoted $\Gamma$ , which consists of isolated rational points.
Intersection Theory and Multiplicity:
- Lemma 2.1: The authors establish a relationship between the degree of the zero-dimensional scheme $\Gamma$ , the number of lines $m$ , and the intersection numbers of these lines. This lemma relies on the adjunction formula and classical intersection theory on surfaces to bound the degree of $\Gamma$ .
- Multiplicity Analysis: A critical step involves determining the multiplicity of points in $X$ . The authors note that rational points on $S$ not lying on a line appear in $X$ with multiplicity at least 6.
- Line Multiplicity: The authors investigate whether the lines in $X$ occur with multiplicity one. While they cannot prove this generally, they verify it for specific families (e.g., Fermat surfaces) and via computer checks on examples.
Combinatorial Counting: For specific families of surfaces (Fermat-type and others), the authors perform explicit combinatorial counting of rational points by analyzing the structure of the defining equations and the properties of roots of unity in $\mathbb{F}_p$ .

Key Contributions and Results

Improved Upper Bound (Theorem 2.2):
The primary theoretical contribution is an improved upper bound for $\sharp S(\mathbb{F}_p)$ when $2 < d < p$ . Assuming the lines in $X$ have multiplicity one, the new bound is:
$\sharp S(\mathbb{F}_p) \le \frac{d(d+p-1)(d+2p-2)}{6} - \frac{3m(d+p-1) - m(m+1)}{6} + m(p+1)$
This improves upon the previous bound (1) whenever the number of lines $m$ satisfies $m \le 3d + 3p - 4$ . Given that a smooth surface of degree $d$ contains at most $11d^2 - 30d + 18$ lines, the improvement holds for sufficiently large $p$ (specifically $p > \frac{11d^2 - 33d + 22}{3}$ ).
Exact Computations for Specific Families:
The authors compute the exact number of rational points for several families of surfaces, demonstrating cases where the bounds are tight or nearly tight:
- Fermat Surfaces: For surfaces defined by $x^d + y^d - z^d - w^d = 0$ (with $p \equiv 1 \pmod d$ and $-1$ being a $d$ -th power), the authors prove the existence of $3d^2$ lines and show that all lines in $X$ have multiplicity one.
- Surfaces without Isolated Points: For the surface defined by $x^{(p-1)/2} + y^{(p-1)/2} - z^{(p-1)/2} - w^{(p-1)/2} = 0$ , the authors derive the exact count $\sharp S(\mathbb{F}_p) = \frac{3}{8}(p^3 - 3p^2 + 11p - 9)$ . Crucially, they prove that for this surface, the scheme $X$ contains no isolated rational points (i.e., $\deg \Gamma = 0$ ), meaning all rational points lie on lines.
- Surfaces without Rational Lines: The authors analyze surfaces defined by equations involving 5-th roots of unity, such as $x^{2(p-1)/5} + y^{2(p-1)/5} + x^{(p-1)/5}y^{(p-1)/5} + z^{2(p-1)/5} + w^{2(p-1)/5} = 0$ . They prove these surfaces contain no $\mathbb{F}_p$ -lines (for $p \neq 11, 41, 61$ ) and provide exact point counts based on the distribution of 5-th roots of unity.
- Other Examples: The paper includes examples involving 7-th roots of unity and other polynomial constructions, providing lower bounds on the number of points for specific primes (e.g., $p=29, 43, 71$ ).

Significance and Claims
The paper claims to improve the known upper bound for the number of rational points on smooth surfaces in $\mathbb{P}^3$ over finite fields, specifically refining the bound dependent on the number of lines contained in the surface. The authors state that their computations for specific families of surfaces "may have independent interest," particularly because they determine exact point counts and the number of lines for these cases.

The authors are modest regarding the generality of their assumptions; specifically, they acknowledge they have not proven that lines in $X$ always occur with multiplicity one under the general hypotheses of Theorem 2.2, though they expect this to be true and have verified it for Fermat surfaces and several examples. The paper does not propose new applications beyond the theoretical advancement of bounds and the enumeration of points on specific algebraic varieties.