Rational points on K3 surfaces of degree 2

Basic Information

• Paper ID: 2505.13262
• Title: Rational points on K3 surfaces of degree 2
• Author: Júlia Martínez-Marín (University of Bristol)
• Classification: math.NT (Number Theory), math.AG (Algebraic Geometry)
• Publication Date: May 2025 (arXiv preprint, revised October 15, 2025)
• Paper Link: https://arxiv.org/abs/2505.13262

Abstract

This paper investigates the problem of rational points on degree 2 K3 surfaces over number fields. The author proves that for degree 2 K3 surfaces (double covers of $\mathbb{P}^2$ branched along a smooth sextic curve), there exists a finite field extension over which the surface has infinitely many rational points, and provides an upper bound on the degree of extension. Furthermore, utilizing van Luijk's ideas and surfaces constructed by Elsenhans-Jahnel, the author provides an explicit family of degree 2 K3 surfaces defined over $\mathbb{Q}$ with geometric Picard number 1 and infinitely many $\mathbb{Q}$-rational points, which is Zariski dense in the moduli space of degree 2 K3 surfaces.

Research Background and Motivation

1. Core Problem: As an important algebraic surface with Kodaira dimension 0, the distribution of rational points on K3 surfaces is a fundamental problem in algebraic geometry and number theory. In particular, whether rational points on K3 surfaces are potentially dense over some finite field extension is an important conjecture.
2. Problem Significance:
   • K3 surfaces are the only known class of surfaces with Kodaira dimension 0 besides rational and ruled surfaces
   • Understanding the distribution of rational points on K3 surfaces is crucial for understanding the arithmetic properties of more general algebraic varieties
   • This is directly related to a problem posed by Swinnerton-Dyer and Poonen in 2002
3. Limitations of Existing Research:
   • While elliptic K3 surfaces (geometric Picard number ≥ 2) are known to have potentially dense rational points
   • For K3 surfaces with geometric Picard number 1, van Luijk first provided examples of quartic K3 surfaces in 2005
   • For degree 2 K3 surfaces, systematic research and explicit constructions are lacking
4. Research Motivation: To generalize van Luijk's method to degree 2 K3 surfaces and provide explicit bounds on the degree of field extension.

Core Contributions

1. Theoretical Result: Proves that for any degree 2 K3 surface X over a number field K, there exists a field extension L/K of degree at most 12 such that X(L) is infinite (Theorem 1.1)
2. Constructive Proof: Provides a constructive method for finding curves of geometric genus 1, proving that such curves can always be found over extensions of degree at most 6 (Lemma 4.1)
3. Application of Elliptic Curve Theory: Proves that for genus 1 curves of the form $y^2 = h(x)$ (where $h$ is a separable quartic polynomial), there exist infinitely many quadratic field extensions such that the curve possesses non-torsion rational points (Proposition 4.2)
4. Explicit Family Construction: Constructs an explicit family of degree 2 K3 surfaces with geometric Picard number 1 and infinitely many $\mathbb{Q}$-rational points, which is Zariski dense in the moduli space (Theorem 1.3)

Detailed Methodology

Problem Definition

Study the distribution of rational points on degree 2 K3 surfaces $X$ defined by the equation $w^2 = f(x,y,z)$, where $f$ is a homogeneous degree 6 polynomial such that the curve $V(f) \subset \mathbb{P}^2$ is smooth. The objectives are:

1. Find an upper bound on the degree of field extension such that $X$ has infinitely many rational points over that extension
2. Construct explicit K3 surface families with specific properties

Core Methodological Framework

Step One: Finding Genus 1 Curves (Lemma 4.1)

1. Tangent Line Method: For a point $P$ on the branch curve $B = V(f) \subset \mathbb{P}^2$, consider the tangent line $\ell$ at $P$
2. Preimage Analysis: Compute the preimage $C = \pi^{-1}(\ell)$ of the tangent line under the double cover $\pi: X \to \mathbb{P}^2$
3. Genus Calculation: Since $\ell$ is tangent to $B$ at $P$, the intersection multiplicity is at least 2, which causes $C$ to have singularities. The normalization $\tilde{C}$ has genus 1

Step Two: Non-Torsion Points on Elliptic Curves (Proposition 4.2)

For genus 1 curves of the form $y^2 = h(x)$ where $h$ is a separable quartic polynomial:

1. Rational Point Construction: For $\alpha \in K$, the curve $C$ acquires a rational point $(α, ±\sqrt{h(α)})$ over the field extension $K_\alpha = K(\sqrt{h(\alpha)})$
2. Application of Merel's Theorem: If all these points were torsion, then by Merel's theorem, their orders would have a uniform bound
3. Proof by Contradiction: However, the $m$-torsion group of an elliptic curve $E[m] \cong \mathbb{Z}/m\mathbb{Z} \times \mathbb{Z}/m\mathbb{Z}$ cannot contain infinitely many distinct points of bounded order

Technical Innovations

1. Constructive Method: Unlike existence proofs, this paper provides a fully constructive proof that allows explicit computation of the required field extension
2. Degree Optimization: Through careful analysis, the required field extension degree is optimized from the naive bound of 24 (6×4) to 12
3. Moduli Space Density: Cleverly utilizes the action of the linear transformation group to prove the density of the constructed family in the moduli space

Experimental Setup

Concrete Example Analysis

Section 3 of the paper provides a concrete example: Surface: $X: w^2 = -x^6 - y^6 - z^6$ in $\mathbb{P}(1,1,1,3)$

Property Verification:

1. This surface has no rational points over $\mathbb{Q}$
2. It has infinitely many rational points over $\mathbb{Q}(i)$: $[1:0:0:i], [0:1:0:i], [0:0:1:i]$
3. It contains the line $V(w-ix^3, y-iz)$

Computational Tools

Uses the Magma computer algebra system for:

1. Computing torsion subgroups of elliptic curves
2. Verifying surface smoothness
3. Computing Picard numbers

Experimental Results

Main Theorem Verification

Theorem 1.1 Verification: For any degree 2 K3 surface, infinitely many rational points can indeed be found over an extension of degree ≤ 12.

Theorem 1.3 Verification: The constructed family $73w^2 = 7(11x^5y + 7x^5z + \ldots + 4z^6 + 15h(x,y,z))$ where $h$ is a homogeneous degree 6 polynomial, satisfies:

1. Geometric Picard number equals 1 (Theorem 5.2)
2. For appropriately chosen $h$, has infinitely many $\mathbb{Q}$-rational points
3. Is Zariski dense in the moduli space

Specific Computational Results

For the case $h = 0$, the elliptic curve obtained is: $\tilde{C}_S: w^2 = \frac{16771780}{1226911}y^4 - \frac{1540220}{175273}y^3 + \frac{81451}{25039}y^2 - \frac{4078}{3577}y + 1$

Computational verification:

• The torsion subgroup of this elliptic curve is trivial
• The rational points $(0, ±1)$ are non-torsion points
• Therefore, the corresponding K3 surface has infinitely many rational points

Related Work

Historical Development

1. Bogomolov-Tschinkel (1998-2000): Proved that elliptic K3 surfaces have potentially dense rational points
2. van Luijk (2005): First constructed explicit examples of K3 surfaces with geometric Picard number 1 (quartic case)
3. Elsenhans-Jahnel (2008): Constructed multiple examples of degree 2 K3 surfaces with geometric Picard number 1

Contributions of This Paper

• Successfully generalizes van Luijk's method to degree 2 K3 surfaces
• Provides more precise bounds on the degree of field extension than previous work
• Offers a fully constructive proof method

Conclusions and Discussion

Main Conclusions

1. Theoretical Bound: Any degree 2 K3 surface has infinitely many rational points over a field extension of degree ≤ 12
2. Constructive Method: Provides an explicit algorithm for finding such extensions
3. Moduli Space Result: Degree 2 K3 surfaces with geometric Picard number 1 and infinitely many rational points are dense in the moduli space

Limitations

1. Optimality of Bound: It remains unclear whether the bound of 12 is optimal
2. Computational Complexity: Finding the concrete field extension for a given K3 surface may require substantial computation
3. Generalization Issues: Whether the method can be extended to K3 surfaces of higher degree remains unclear

Future Directions

1. Improve the bounds on the degree of field extension
2. Study similar problems for K3 surfaces of other degrees
3. Explore more general criteria for Zariski density of rational points

In-Depth Evaluation

Strengths

1. Theoretical Depth: Cleverly combines algebraic geometry, number theory, and elliptic curve theory
2. Methodological Innovation: The constructive proof method is practically implementable
3. Complete Results: Forms a complete theoretical system, from existence to constructivity, from individual examples to moduli space density
4. Technical Excellence: Contains technical innovations at multiple critical steps, particularly in optimizing the degree bound

Weaknesses

1. Sharpness of Bound: The bound of 12 may not be optimal
2. Computational Efficiency: May face complexity issues in practical computation
3. Generalizability: The applicability of the method to other types of surfaces is limited

Impact

1. Theoretical Contribution: Advances the development of arithmetic theory of K3 surfaces
2. Methodological Significance: Provides new tools for studying rational points on algebraic varieties
3. Practical Value: The constructive method facilitates computer verification and further research

Applicable Scenarios

This method is particularly suitable for:

1. Studying rational points on degree 2 K3 surfaces
2. Explicitly constructing K3 surfaces with specific properties
3. Investigating moduli space density problems

References

The paper cites 17 important references, including:

• van Luijk's pioneering work 13
• Elsenhans-Jahnel's construction 7
• Bogomolov-Tschinkel's work on elliptic K3 surfaces 2,4
• Classical results by Merel and Mazur on torsion points of elliptic curves 15,14
• Huybrechts' monograph on K3 surfaces 10

This paper is an important contribution at the intersection of algebraic geometry and number theory. Through clever constructive methods, it resolves fundamental problems concerning rational points on degree 2 K3 surfaces, laying a solid foundation for further research.