Rational points on K3 surfaces of degree 2

Júlia Martínez-Marín · 2025

Read Rational points on K3 surfaces of degree 2 by Júlia Martínez-Marín (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-13262.txt.

Introduction

Júlia Martínez-Marín studies rational points on a special class of surfaces in a 2025 mathematics paper. A K3 surface over a number field always gains infinitely many rational points after passing to some larger field. The question is how large that field must be. Her surfaces have degree two, built as double covers of the plane branched over a six-degree curve.

She bounds the degree of the field extension needed. Borrowing ideas from van Luijk and a surface built by Elsenhans and Jahnel, she also constructs an explicit family of such surfaces over the rationals. The family has the smallest possible Picard number yet infinitely many rational points, and it fills the moduli space densely. The work sharpens number theory's grip on these surfaces.

Catalog record

CREATORJúlia Martínez-Marín
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-13262

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Rational points on K3 surfaces of degree 2 Júlia Martínez-Marín Abstract . A K3 surface over a number field has infinitely many rational points over a finite field extension. For K3 surfaces of degree 2, arising as double covers of ℙ 2 \mathbb{P}^{2} branched along a smooth sextic curve, we give a bound for the degree of such an extension. Moreover, using ideas of van Luijk and a surface constructed by Elsenhans and Jahnel, we give an explicit family of K3 surfaces of degree 2 defined over ℚ \mathbb{Q} with geometric Picard number 1 and infinitely many ℚ \mathbb{Q} -rational points that is Zariski dense in the moduli space of K3 surfaces of degree 2. 2020 Mathematics Subject Classification Primary 14G05; Secondary 14J28, 11G35 1. Introduction A K3 surface over an algebraically closed field contains a rational (possibly singular) curve [ 16 , Appendix] . Therefore a K3 surface over a number field K has infinitely many rational points over a finite field extension of K over which the rational curve is defined and has a rational point. For a K3 surface X of degree 2, i.e. a double cover of \mathbb{P}^{2} branched along a smooth sextic curve, we find a bound (independent of X and K ) for the degree of an extension L of K such that X has infinitely many L -rational points. It remains open whether the bound is sharp or not. Theorem 1.1 . Let X be a K3 surface of degree 2 over a number field K . Then there is a field extension L/K with [L:K]\leq 12 such that X(L) is infinite. We may then wonder whether the set of rational points over a finite field extension of K is Zariski dense in X . If so, we say that the rational points on X are potentially dense . Apart from K3 surfaces, all other surfaces of Kodaira dimension \kappa=0 have potentially dense rational points [ 9 , Proposition 4.2] , [ 3 , Theorem 3.1] . For K3 surfaces, it is conjectured that the rational points are also potentially dense. In [ 4 ] , Bogomolov and Tschinkel show that any double cover of \mathbb{P}^{2} branched along a sextic curve with a singular point has potentially dense rational points. The blow-up in the preimage of the singular point under the double cover is an elliptic K3 surface. More generally, Bogomolov and Tschinkel proved potential density of rational points for all elliptic K3 surfaces [ 2 ] . Such surfaces have geometric Picard number \rho\geq 2 . In [ 13 ] , van Luijk gave the first explicit example of a K3 surface over \mathbb{Q} with geometric Picard number \rho=1 . With this K3 surface, which is of degree 4, he also gave a positive answer to the question of whether there exists a K3 surface over a number field with infinitely many rational points and geometric Picard number 1, asked by Swinnerton-Dyer and Poonen at a workshop in 2002. Using this, he proved the following theorem. Theorem 1.2 (van Luijk, 2005 [ 13 ] ) . In the moduli space of K3 surfaces of degree 4, the set of surfaces defined over \mathbb{Q} with geometric Picard number 1 and infinitely many \mathbb{Q} -rational points is Zariski dense. Using van Luijk’s method, Elsenhans and Jahnel constructed many examples of K3 surfaces of degree 2 with geometric Picard number 1 in [ 7 ] . Modifying slightly one of their examples and using the method developed in the proof of Theorem 1.1 to find infinitely many rational points, we prove the analogue of van Luijk’s theorem for K3 surfaces of degree 2. Theorem 1.3 . In the (coarse) moduli space of K3 surfaces of degree 2, the set of surfaces defined over \mathbb{Q} with geometric Picard number 1 and infinitely many \mathbb{Q} -rational points is Zariski dense. The proof of Theorem 1.1 is divided into two steps. The first consists of finding a curve of geometric genus 1 lying on the K3 surface. This can always be done up to enlarging the base field by an extension of degree at most 6, as we prove in Lemma 4.1 . The proof of this lemma is constructive, so given an explicit K3 surface X of degree 2 over a number field, one can find an explicit extension over which X contains a curve of geometric genus 1. In fact, in most cases it will suffice to take the smallest extension L of K over which the branch curve B of X\rightarrow\mathbb{P}^{2} has B(L)\neq\emptyset . This gives a way of finding examples of K3 surfaces (of degree 2) with infinitely many rational points over their base field, and this is the key to proving Theorem 1.3 . See Remark 4.4 for more details on this. The second ingredient to prove Theorem 1.1 involves showing that there is a degree 2 extension L of a number field K over which a smooth curve C of genus 1 given by y^{2}=h(x) for a polynomial h\in K[x] of degree 4 has infinitely many L -rational points. It is not hard to see that this happens if we allow an extension of degree 4 since we can find a quadratic extension K^{\prime} of K such that C(K^{\prime})\neq\emptyset , so C becomes an elliptic curve over K^{\prime} . Then we can use Merel’s theorem [ 17 , p. 242] to see that C_{K^{\prime}} must have a non-torsion point over some quadratic extension K^{\prime\prime} of K^{\prime} , i.e. an extension of K of degree 4. In Proposition 4.2 we do both steps at once and show that we only require a degree 2 extension. 1.1. Outline We begin in Section 2 with a brief review on K3 surfaces and relevant invariants of algebraic surfaces. In Section 3 , we discuss an example of a K3 surface of degree 2 with no rational points over \mathbb{Q} but Zariski dense rational points over a quadratic field extension. In Section 4 , we prove Lemma 4.1 , which says that a K3 surface of degree 2 over a number field K contains a singular curve of geometric genus 1 over an extension K^{\prime} of K of degree at most 6. We show that the normalisation C of this curve is given by y^{2}=h(x) , for a separable polynomial h\in K^{\prime}[x] of degree 4. Proposition 4.2 now says that C becomes an elliptic curve with a non-torsion point over infinitely many quadratic extensions. These two results are combined to prove Theorem 1.1 . In Section 5 , we give an explicit family of K3 surfaces of degree 2 defined over \mathbb{Q} with geometric Picard number 1 and infinitely many rational points. Theorem 1.3 follows once we show that this family is Zariski dense in the moduli space of K3 surfaces of degree 2. 2. Preliminaries Let X be a smooth, projective surface over a field k . The Néron–Severi group of X is the quotient \operatorname{NS}X\coloneq\operatorname{Pic}X/\operatorname{Pic}^{0}X, where \operatorname{Pic}^{0}X denotes the connected component of the Picard variety \operatorname{Pic}X containing the identity element. The Néron–Severi group \operatorname{NS}X is a finitely generated abelian group [ 10 , p. 7] . The Picard number of X is \operatorname{rk}\left(\operatorname{NS}X\right) and its

Read the complete text