Solubility of a family of conics with polynomial coefficients in many variables
Mathieu Da Silva · 2025
Read Solubility of a family of conics with polynomial coefficients in many variables by Mathieu Da Silva (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-2511-20282.txt.
Introduction
Da Silva counts how often a family of conic curves has rational solutions. His 2025 paper studies conics with polynomial coefficients. Each conic mixes three squared variables with polynomial weights. The weights depend on many further variables. Whether a rational point exists varies across the family.
An asymptotic formula counts the solvable cases. Parameters of bounded height are tallied precisely. The result matches two standing conjectures in the area. Loughran-Smeets and Loughran-Rome-Sofos both predicted this behavior. The proof leans on a recent circle method advance. Destagnol, Lyczak, and Sofos supplied that key input. Arithmetic functions are averaged over many-variable polynomials. A simpler diagonal conic family is analyzed first. Its coefficients range over arithmetic progressions. Solubility proportions there feed the main count. Analytic number theory gains a sharp family census. Rational point statistics grow more predictable. The circle method stretches to broader polynomial families.
Catalog record
| CREATOR | Mathieu Da Silva |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 120 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-20282
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Solubility of a family of conics with polynomial coefficients in many variables Mathieu Da Silva Abstract We study the proportion of conics given by (\mathcal{C}_{\bm{F},\bm{y}}):F_{0}(\bm{y})x_{0}^{2}+F_{1}(\bm{y})x_{1}^{2}=F_{2}(\bm{y})x_{2}^{2} which have a rational point \bm{x}=(x_{0}:x_{1}:x_{2})\in\mathbb{P}^{2}(\mathbb{Q}) , where \bm{y}=(y_{0}:\dots:y_{n})\in\mathbb{P}^{n}(\mathbb{Q}) and F_{0},F_{1},F_{2}\in\mathbb{Z}[X_{0},\ldots,X_{n}] are homogeneous polynomials in many variables of the same degree d . We provide an asymptotic formula for the number of \bm{y} of bounded height such that the corresponding conic (\mathcal{C}_{\bm{F},\bm{y}}) has a rational point. In particular, our result agrees with the Loughran–Smeets and the Loughran–Rome–Sofos conjectures. Our strategy is based on a recent result of Destagnol–Lyczak–Sofos relying on the circle method to estimate the average of an arithmetic function over polynomials in many variables. To this end, we study the proportion of conics t_{0}x_{0}^{2}+t_{1}x_{1}^{2}+t_{2}x_{2}^{2}=0 having a rational point, and coefficients t_{0},t_{1},t_{2} in arithmetic progressions. Notation • The letter p is exclusively used for prime numbers. We write \mathcal{P} for the set of primes. If n,m\in\penalty\ \mathbb{Z}_{>0} , the notation p^{m}\mid\mid n means that p^{m}\mid n and p^{m+1}\nmid n . If 2^{m}\mid\mid n , we denote by n_{\mathrm{odd}} the integer 2^{-m}n . • If {\bm{n}}=(n_{1},\ldots,n_{k})\in\mathbb{Z}^{k} , we write \left|\bm{n}\right|:=n_{1}+\cdots+n_{k} and \bm{n}_{\mathrm{odd}}:=((n_{0})_{\mathrm{odd}},\dots,(n_{k})_{\mathrm{odd}}) . • For k\geqslant 2 , the notation \log_{k} stands for the k^{\text{th}} iterate of \log . • If f:(\mathbb{Z}_{>0})^{k}\to\mathbb{C} is an arithmetic function, we write f({\bm{n}}) for f(n_{1},\ldots,n_{k}) and we denote its associated Dirichlet series by D_{f}({\bm{s}})=D_{f}(s_{1},\ldots,s_{k}) . We call f multiplicative if it satisfies the equality f(a_{1}b_{1},\dots,a_{k}b_{k})=f(\bm{a})f(\bm{b}) for all \bm{a},\bm{b}\in(\mathbb{Z}_{>0})^{k} such that \gcd(a_{1}\cdots a_{k},b_{1}\cdots b_{k})=1 . • If s is a complex number, we denote by \sigma its real part. • For z\in\mathbb{C} we denote by \tau_{z}(n) the n -th coefficient in the development of \zeta(s)^{z} ( \sigma>1 ) as a Dirichlet series. Its existence is justified for instance in [ 25 , II.5.1] . If {\bm{z}}=(z_{1},\ldots,z_{k})\in\mathbb{C}^{k} , we introduce \bm{\tau}_{\bm{z}}({\bm{n}}):=\prod_{i=1}^{k}\tau_{z_{i}}(n_{i}) . The latter defines an arithmetic function in k variables which is multiplicative and generalises the \tau_{z} function. • The Gamma function is the function defined by \Gamma(z):=\int_{0}^{+\infty}e^{-t}t^{z-1}dt for \sigma>0 . We recall that it admits an analytic continuation that is meromorphic with simple poles at each non positive integer. • If \ell_{0},\ell_{1},\ell_{2} ( resp. m_{01},m_{02},m_{12} ) are three integers, we set \ell_{012}:=\ell_{0}\ell_{1}\ell_{2} ( resp. m_{012}:=m_{01}m_{02}m_{12} ). • The notation \mathbb{A}_{\mathbb{Q}} stands for the ring of adeles of \mathbb{Q} , that is the set of sequences (a_{\infty},a_{2},a_{3},\dots,a_{p},\dots) where a_{\infty}\in\mathbb{R} , and a_{p}\in\mathbb{Z}_{p} for all but finitely many p for which we have a_{p}\in\mathbb{Q}_{p} . • If K is a field and if \bm{t}=(t_{0},t_{1},t_{2})\in K^{3} , we set \vartheta_{K}(\bm{t}):=\mathds{1}\left(t_{0}x_{0}^{2}+t_{1}x_{1}^{2}=t_{2}x_{2}^{2}\text{ has a solution $\bm{x}\in K^{3}\setminus\{(0,0,0)\}$}\right). ( 1 ) If p is a prime and \bm{t}\in\mathbb{Q}^{3} , we simply write \vartheta_{p}(\bm{t}) for \vartheta_{\mathbb{Q}_{p}}(\bm{t}) . We denote by (\cdot,\cdot)_{K} the Hilbert symbol over K (see [ 22 , chap. 3] ). If K=\mathbb{Q}_{p} , we prefer the notation (\cdot,\cdot)_{p} instead of (\cdot,\cdot)_{\mathbb{Q}_{p}} . • In this paper, an algebraic variety is a \mathbb{Q} -scheme which is reduced, separated and of finite type. If X is an algebraic variety, we denote by K_{X} its canonical divisor and by \Lambda_{\mathrm{eff}}(X) its pseudo-effective cone, i.e. the closure of the cone of effective divisors in \mathrm{Pic}(X)\otimes_{\mathbb{Z}}\mathbb{R} . For each point x of a scheme X , we denote by \kappa(x) its residue field. The set of codimension one points of X is denoted X^{(1)} . • Finally, we denote by \mu_{\infty} the Lebesgue measure on \mathbb{R}^{n+1} and by \mu_{p} the Haar measure on \mathbb{Q}_{p}^{n+1} , normalised by \mu_{p}(\mathbb{Z}_{p}^{n+1})=1 . 1 Introduction Motivations and first definitions. Let n\in\mathbb{Z}_{>0} and \bm{y}\in\mathbb{P}^{n}(\mathbb{Q}) . We recall that the anti-canonical height on \mathbb{P}^{n}(\mathbb{Q}) is defined by H({\bm{y}}):=\max(\left|y_{0}\right|,\ldots,\left|y_{n}\right|)^{n+1}, where (y_{0}:\cdots:y_{n}) is a representative of \bm{y} with coprime coordinates. The integer H(\bm{y}) does not depend on the choice of such a representative of \bm{y} . Let X be a smooth and proper projective variety over \mathbb{Q} and \pi\penalty\ :\penalty\ X\penalty\ \to\penalty\ \mathbb{P}^{n} a dominant morphism whose generic fibre is geometrically integral. The fibres of \pi define a family of varieties. The problem we are interested in consists in estimating, when B goes to +\infty , the quantity N_{\mathrm{loc}}(\pi,B):=\#\left\{\bm{y}\in\mathbb{P}^{n}(\mathbb{Q}):H(\bm{y})\leqslant B,\bm{y}\in\pi(X(\mathbb{A}_{\mathbb{Q}}))\right\}, which is a natural upper-bound for the quantity N_{\text{glob}}(\pi,B):=\#\left\{\bm{y}\in\mathbb{P}^{n}(\mathbb{Q}):H(\bm{y})\leqslant B,\bm{y}\in\pi(X(\mathbb{Q}))\right\}. The Loughran–Smeets conjecture. In 2016, Loughran and Smeets [ 15 , th. 1.2 and 1.3] used the large sieve and the Ekedahl sieve to obtain an upper-bound for N_{\mathrm{loc}}(\pi,B) (and hence for N_{\mathrm{glob}}(\pi,B) ), which is valid for every dominant morphism \pi:X\to\mathbb{P}^{n} whose generic fibre is geometrically integral, where X is a smooth and proper projective variety over a number field. The Loughran–Smeets conjecture [ 15 , 1.6] states that this upper-bound is sharp for N_{\mathrm{loc}}(\pi,B) under some mild assumptions. More precisely, it predicts that N_{\mathrm{loc}}(\pi,B)\underset{B\to+\infty}{\sim}c{B\over(\log B)^{\Delta(\pi)}} for some constant c>0 and \Delta(\pi)\in\mathbb{Q}^{+} defined by [ 15 , (1.3)] . This generalises a previous work of Serre [ 21 ] from 1990 where, using the large sieve as well, the author obtained a sharp upper-bound for the quantity N(B):=\#\left\{{\bm{t}}\in\left[-B,B\right]^{3}:t_{0}x_{0}^{2}+t_{1}x_{1}^{2}=t_{2}x_{2}^{2}\text{ has a solution $\bm{x}\in\mathbb{Q}^{3}\setminus\{(0,0,0)\}$}\right\}\quad(B\geqslant 2). Note that, here, studying the quantity N(B) is equivalent to studying the quantity N_{\mathrm{loc}}(\pi,B) in the case of the family described by X:\quad t_{0}x_{0}^{2}+t_{1}x_{1}^{2}=t_{2}x_{2}^{2}\quad\subseteq\mathbb{P}^{2}\times\mathbb{P}^{2}, with \pi:(\bm{x},\bm{t})\in X\mapsto\bm{t}\in\mathbb{P}^{2} . Between 1990 and 2000 matching lower bounds for N(B) were obtained independently by Guo [ 8 ] and Hooley [ 10 , 11 ] . In 2023, Loughran, Rome and Sofos promoted those bounds to an asymptotic by showing the estimate N(B)\underset{B\to+\infty}{\sim}c{B^{3}\over(\log B)^{3/2}} where c is an explicit positive constant. They also give a prediction for the shape of the constant c in general. We now recall some terminology in order to state conjecture [ 16 , 3.8] . For any weak Fano variety Y (see [ 16 , def 3.1] ), we denote by \Lambda_{\mathrm{eff}}(Y)^{\widecheck{}}\subseteq\mathrm{Pic}(Y)^{\widecheck{}} the dual of the pseudo-effective cone equipped with the Haar measure such that the dual lattice \mathrm{Pic}(Y)^{\widecheck{}} has covolume 1 . The purity exact sequence [ 3 , 3.7.2] yields 0\longrightarrow\brau(Y)\longrightarrow\brau(\mathbb{Q}(Y))\overset{\oplus\partial_{D}}{\longrightarrow}\bigoplus_{D\in Y^{(1)}}H^{1}(\mathbb{Q}(D),\mathbb{Q}/\mathbb{Z}) where \partial_{D}:\brau(\mathbb{Q}(Y))\longrightarrow H^{1}(\mathbb{Q}(D),\mathbb{Q}/\mathbb{Z}) is the residue map at D . If b is such that \delta_{D}(b)=0 , one says that b is unramified along D . Let X be a smooth and proper projective variety over a number field and \pi:X\to Y be a dominant morphism. The subordinate Brauer \brausub(Y,\pi) is defined by \brausub(Y,\pi):=\bigcap_{D\in Y^{(1)}}\left\{b\in\brau\mathbb{Q}(Y):\begin{tabular}[]{c}$\partial_{E}\;\pi^{\ast}b=0\text{ for all irreducible components }$\\ $E\subseteq\pi^{-1}(D)\text{ of multiplicity }1$\end{tabular}\right\}. Note that these two notions come from [ 16 ] . We also need the following definitions. Definition 1.1 . Let K be a number field or a finite field, let \overline{K} be its algebraic closure, and let X be a finite type scheme over K . The absolute Galois group \mathrm{Gal}(\overline{K}/K) acts on the geometric irreducible components of X , i.e. the irreducible components of X\times_{K}\overline{K} . We say that X is split (resp. pseudo-split) if \mathrm{Gal}(\overline{K}/K) (resp. every element of \mathrm{Gal}(\overline{K}/K) ) fixes some geometric irreducible component of multiplicity 1. Definition 1.2 . Let \pi:X\to Y be a dominant proper morphism of smooth varieties over \mathbb{Q} . For each point D of Y , we denote by \kappa(D) its residue field and we choose some finite group \Gamma_{D} through which the absolute Galois group \mathrm{Gal}(\overline{\kappa(D)}/\kappa(D)) acts on the geometric irreducible components of \pi^{-1}(D) . We define \delta_{D}(\pi)={\#\{\gamma\in\Gamma_{D}:\gamma\text{ fixes a geometric irreducible component of }\pi^{-1}(D)\text{ of multiplicity 1}\}\over\#\Gamma_{D}}. This quantity measures