Local-global principles for semi-integral points on Markoff orbifold pairs

Vladimir Mitankin, Justin Uhlemann · 2024

Read Local-global principles for semi-integral points on Markoff orbifold pairs by Vladimir Mitankin, Justin Uhlemann (2024) 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-2411-02629.txt.

Description

A 2024 arXiv research paper by Vladimir Mitankin, Justin Uhlemann. We study local-global principles for semi-integral points on orbifold pairs of Markoff type. In particular, we analyse when these orbifold pairs satisfy weak weak approximation, weak approximation and strong approximation off a finite set of places. We show that Markoff orbifold pairs satisfy the semi-integral Hasse principle and we measure how often such orbifold pairs have strict semi-integral points but the corresponding Markoff surface lacks integral points. Introduction

Let m\neq 0,4 be an integer.

Catalog record

CREATORVladimir Mitankin, Justin Uhlemann
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2411-02629

Browse the full library · About the Tower of Babel

Full text

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

Local-global principles for semi-integral points on Markoff orbifold pairs Thanks: 2020 Mathematics Subject Classification 14G05, 14G12, 11D25 (primary), 11G35 (secondary). Vladimir Mitankin, Justin Uhlemann Abstract . We study local-global principles for semi-integral points on orbifold pairs of Markoff type. In particular, we analyse when these orbifold pairs satisfy weak weak approximation, weak approximation and strong approximation off a finite set of places. We show that Markoff orbifold pairs satisfy the semi-integral Hasse principle and we measure how often such orbifold pairs have strict semi-integral points but the corresponding Markoff surface lacks integral points. 1. Introduction Let m\neq 0,4 be an integer. A smooth affine Markoff surface U_{m} over \mathbb{Q} and its \mathbb{Z} -model \mathcal{U}_{m} , given by the same equation, are defined by (1.1) U_{m}:\quad u_{1}^{2}+u_{2}^{2}+u_{3}^{2}-u_{1}u_{2}u_{3}=m. Ghosh and Sarnak’s pioneering work [ 24 ] conjectured the amount of m with \mathcal{U}_{m}(\mathbb{Z})=\emptyset but no local obstructions to that, which sparked a lot of interest in the number theory community. Soon after, two independent studies [ 33 ] and [ 17 ] examined that conjecture together with the density of integral points on \mathcal{U}_{m} . The behaviour of integral points on singular and generalised versions of Markoff surfaces have been investigated in [ 56 ] , [ 51 ] , [ 2 ] , [ 15 ] , [ 8 ] , [ 21 ] , [ 30 ] , [ 22 ] , [ 38 ] . This work pursues a different direction by exploring for the first time Campana and Darmon points, collectively abbreviated semi-integral, on orbifold pairs defined by ( 1.1 ). Semi-integrality (Definition 2.7 ) is a concept generalising integral and rational points and at the same time interpolating between them. We conduct a comprehensive study of the semi-integral local-global principles, as introduced in [ 39 , §2] , for Markoff orbifold pairs and compare their status with their rational and integral counterparts. This introduces the investigation of pairs with multiple irreducible components of the boundary. Local-global principles To put things in perspective, consider the smooth compactification X_{m}\subset\mathbb{P}_{\mathbb{Q}}^{3} of U_{m} , with a \mathbb{Z} -model \mathcal{X}_{m} defined by the same equation, given by (1.2) X_{m}:\quad x_{0}(x_{1}^{2}+x_{2}^{2}+x_{3}^{2})-x_{1}x_{2}x_{3}=mx_{0}^{3}. The existence of a rational point (0:0:1:1)\in X_{m}(\mathbb{Q}) implies that X_{m} is unirational [ 31 ] and that it satisfies weak weak approximation [ 52 ] . In fact, X_{m} is rational, and thus satisfies weak approximation, if and only if m-4\in\mathbb{Q}(\sqrt{m})^{*2} by [ 33 , Lem. 3.3.] , which is equivalent to m-4\in\mathbb{Q}^{*2} (see Remark 3.10 ). The reverse implication also holds. Theorem 1.1 . The variety X_{m} satisfies weak approximation for \mathbb{Q} -rational points if and only if X_{m} is rational. To set-up the notion of semi-integrality, let D=X_{m}\setminus U_{m} . Fixing weights \omega_{i}\in\mathbb{Z}_{\geq 2}\cup\{\infty\} for the irreducible components D_{i}=\{x_{0}=x_{i}=0\} , i=1,2,3 of D defines a \mathbb{Q} -divisor D_{\uomega}=\sum_{i=1}^{3}(1-1/\omega_{i})D_{i} that gives rise to a Campana orbifold structure (X_{m},D_{\uomega}) . If \mathcal{D}_{i} is the Zariski closure of D_{i} in \mathcal{X}_{m} for i=1,2,3 and \mathcal{D}_{\uomega}=\sum_{i=1}^{3}(1-1/\omega_{i})\mathcal{D}_{i} , the \mathbb{Z} -model (\mathcal{X}_{m},\mathcal{D}_{\uomega}) of (X_{m},D_{\uomega}) carries the data of semi-integral points on the pair (Definition 2.7 ). Selecting \omega_{1}=\omega_{2}=\omega_{3}=\infty or an empty boundary divisor recovers integral points on \mathcal{U}_{m} and rational points on X_{m} , respectively. Our results go beyond the classical setting and consider the case where at least one \omega_{i}<\infty . The discrepancy in the behaviour of semi-integral and rational points is observed in the next result. Theorem 1.2 . The following hold. (i) Assume that at least two of \omega_{1},\omega_{2},\omega_{3} are finite. Then each pair (\mathcal{X}_{m},\mathcal{D}_{\uomega}) fails weak weak approximation for semi-integral points. (ii) Assume that exactly one of \omega_{1},\omega_{2},\omega_{3} is finite. If m-4\in\mathbb{Q}^{*2} , then (\mathcal{X}_{m},\mathcal{D}_{\uomega}) satisfies weak approximation. On the other hand, if m-4\notin\mathbb{Q}^{*2} , then (a) (\mathcal{X}_{m},\mathcal{D}_{\uomega}) fails Campana weak approximation; (b) (\mathcal{X}_{m},\mathcal{D}_{\uomega}) fails Darmon weak approximation if the finite weight is even. Theorem 1.2 (i) follows from a result about weak approximation for general orbifold pairs, namely Theorem 2.15 . To the best of our knowledge this is the first result on local-global principles for Campana orbifolds with multiple irreducible components of the boundary. Remark 1.3 . Currently, there is no uniform notion for the semi-integral adelic space across the literature. We follow the convention of [ 39 ] . The proofs of Theorems 1.2 and 2.15 do not adapt to the other notions of semi-integral adeles, e.g the one given in [ 46 ] . Theorem 1.2 implies that, because of the boundary, strong approximation off any finite set fails if at least two of \omega_{1},\omega_{2},\omega_{3} are finite. The next result exhibits this failure through semi-integral points on U_{m} in the case of an arbitrary number of finite weights. Theorem 1.4 . Let S(m)=\{p\in\Omega_{\mathbb{Q}}\ :\ m-4\in\mathbb{Q}_{p}^{*2}\} . If m-4\notin\mathbb{Q}^{*2} and some \omega_{i} , i=1,2,3 is finite, then (\mathcal{X}_{m},\mathcal{D}_{\uomega}) fails strong approximation off any finite set T\subset S(m) . Remark 1.5 . Assume that not all of \omega_{1},\omega_{2},\omega_{3} are infinite. Then the family of orbifold pairs (\mathcal{X}_{m},\mathcal{D}_{\uomega}) satisfies the semi-integral Hasse principle as (0:0:1:1)\in X_{m}(\mathbb{Q}) is semi-integral on (\mathcal{X}_{m},\mathcal{D}_{\uomega}) by Proposition 3.1 . Thus for any choice of weights neither Campana nor Darmon points on (\mathcal{X}_{m},\mathcal{D}_{\uomega}) behave in the way that \mathcal{U}_{m}(\mathbb{Z}) does, as \mathcal{U}_{m}(\mathbb{Z}) is infinitely often empty without local obstruction to that as shown in [ 24 ] , [ 33 ] and [ 17 ] . Strict points Integral points are better approximated by strict semi-integral points (\mathcal{X}_{m},\mathcal{D}_{\uomega})_{\str}^{*}(\mathbb{Z})=(\mathcal{X}_{m},\mathcal{D}_{\uomega})^{*}(\mathbb{Z})\cap U_{m}(\mathbb{Q}) (Definitions 2.7 and 2.11 ), where \ast=\cc to indicate Campana points or \ast=\dd for Darmon points. For example (\mathcal{X}_{m},\mathcal{D}_{\uomega})_{\str}^{*}(\mathbb{Z}) detects failures of strong approximation featuring in the integral case [ 33 , Thm. 1.1] , e.g. Theorem 1.4 . A powerful tool to study local-global principles is the Brauer–Manin obstruction. Its semi-integral version has been defined in [ 39 , §3] along with its relation to its rational and integral counterparts. As is evident from Remark 1.5 such obstructions to the Hasse principle are not present in the family (\mathcal{X}_{m},\mathcal{D}_{\uomega}) as its non-strict part vanishes. We perform a complete analysis of the values of the local invariant maps for the relevant Brauer elements in § 3 to examine this obstruction for strict semi-integral points. Theorem 1.6 . Assume that \omega_{i}<\infty for some i=1,2,3 . Then no algebraic Brauer–Manin obstruction to the Hasse principle for strict points on (\mathcal{X}_{m},\mathcal{D}_{\uomega}) is present. Moreover, if ( \ast ) -(m-4)\in\mathbb{Q}^{2}\implies\frac{\sqrt{m}+\sqrt{m-4}}{2}\notin\mathbb{Q}(\sqrt{m},\sqrt{m-4})^{2} is met, then there is no Brauer–Manin obstruction to the Hasse principle for strict semi-integral points on (\mathcal{X}_{m},\mathcal{D}_{\uomega}) . The condition ( ∗ ‣ 1.6 ) is not very restrictive on m , in fact \#\{|m|\leq B\ :\ \text{\eqref{eqn:star} fails}\}=O(B^{1/4}) by [ 33 , Prop 5.2] . The lack of a Brauer–Manin obstruction does not necessarily imply the existence of strict semi-integral points. By [ 24 , Thm 1.2(ii) and Prop. 6.1] asymptotically 7/12 of the total number of Markoff orbifold pairs trivially have a strict semi-integral point, as \mathcal{U}_{m}(\mathbb{Z})\neq\emptyset . For the remaining 5/12 values of m we have \mathcal{U}_{m}(\mathbb{Z})=\emptyset because no 2 -adic or 3 -adic integral points exist. However, it may still be the case that there are global semi-integral points, as explained in the paragraph after Question 1.7 in [ 39 ] . It is natural to focus on those remaining m with empty integral adeles.

Read the complete text