Quantitative strong approximation for quaternary quadratic forms
Zhizhong Huang and Damaris Schindler and Alec Shute · 2025
Read Quantitative strong approximation for quaternary quadratic forms by Zhizhong Huang and Damaris Schindler and Alec Shute (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-2501-16766.txt.
Introduction
Zhizhong Huang and co-authors count rational points on quadratic surfaces with precise error control, a question in number theory. Such counting tests how evenly points spread over a shape. Strong approximation asks whether local solutions glue globally.
The team proves formulas for any smooth non-split surface. The error term at the real place is optimal. A second thread counts whole-number points on a punctured cone. That count quantifies strong approximation with an obstruction. The Brauer-Manin idea explains when gluing can fail. Results serve arithmetic geometry of quadrics. They sharpen tools used across Diophantine problems. The paper joins two counting stories in one for the field.
Catalog record
| CREATOR | Zhizhong Huang and Damaris Schindler and Alec Shute |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 96 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-16766
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantitative strong approximation for quaternary quadratic forms Zhizhong Huang and Damaris Schindler and Alec Shute Abstract . The purpose of this article is twofold. On the one hand, we prove asymptotic formulas for the quantitative distribution of rational points on any smooth non-split projective quadratic surface. We obtain the optimal error term for the real place. On the other hand, we also study the growth of integral points on the three-dimensional punctured affine cone, as a quantitative version of strong approximation with Brauer–Manin obstruction for this quasi-affine variety. 1. Introduction Let V\subseteq{\mathbb{P}}_{\mathbb{Q}}^{n} be a smooth projective quadric defined by a non-degenerate integral quadratic form F in (n+1) -variables. We always assume that V has local points at all places of \mathbb{Q} . Let W\subseteq{\mathbb{A}}^{n+1} be its corresponding affine cone. Let {\mathscr{V}}\subseteq{\mathbb{P}}_{{\mathbb{Z}}}^{n},{\mathscr{W}}\subseteq{\mathbb{A}}_{{\mathbb{Z}}}^{n+1} be respectively their integral models defined by F . In [ 10 ] , assuming n\geqslant 4 , on employing the delta method in a form developed by Heath-Brown [ 9 ] , we prove asymptotic formulas for the distribution of rational points on V , with optimal error terms. This is done via counting integral points of {\mathscr{W}} inside of an arbitrary adelic neighbourhood, with additional assumptions regarding the real condition. From this we deduce the best growth rate of the “size” of local conditions for which equidistribution is preserved. Adopting the approach from [ 10 ] , the goal of the present concurrent article is to focus on the non-split quaternary case, i.e. we assume that n=3 and \Delta_{F} , the discriminant of F , is not a square. In this case the delta method has already been applied by Lindqvist [ 11 ] in the context of strong approximation for the affine cone and she found additional main terms which did not have any clear interpretation yet. We provide a generalisation of her work to quantitative strong approximation, we compute the additional appearing terms explicitly and we show how they are related to the Brauer-Manin obstruction. This is the first instance that we are aware of in which the circle method detects the Brauer-Manin obstruction in a quantitative way with an explicit leading constant. 1.1. Effective equidistribution on the non-split projective quadratic surface Let us fix a parameter space S and a family of smooth weight functions \mathscr{C}(S) as in [ 9 ] . As in [ 10 , (1.1)] , we define a real zoom condition in terms of a fixed real point \boldsymbol{\xi}\in V({\mathbb{R}}) , a neighborhood U_{0} of \boldsymbol{\xi} , and a scaling parameter R\geqslant 1 . For B\gg 1 , we choose w_{B,R}:{\mathbb{R}}^{4}\to{\mathbb{R}} (see ( 2.7 )), constructed from a certain fixed w\in\mathscr{C}(S) as in [ 10 ] , which describes the bounded height condition depending on B and the real zoom condition. It satisfies (1.1) w_{B,R}(-\boldsymbol{x})=w_{B,R}(\boldsymbol{x}). See § 2.1 for a sketch and [ 10 , §2] for more details. By abuse of notation, for P\in{\mathbb{P}}^{3}({\mathbb{Q}}) , w_{B,R}(P) means that we choose a lift of P into a primitive integer vector of {\mathbb{Z}}^{4} and apply w_{B,R} . Our first result is on equidistribution with only a real zoom condition, and provides an optimal error term. We consider the counting function (1.2) {\mathscr{N}}_{V}(w_{B,R}):=\sum_{\begin{subarray}{c}P\in V({\mathbb{Q}})\end{subarray}}w_{B,R}(P). Theorem 1.1 . Assume ( 1.1 ), \Delta_{F}\neq\square , and (1.3) R^{2}\asymp B^{1-\tau}\quad\text{for some }0<\tau<1. Assume that V has local points at all places of \mathbb{Q} . Then {\mathscr{N}}_{V}(w_{B,R})=B^{2}\frac{{\mathscr{I}}(w_{R})}{2}\mathfrak{S}(V)\left(1+O_{\varepsilon}\left(B^{-\tau+\varepsilon}\right)\right), where {\mathscr{I}}(w_{R}) is the “weighted singular integral” (see ( 2.24 ) for its definition) and \mathfrak{S}(V) is the finite Tamagawa measure of {\mathscr{V}}(\widehat{{\mathbb{Z}}}) (à la Peyre [ 12 ] , with respect to the naive metric). For L\in{\mathbb{N}} , we let \boldsymbol{\Lambda}=(\boldsymbol{\Lambda}_{p})_{p\mid L} , where \boldsymbol{\Lambda}_{p}\in{\mathscr{V}}({\mathbb{Z}}_{p}) for every p\mid L . The pair (L,\boldsymbol{\Lambda}) defines a finite adelic open subset {\mathscr{D}}(L,\boldsymbol{\Lambda})\subset{\mathscr{V}}(\widehat{{\mathbb{Z}}}):=\prod_{p}{\mathscr{V}}({\mathbb{Z}}_{p}) (see [ 10 ] ), and gives rise to a p -adic zoom condition for each p\mid L . We now describe our second result of counting with only p -adic zoom conditions (i.e. we take R=1 ). We shall write w_{B}=w_{B,1} and consider (1.4) {\mathscr{N}}_{V}(w_{B};(L,\boldsymbol{\Lambda})):=\sum_{\begin{subarray}{c}P\in V({\mathbb{Q}})\\ P\in{\mathscr{D}}(L,\boldsymbol{\Lambda})\end{subarray}}w_{B}(P). Theorem 1.2 . Assume ( 1.1 ), \Delta_{F}\neq\square , and (1.5) L^{2}\asymp B^{1-\tau}\text{ for some }\frac{1}{2}<\tau<1. Assume that V has local points at all places of \mathbb{Q} . Then we have {\mathscr{N}}_{V}(w_{B};(L,\boldsymbol{\Lambda}))=B^{2}\frac{{\mathscr{I}}(w)}{2}\mathfrak{S}_{L,\boldsymbol{\Lambda}}({\mathscr{V}})\left(1+O_{\varepsilon}\left(B^{-2\tau+1+\varepsilon}\right)\right), where \mathfrak{S}_{L,\boldsymbol{\Lambda}}({\mathscr{V}}) is the finite Tamagawa measure of {\mathscr{V}}(\widehat{{\mathbb{Z}}}) regarding the local condition (L,\boldsymbol{\Lambda}) . Remark 1.3 . Theorems 1.1 and 1.2 have the following consequence on the “ essential approximation constant ” \alpha_{\operatorname{ess}}^{\nu} (see [ 10 , Appendix A] for more details): For any real point \boldsymbol{\xi}\in V({\mathbb{R}}) , we have \alpha_{\operatorname{ess}}^{{\mathbb{R}}}(\boldsymbol{\xi})\leqslant 2; For any non-archimedean place \nu and any \nu -adic point \xi_{\nu}\in V({\mathbb{Q}}_{\nu}) , we have \alpha_{\operatorname{ess}}^{\nu}(\xi_{\nu})\leqslant 4. Note that in [ 10 , Appendix A] we show that \alpha_{\operatorname{ess}}^{\nu}(\xi)=2 when \xi\in V(\mathbb{Q}) . So Theorem 1.1 is optimal if \boldsymbol{\xi} is a rational point. It would be very interesting to determine the exact value of such an essential constant for non-archimedean places. This is related to the best error term one can obtain in Theorem 1.2 . 1.2. Counting integral points on the punctured affine cone The second goal of this article is to study the asymptotic growth of integral points on the punctured affine cone {\mathscr{W}}^{o}:={\mathscr{W}}\setminus\overline{\boldsymbol{0}}\subseteq{\mathbb{A}}^{4}_{{\mathbb{Z}}} . In stark contrast to the higher dimensional cases, under the assumption that \Delta_{F}\neq\square , there can be a Brauer–Manin obstruction to strong approximation on {\mathscr{W}}^{o} . On {\mathscr{W}}^{o}(\widehat{{\mathbb{Z}}}):=\prod_{p}{\mathscr{W}}^{o}({\mathbb{Z}}_{p}) , we can associate a (normalised) Tamagawa measure, denoted by \omega_{f} locally given by the “modified singular series” (see ( 5.3 ) and see also Appendix A). Let W_{i},i\in I be the connected components of the real locus W^{o}({\mathbb{R}}) . If \Delta<0 (resp. \Delta>0 ) we then have \#I=2 (resp. \#I=1 ). (See Lemma 5.4 .) We shall establish in § 5 the following result. Theorem 1.4 . Assume \Delta_{F}\neq\square . Then for every i\in I , there exists a locally constant function \Xi_{W_{i}}:{\mathscr{W}}^{o}(\widehat{{\mathbb{Z}}})\to\{0,2\} such that for any non-empty open compact {\mathscr{E}}\subseteq{\mathscr{W}}^{o}(\widehat{{\mathbb{Z}}}) , if the support of w_{B} is contained in W_{i} , we have as B\to\infty , \sum_{\boldsymbol{x}\in{\mathscr{W}}^{o}({\mathbb{Z}})\cap(W_{i}\times{\mathscr{E}})}w_{B}(\boldsymbol{x})\sim{\mathscr{I}}(w)\left(\int_{{\mathscr{E}}}\Xi_{W_{i}}\operatorname{d}\omega_{f}\right)B^{2}. Theorem 1.4 implies a quantitative version of the Brauer–Manin obstruction to strong approximation of integral points on {\mathscr{W}}^{o} . According to the authors’ knowledge, this is the first instance where the circle method furnishes asymptotic formulas that encode information from the Brauer–Manin obstruction. Note that Theorem 1.4 also encodes a failure of strong approximation due to a Brauer-Manin obstruction in cases where the locally constant function \Xi_{W_{i}} is equal to zero. This shows that an analogue of Corollary 3 in [ 9 ] for primitive integer vectors in a strict sense is not true due to the possible existence of a Brauer-Manin obstruction (compare also with the paragraph after Corollary 3 in [ 9 ] where it is claimed that a generalisation would be possible.) Remarks 1.5 . (1) In the terminology of Borovoi–Rudnick [ 5 ] , up to the weighted singular integral, (an extension of) Theorem 1.4 shows that (see also [ 6 , Proposition 2.4] ): The quasi-affine variety W^{o} is relatively Hardy–Littlewood ( [ 5 , p. 39] ) with density function \Xi_{W_{i}} for each connected component W_{i} . (2) Applying results on the geometric sieve for quadrics ( [ 4 , Theorem 1.3] or