Hostname: page-component-586b7cd67f-gb8f7 Total loading time: 0 Render date: 2024-11-25T22:04:44.970Z Has data issue: false hasContentIssue false

AN APPLICATION OF BINARY QUADRATIC FORMS OF DISCRIMINANT $\boldsymbol {-31}$ TO MODULAR FORMS

Published online by Cambridge University Press:  08 February 2022

ZAFER SELCUK AYGIN
Affiliation:
Centre for Research in Algebra and Number Theory, School of Mathematics and Statistics, Carleton University, Ottawa, Ontario K1S 5B6, Canada and University Studies Department, Northern Lakes College, Slave Lake, Alberta T0G2A3, Canada e-mail: [email protected]
KENNETH S. WILLIAMS*
Affiliation:
Centre for Research in Algebra and Number Theory, School of Mathematics and Statistics, Carleton University, Ottawa, Ontario K1S 5B6, Canada
*
Rights & Permissions [Opens in a new window]

Abstract

In this note, we use Dedekind’s eta function to prove a congruence relation between the number of representations by binary quadratic forms of discriminant $-31$ and Fourier coefficients of a weight $16$ cusp form. Our result is analogous to the classical result concerning Ramanujan’s tau function and binary quadratic forms of discriminant $-23$ .

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of Australian Mathematical Publishing Association Inc.

The Ramanujan tau function $\tau $ is defined by

$$ \begin{align*} \Delta(z):=q \prod_{n=1}^{\infty} (1-q^{n})^{24} = \sum_{n=1}^{\infty} \tau(n) q^{n}, \end{align*} $$

where $q:=e^{2 \pi i z}$ ( $z \in \mathbb {C}$ , Im $(z)>0$ ). In 1930, Wilton [Reference Wilton5] determined $\tau (n)$ modulo $23$ for all positive integers n. In 2006, Sun and Williams [Reference Sun and Williams3, Corollary 2.2, page 357] obtained Wilton’s congruence for $\tau (n)$ modulo $23$ as a consequence of their work on binary quadratic forms. Recently Dr. Pieter Moree of the Max Planck Institute for Mathematics in Bonn, Germany, in relation to his recent work [Reference Ciolan, Languasco and Moree1] with Ciolan and Languasco, asked the second author if the analogous congruence modulo $31$ could be obtained using the ideas of [Reference Sun and Williams3] for the function $\tau _{16}(n)$ , where

$$ \begin{align*} \Delta(z) E_{4}(z):=\sum_{n=1}^{\infty} \tau_{16}(n) q^{n} \end{align*} $$

and $E_{4}(q)$ is the Eisenstein series

$$ \begin{align*} E_{4}(z) := 1+ 240 \sum_{n=1}^{\infty} \sigma_{3}(n) q^{n}. \end{align*} $$

Swinnerton-Dyer [Reference Swinnerton-Dyer, Kuijk and Serre4, page 34], before giving the arguments that prove the congruence relation (3), notes that ‘there seems little prospect’ of proving this congruence using Dedekind’s eta function. In this note, we show that it can be done by giving an explicit proof of the congruence relation (3) using Dedekind’s eta function. Then we combine our results with [Reference Sun and Williams2, Theorem 10.2, page 166] to obtain the following congruence for $\tau _{16}(n)$ modulo $31$ .

Theorem 1. For any positive integer n,

$$ \begin{align*} \tau_{16}(n) \equiv \begin{cases} 0 \pmod{31} \hspace{0.5cm} \mbox{if there is a prime }p \mid n\mbox{ with }\scriptsize{\big(\tfrac{p}{31}\big) }=-1\\ \hspace{3cm} \mbox{and }\nu_{p}(n) \equiv 1\pmod{2}\mbox{, or }\scriptsize{\big(\tfrac{p}{31}\big) }=1\mbox{, }\\ \hspace{3cm} p= 2x^{2}+xy+4y^{2}\mbox{ and }\nu_{p}(n) \equiv 2 \pmod{3}\mbox{,}\\ \hspace{2cm} \mbox{ }\\ \displaystyle (-1)^{\mu} \hspace{-15pt}\prod_{\substack{p \mid n,\\ \scriptsize{(\frac{p}{31}) }=1,\\ p= x^{2}+xy+8y^{2}}}\hspace{-15pt} (1+ \nu_{p}(n)) \pmod{31} \hspace{0.5cm} \mbox{otherwise,} \end{cases} \end{align*} $$

where

$$ \begin{align*} \mu= \hspace{-20pt}\sum_{\substack{p \mid n,\\ \scriptsize{(\frac{p}{31}) }=1,\\ p= 2x^{2}+xy+4y^{2},\\ \nu_{p}(n) \equiv 1 \pmod{3}}}\hspace{-20pt} 1. \end{align*} $$

Proof. We use the Dedekind eta function which is defined by

$$ \begin{align*} \eta(z):=q^{1/24} \prod_{n=1}^{\infty} (1-q^{n}). \end{align*} $$

We have

$$ \begin{align*} \Delta(z) E_{4}(z) &= \frac{\eta^{32}(z) \eta^{32}(4z)}{\eta^{32}(2z)} + \frac{1}{2}\bigg( \frac{\eta^{64}(2z)}{\eta^{32}(4z)} - \frac{\eta^{64}(z)}{\eta^{32}(2z)} \bigg) \\[6pt] &\quad + 31 \bigg( 69271552 \frac{\eta^{64}(4z)}{\eta^{32}(2z)} -34095104 \frac{\eta^{48}(4z)}{\eta^{8}(z)\eta^{8}(2z)} + 7915008 \frac{\eta^{16}(2z)\eta^{32}(4z)}{\eta^{16}(z)} \\[6pt] &\quad -1050688 \frac{\eta^{40}(2z)\eta^{16}(4z)}{\eta^{24}(z)} + 82977 \frac{\eta^{64}(2z)}{\eta^{32}(z)} -3840 \frac{\eta^{88}(2z)}{\eta^{40}(z)\eta^{16}(4z)}\\[6pt] &\quad + 96 \frac{\eta^{112}(2z)}{\eta^{48}(z)\eta^{32}(4z)} - \frac{\eta^{136}(2z)}{\eta^{56}(z)\eta^{48}(4z)} \bigg). \end{align*} $$

All of the functions in this modular equation are in $M_{16}(\Gamma _{0}(4))$ and the identity can be proved using Sturm’s theorem. Thus, we have

(1) $$ \begin{align} \Delta(z) E_{4}(z) \equiv \frac{\eta^{32}(z) \eta^{32}(4z)}{\eta^{32}(2z)} + \frac{1}{2}\bigg( \frac{\eta^{64}(2z)}{\eta^{32}(4z)} - \frac{\eta^{64}(z)}{\eta^{32}(2z)} \bigg) \ \pmod{31}. \end{align} $$

If $f:=ax^{2}+bxy+cy^{2}$ is a positive-definite integral binary quadratic form, we denote by $r(f;n)$ the number of representations of a nonnegative integer n by f. We set

$$ \begin{align*} \phi_{2}(z)&:=\frac{1}{2}\sum_{n \geq 0} (r(x^{2}+xy+8 y^{2};n)-r(2x^{2}+xy+4 y^{2};n)) q^{n}\\[6pt] &\ = \frac{1}{2} \sum_{x,y=-\infty}^{\infty} (q^{x^{2}+xy+8 y^{2}} - q^{2x^{2}+xy+4 y^{2}}). \end{align*} $$

The theta functions $\sum _{x,y=-\infty }^{\infty } q^{x^{2}+xy+8 y^{2}}$ and $\sum _{x,y=-\infty }^{\infty } q^{2x^{2}+xy+4 y^{2}}$ belong to the space $M_{1}(\Gamma _{0}(124),\scriptsize {(\frac{-31}{*}) })$ as do the eta quotients

$$ \begin{align*} \frac{\eta(z) \eta(4z) \eta(31z)\eta(124z)}{\eta(2z)\eta(62z)}, \quad \frac{\eta^{2}(2z)\eta^{2}(62z)}{\eta(4z)\eta(124z)} \quad\mbox{and}\quad \frac{\eta^{2}(z)\eta^{2}(31z)}{\eta(2z)\eta(62z)}. \end{align*} $$

Then it is straightforward to prove the modular identity

$$ \begin{align*} \phi_{2}(z) = \frac{\eta(z) \eta(4z) \eta(31z)\eta(124z)}{\eta(2z)\eta(62z)} + \frac{1}{2}\bigg( \frac{\eta^{2}(2z)\eta^{2}(62z)}{\eta(4z)\eta(124z)} - \frac{\eta^{2}(z)\eta^{2}(31z)}{\eta(2z)\eta(62z)} \bigg) \end{align*} $$

using Sturm’s theorem. We have $1-A^{31} \equiv (1-A)^{31}$ (mod $31$ ) by the binomial theorem, so that

(2) $$ \begin{align} \phi_{2}(z) \equiv \frac{\eta^{32}(z) \eta^{32}(4z)}{\eta^{32}(2z)} + \frac{1}{2}\bigg( \frac{\eta^{64}(2z)}{\eta^{32}(4z)} - \frac{\eta^{64}(z)}{\eta^{32}(2z)} \bigg) \ \pmod{31}. \end{align} $$

From (1) and (2), we deduce that

(3) $$ \begin{align} \Delta(z) E_{4}(z) \equiv \phi_{2}(z) \ \pmod{31}. \end{align} $$

Appealing to the formula for $\tfrac 12(r(x^{2}+xy+8 y^{2};n)-r(2x^{2}+xy+4 y^{2};n))$ given in [Reference Sun and Williams2, Theorem 10.2, page 166], we obtain from (3) the congruence for $\tau _{16}(n)$ stated in the theorem.

References

Ciolan, A., Languasco, A. and Moree, P., ‘Landau and Ramanujan approximations for divisor sums and coefficients of cusp forms’, Preprint, 2021, arXiv:2109.03288, 43 pages.Google Scholar
Sun, Z.-H. and Williams, K. S., ‘On the number of representations of $n$ by $a{x}^2+bxy+c{y}^2$ ’, Acta Arith. 122(2) (2006), 101171.CrossRefGoogle Scholar
Sun, Z.-H. and Williams, K. S., ‘Ramanujan identities and Euler products for a type of Dirichlet series’, Acta Arith. 122(4) (2006), 349393.CrossRefGoogle Scholar
Swinnerton-Dyer, H. P. F., ‘On $\ell$ -adic representations and congruences for coefficients of modular forms’, in: Modular Functions of One Variable III (Proceedings International Summer School University of Antwerp, RUCA July 17–August 3, 1972), Lecture Notes in Mathematics, 350 (eds. Kuijk, W. and Serre, J.-P.) (Springer, Berlin, 1973), 155.Google Scholar
Wilton, J. R., ‘Congruence properties of Ramanujan’s function $\tau (n)$ ’, Proc. Lond. Math. Soc. (3) 31 (1930), 110.CrossRefGoogle Scholar