Quadratic Euler-Kronecker constants in positive characteristic

Amir Akbary, Félix Baril Boudreau · 2025

Read Quadratic Euler-Kronecker constants in positive characteristic by Amir Akbary, Félix Baril Boudreau (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-2503-00288.txt.

Introduction

Amir Akbary and Félix Baril Boudreau study how a set of constants, tied to prime-based functions, spread across a family of fields. Their paper, from two thousand twenty-five, sits in number theory. The constants grow out of a famed one for the whole counts. The team reads their values over squared growths of a base field, through a linked function's slope at one point. With a chance model, they show the values tend to a smooth spread as the fields grow.

Catalog record

CREATORAmir Akbary, Félix Baril Boudreau
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-00288

Browse the full library · About the Tower of Babel

Full text

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

Quadratic Euler-Kronecker constants in positive characteristic Thanks: Amir Akbary was supported by the NSERC discovery grant RGPIN-2021-02952. The project started when Félix Baril Boudreau was supported by a PIMS postdoctoral fellowship at the University of Lethbridge. This research was also funded in part by the Luxembourg National Research Fund (FNR) grant reference 501100001866-17921905. For the purpose of open access, and in fulfilment of the obligations arising from the grant agreement, the authors have applied a Creative Commons Attribution 4.0 International (CC BY 4.0) license to any Author Accepted Manuscript version arising from this submission. Amir Akbary, Félix Baril Boudreau Abstract . In 2006, Ihara defined and systematically studied a generalization of the Euler-Mascheroni constant for all global fields, named the Euler-Kronecker constants. This paper examines their distribution across geometric quadratic extensions of a rational global function field, via the values of logarithmic derivatives of Dirichlet L L -functions at 1. Using a probabilistic model, we show that the values converge to a limiting distribution with a smooth, positive density function, as the genii of quadratic fields approach infinity. We then prove a discrepancy theorem for the convergence of the frequency of these values, and obtain information about the proportion of the small values. Finally, we prove omega results on the extreme values. Our theorems imply new distribution results on the stable Taguchi heights and logarithmic Weil heights of rank 2 Drinfeld modules with CM. 2020 Mathematics Subject Classification 11R59, 11R58, 11G09 1. Introduction In [ 16 ] , Ihara defines the Euler-Kronecker constant of a global field F as the limit \gamma_{F}=\lim_{s\to 1}\left(\frac{\zeta^{\prime}(s,F)}{\zeta(s,F)}+\frac{1}{s-1}\right), where \zeta(s,F) is the Dedekind zeta function of F in the complex variable s (see also [ 15 ] *pp.496-497 for the number field case). If F is a global function field, i.e., F is the function field of a proper, smooth, and geometrically connected curve C defined over a finite field \mathbb{F}_{q} , it is then shown in [ 16 ] *(1.4.3)) that (1.1) \frac{\gamma_{F}}{\log{q}}=\sum_{m=1}^{\infty}\left(\frac{q^{m}+1-N_{m}}{q^{m}}\right)+\frac{q-3}{2(q-1)}, where N_{m} denotes the number of \mathbb{F}_{q^{m}} -rational points of C . This formula allows explicit computation of \gamma_{F} . For instance, when F=\mathbb{F}_{q}(t) , then N_{m}=q^{m}+1 for all m\geq 1 , hence \gamma_{q}:=\gamma_{F}=(q-3)\log(q)/(2(q-1)) . Now, consider a supersingular elliptic curve E/\mathbb{F}_{p} for a prime p\geq 5 with Weierstrass equation y^{2}=D for some monic D in \mathbb{F}_{p}[t] . For each integer m\geq 1 , N_{m} equals p^{m}+1 if m is odd, and \left(p^{\frac{m}{2}}-(-1)^{\frac{m}{2}}\right)^{2} if m is even (see [ 29 ] *Exercise 5.15). Therefore, if F is the function field of E/\mathbb{F}_{p} , then \gamma_{F}=\left(\sum_{m=1}^{\infty}\left(\frac{p^{m}+1-N_{m}}{p^{m}}\right)+1-\frac{p+1}{2(p-1)}\right)\log(p)=\frac{p-5}{2(p+1)}. More generally, Ihara observes from ( 1.1 ) that if N_{m} is large for small m (in particular for m=1 ), then \gamma_{F} tends to be negative (see also [ 16 ] *Theorem 4). Ihara puts forward \gamma_{F} as an invariant of F and investigates its cardinality. If F has genus g_{F} , then one has (see [ 16 ] *Theorem 1 and Proposition 3) (1.2) -\log\left(q^{g_{F}-1}\right)+O_{q}(1)\leq\gamma_{F}\leq 2\log\left(\log\left(q^{g_{F}-1}\right)\right)+O_{q}(1). Moreover, for an infinite family of global function fields F of unbounded genii g_{F} and fixed degree N\geq 1 over a given rational function field \mathbb{F}_{q}(t) , Ihara shows [ 16 ] *Corollary 2 (1.3) \gamma_{F}>-2(N-1+\varepsilon)\log\left(\log\left(q^{g_{F}-1}\right)\right), for any real number \varepsilon>0 . Let q be a power of an odd prime number. Let \mathbb{F}_{q}[t] be the ring of polynomials in the variable t with coefficients in \mathbb{F}_{q} . For each integer n\geq 1 , let \mathcal{H}_{n} be the subset of degree n square-free monic polynomials in \mathbb{F}_{q}[t] . In this paper, we study the magnitude of \gamma_{D} , the Euler-Kronecker constant of K_{D}:=\mathbb{F}_{q}(t)(\sqrt{D}) , the function field of the hyperelliptic curve of affine model y^{2}=D as D varies in the set \mathcal{H}_{n} . From the Hurwitz genus formula, the genus of K_{D} is (1.4) g_{D}=\begin{cases}\frac{n}{2}-1&\text{ if }n\text{ is even},\\ \frac{n-1}{2}&\text{ if }n\text{ is odd}.\end{cases} Now, let \chi_{D} be the quadratic Dirichlet character associated with a given D\in\mathcal{H}_{n} and let L(s,\chi_{D}) be the corresponding Dirichlet L -function. Its logarithmic derivative at s=1 satisfies the relation (1.5) \gamma_{D}=\frac{L^{\prime}(1,\chi_{D})}{L(1,\chi_{D})}+\widetilde{\gamma_{q}}, where \widetilde{\gamma_{q}}=\begin{cases}\gamma_{q}&\text{if }n\text{ is odd},\\ \gamma_{q}+\zeta_{q}(2)&\text{if }n\text{ is even},\\ \end{cases} with \zeta_{q}(2):=\zeta(2,\mathbb{F}_{q}[t])=q/(q-1) . (See Subsection 2.2 for a proof of ( 1.5 ).) As we see from ( 1.5 ), the magnitudes of \gamma_{D} and L^{\prime}(1,\chi_{D})/L(1,\chi_{D}) are intimately related. From [ 17 ] *(6.8.20), we have (1.6) \left|\frac{L^{\prime}(1,\chi_{D})}{L(1,\chi_{D})}\right|\ll_{q}\log\left(\log{q^{n}}\right), and so \left|\gamma_{D}\right|\ll_{q}\log\left(\log{q^{n}}\right) . In this paper, we show that for a proportion of polynomials D in \mathcal{H}_{n} the values of |{L^{\prime}(1,\chi_{D})}/{L(1,\chi_{D})}| are significantly smaller than the upper bound in ( 1.6 ). Theorem 1.1 . For n\geq 1 , we have \min\limits_{D\in\mathcal{H}_{n}}\left(\left|\frac{L^{\prime}(1,\chi_{D})}{L(1,\chi_{D})}\right|\right)\ll_{q}\frac{\log^{2}(\log(q^{n}))}{\log(q^{n})}. More precisely, there are \gg_{q}q^{n}\log^{2}(\log(q^{n}))/\log(q^{n}) polynomials D\in\mathcal{H}_{n} for which \left|\gamma_{D}-\widetilde{\gamma}_{q}\right|=\left|\frac{L^{\prime}(1,\chi_{D})}{L(1,\chi_{D})}\right|\ll_{q}\frac{\log^{2}(\log(q^{n}))}{\log(q^{n})}. We also obtain unconditional omega results which imply that Ihara’s upper bound ( 1.2 ) and lower bound ( 1.3 ) for \gamma_{D} cannot be improved to \mid\gamma_{D}\mid\leq\log(\log(q^{g_{D}}))+O_{q}(1) . Theorem 1.2 . Let \mathcal{P}_{n} be the set of irreducible elements in \mathcal{H}_{n} . For any \varepsilon>0 and all large n , there are \gg q^{\frac{n}{2}} elements Q\in\mathcal{P}_{n} such that \frac{L^{\prime}(1,\chi_{Q})}{L(1,\chi_{Q})}\geq\log(\log(q^{n}))+\log(\log(\log(q^{n})))-A_{q}-\varepsilon, and \gg q^{\frac{n}{2}} elements Q\in\mathcal{P}_{n} such that \frac{L^{\prime}(1,\chi_{Q})}{L(1,\chi_{Q})}\leq-\log(\log(q^{n}))-\log(\log(\log(q^{n})))+A_{q}+\varepsilon, where A_{q}:=(2.61)\log(q)+\log(2\log(2)\zeta_{q}(2))>0 . Bounds for \gamma_{D} are obtained by replacing L^{\prime}(1,\chi_{Q})/L(1,\chi_{Q}) with \gamma_{D} and A_{q} with A_{q}+\tilde{\gamma} . We prove Theorem 1.1 following the number field method in [ 1 ] , which drew ideas from [ 21 ] , [ 20 ] , [ 19 ] and [ 14 ] . More precisely, we establish an asymptotic formula for the r -th integral moments of -L’(1,\chi_{D})/L(1,\chi_{D}) (Proposition 3.3 ), uniform on a certain range for r , and use it to compute the Laplace transform of -L’(1,\chi_{D})/L(1,\chi_{D}) asymptotically (Proposition 3.5 ). We then introduce a random series -\frac{L^{\prime}(1,\mathbb{X})}{L(1,\mathbb{X})} as the global function field analogue of the model in [ 13 ] for integers, and compare its Laplace transform to that of -L^{\prime}(1,\chi_{D})/L(1,\chi_{D}) (Proposition 5.2 ). This yields, together with an application of Berry-Esseen inequality, a discrepancy estimate between the characteristic function of the series -\frac{L^{\prime}(1,\mathbb{X})}{L(1,\mathbb{X})} and the sequence of characteristic functions of some arithmetic functions (Proposition 7.1 ). The proof of Theorem 1.1 results from the positivity of the density function associated with the distribution of the random series (Proposition 7.3 ) following an argument analogous to [ 19 ] *p.368 (see also [ 1 ] *Proof of Theorem 1.5 and [ 14 ] *Corollary 1.4). Proposition 3.2, a crucial result in the proof of Theorem 1.1, provides an approximation for the value -L^{\prime}(s,\chi)/L(s,\chi) for a non-trivial Dirichlet character \chi and \re(s)>1/2 in terms of a short Dirichlet polynomial. Although we only need such a result for s=1 and a quadratic Dirichlet character, we prove the general statement as a worthwhile addition to the literature. The proof of this key assertion substantially differs from its number field analogue. We exploit the fact that L(s,\chi) is a polynomial in q^{-s} with roots \alpha_{i}(\chi) satisfying |\alpha_{i}(\chi)|\leq q^{1/2} by the proven Riemann Hypothesis for global function fields. Consequently, we establish an optimal error term in Proposition 3.2, which gives better error terms and more flexible range of moments for global function field counterparts of some of the assertions of [ 21 ] (see Propositions 3.3 , 3.5 , 5.2 ). We prove Theorem 1.2 using Granville–Soundararajan’s strategy [ 13 ] *Proposition 9.1 and Theorem 5a.

Read the complete text