On mean value formulas for the approximate functional equation of the Riemann zeta-function (Q1614937)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: On mean value formulas for the approximate functional equation of the Riemann zeta-function |
scientific article; zbMATH DE number 1798889
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | On mean value formulas for the approximate functional equation of the Riemann zeta-function |
scientific article; zbMATH DE number 1798889 |
Statements
On mean value formulas for the approximate functional equation of the Riemann zeta-function (English)
0 references
10 September 2002
0 references
Let \(s = \sigma + it, 0 \leq \sigma \leq 1, t \geq 1\), and \[ \zeta(s) = \sum_{n\leq\sqrt{t/(2\pi)}}n^{-s} + \chi(s)\sum_{n\leq\sqrt{t/(2\pi)}}n^{s-1} + R_1(s), \] so that \(R_1(s)\) denotes the error term in the approximate functional equation. As usual, \(\chi(s)\) is the function appearing in the functional equation for \(\zeta(s)\), namely \[ \chi(s) = {\zeta(s)\over\zeta(1-s)} = 2^s\pi^{s-1}\sin({\textstyle {1\over 2}}\pi s)\Gamma(1-s). \] The authors provide asymptotic formulas for the \(2k\)th moments of \(|R_1(s)|\), where \(k\) is a fixed natural number. Namely for \(0\leq \sigma \leq 1/(2k)\) they obtain \[ \int_1^T|R_1(s)|^{2k} dt = {(2\pi)^{k\sigma}C_k\over 1-k\sigma} T^{1-k\sigma} + Y_{k,\sigma}(T), \] where \[ C_k = \int_0^1{\left({\cos(2\pi(y^2-y-{1\over 16}))\over\cos(2\pi y)} \right)}^{2k} dy,\quad Y_{k,\sigma} = O(T^{{1\over 2}-k\sigma}). \] A formula valid for the range \(1/(2k) \leq \sigma \leq 1\) is also obtained. The main tool in the proof is the classical Riemann-Siegel formula for \(R_1(s)\). The most interesting ensuing integral is \[ \int_{T_1}^{T_2}\left({t\over 2\pi}\right)^{-k\sigma} {\left({\cos(2\pi(\delta^2-\delta-{1\over 16}))\over\cos(2\pi \delta)} \right)}^{2k}\text{ d}t,\quad T_1 \asymp T_2,\quad\delta = \sqrt{t\over 2\pi} - \left[\sqrt{t\over 2\pi}\right], \] where \([x]\) is the integer part of \(x\). After a change of variables \(\sqrt{t\over 2\pi} = y\) the authors make use of the fact that the new integrand is a periodic function of \(y\) with period 1. This enables them to obtain a sum of \((y + n)^{1-2k\sigma}\) over a suitable \(n\)-interval, which can be evaluated by the Euler-Maclaurin summation formula, and thus complete the proof.
0 references
Riemann zeta-function
0 references
approximate functional equation
0 references
Riemann-Siegel formula
0 references