On the function S ( t )

From MaRDI portal
Publication:4368464





Let, as usual, \(N(\sigma,T)\) denote the number of zeros \(\rho = \beta + i\gamma\) of the Riemann zeta-function \(\zeta(s)\) such that \(\beta \geq \sigma, 0 < \gamma \leq T\), and \[ S(T) = {1\over\pi}\arg\zeta({\textstyle{1\over 2}} + iT), \qquad S_1(T) = \int_0^T S(t) dt. \] These functions play an important rôle in zeta-function theory, and fundamental results on \(N(\sigma,T)\), \(S(T)\), \(S_1(T)\) were obtained long ago by \textit{A. Selberg} [Skr. Norske Vid. Akad., Oslo 10 (1942; Zbl 0028.11101) and Arch. Math. Naturvid. 48, 89-155 (1946; Zbl 0061.08402)]. In particular Selberg proved the asymptotic formula \[ \int_T^{T+H}| S(t)| ^{2k} dt = {(2k)!\over k!(2\pi)^{2k}}H(\log\log T)^k + O_k(H(\log\log T)^{k-{1\over 2}}),\tag{1} \] where \(k\) is a natural number and \(T^a \leq H \leq T, {1\over 2} < a \leq 1\). The author sharpens (1) and other corresponding results of Selberg by showing that one can take a substantially smaller \(H\), namely \[ H= T^{{27\over 82}+\varepsilon}\qquad \biggl( 0<\varepsilon< {1\over 1000}\biggr),\tag{2} \] and the \(O\)-constant in question will additionally depend on \(\varepsilon\). The basic method of proof is Selberg's, but it is combined with the author's method of dealing with the number of zeros of \(\zeta(s)\) on the critical line in short intervals [\textit{A. Karatsuba}, Izv. Akad. Nauk SSSR, Ser. Mat. 48, 569-584 (1984; Zbl 0545.10026)]. The author displays considerable skill in dealing with various technical problems. As in his earlier work (op. cit.), the constant ``27/82 in (2) could be somewhat decreased by more careful estimation of the relevant exponential sums by the theory of exponent pairs.











This page was built for publication: On the function S ( t )

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4368464)