Explicit formulas for the pair correlation of vertical shifts of zeros of the Riemann zeta-function (Q2910123)

From MaRDI portal





scientific article; zbMATH DE number 6079049
Language Label Description Also known as
English
Explicit formulas for the pair correlation of vertical shifts of zeros of the Riemann zeta-function
scientific article; zbMATH DE number 6079049

    Statements

    0 references
    0 references
    7 September 2012
    0 references
    explicit formula
    0 references
    pair correlation
    0 references
    the Riemann zeta-function
    0 references
    Hardy--Littlewood circle method
    0 references
    Explicit formulas for the pair correlation of vertical shifts of zeros of the Riemann zeta-function (English)
    0 references
    The authors obtain three theorems involving the pair correlation of the zeros of the function NEWLINE\[NEWLINE H_\lambda(s) := \zeta(s-\tfrac{1}{2}i\lambda)\zeta(s+\tfrac{1}{2}i\lambda), NEWLINE\]NEWLINE where \(\zeta(s)\) is the Riemann zeta-function, and \(\lambda>0\) is fixed number. The first of these that, for \(2\leq x \leq T\), \(c>0\) a suitable constant, NEWLINE\[NEWLINE \sum_{H_\lambda(\rho) =H_\lambda(\rho')=0,|\Im \rho, \Im\rho'|\leq T}{\frac{x^{\rho+\rho'}}{\rho+\rho'}} = {\frac{2xT}{\pi}}\Bigl\{\Bigl[1 + \Re\Bigl({\frac{x^{i\lambda}}{1+i\lambda}}\Bigr)\Bigr]\log x - \Re\Bigl({\frac{x^{i\lambda}}{(1+i\lambda)^2}}\Bigr)-1\Bigr\}NEWLINE\]NEWLINE NEWLINE\[NEWLINE+ O_\lambda\Bigl(xT\exp(-c(\log x)^{3/5}(\log\log x)^{-1/5})\Bigr)NEWLINE\]NEWLINE NEWLINE\[NEWLINE+ O\left(x^2(\log T)^4\right) + O\left(T(\log T)^3\right). NEWLINE\]NEWLINE Here the error term with \(\ldots(\log x)^{3/5}\ldots\) clearly comes from the strongest known error term in the prime number theory. As an application of this theorem, the authors obtain two further results involving the pair correlation of the zeros of \(H_\lambda(s)\). In the proof they use a version of the Landau-Gonek formula (see e.g., \textit{S. M. Gonek} [Contemp. Math. 143, 395--413 (1993; Zbl 0791.11043)]) which has a good error term. This entails expressions involving the function \(n_x\), the closest prime power to \(x\), which they successfully evaluate.
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references