The Riemann hypothesis for certain integrals of Eisenstein series

Jeffrey C. Lagarias,Masatoshi Suzuki
2005-11-03
Abstract: This paper studies the non-holomorphic Eisenstein series E(z,s) for the modular surface, and shows that integration with respect to certain non-negative measures gives meromorphic functions of s that have all their zeros on the critical line Re(s) = 1/2. For the constant term of the Eisenstein series it shows that all zeros are on the critical line for fixed y= Im(z) \ge 1, except possibly for two real zeros, which are present if and only if y > 4 \pi e^{-\gamma} = 7.0555+. It shows the Riemann hypothesis holds for all truncation integrals with truncation parameter T \ge 1. For T=1 this proves the Riemann hypothesis for a zeta function recently introduced by Lin Weng, attached to rank 2 semistable lattices over the rationals.
Number Theory
What problem does this paper attempt to address?
This paper attempts to solve the Riemann hypothesis problem related to the integral of Eisenstein series. Specifically, the author studies the properties of the non - holomorphic Eisenstein series \(E(z, s)\) for the modular surface \(PSL(2,\mathbb{Z})\backslash\mathbb{H}\), and shows that the function \(F_{\mu}(s)\) given by the integral under certain non - negative measures \(\mu(z)\) has all its zeros on the critical line \(\Re(s)=\frac{1}{2}\). ### Main problems and conclusions 1. **For specific truncated integrals**: - The author proves that for the truncation parameter \(T\geq1\), the integral \[ I(T, s):=\int_{D_T}E^*(x + iy, s)\frac{dxdy}{y^2} \] where \(D_T = \{z = x+iy\mid-\frac{1}{2}\leq x\leq\frac{1}{2}, y\geq T\}\), all zeros are on the critical line \(\Re(s)=\frac{1}{2}\) (see Theorem 2). 2. **For the zeta function introduced by Weng**: - When \(T = 1\), the above result proves that the zeta function \(Z_{2,\mathbb{Q}}(s)\) recently introduced by Weng and related to rank - 2 semi - stable lattices satisfies the Riemann hypothesis (see Theorem 1): \[ Z_{2,\mathbb{Q}}(s):=\zeta^*(2s)\left(\frac{1}{s - 1}\right)-\zeta^*(2 - 2s)\left(\frac{1}{s}\right) \] All its zeros are on the critical line \(\Re(s)=\frac{1}{2}\). 3. **For the constant term of the Eisenstein series**: - For \(y\geq1\), the constant term \(a_0(y, s)\) of the Eisenstein series satisfies the modified Riemann hypothesis (see Theorem 3). That is: \[ a_0(y, s):=\zeta^*(2s)y^s+\zeta^*(2 - 2s)y^{1 - s} \] - When \(1\leq y\leq y^*\approx7.0555\), all zeros are on the critical line \(\Re(s)=\frac{1}{2}\). - When \(y > y^*\), in addition to on the critical line, there are two real zeros \(\rho_y, 1-\rho_y\) and \(\frac{1}{2}<\rho_y<1\), and as \(y\rightarrow\infty\), \(\rho_y\rightarrow1\). ### Key formulas - The complete Riemann zeta function: \[ \zeta^*(s):=\pi^{-\frac{s}{2}}\Gamma\left(\frac{s}{2}\right)\zeta(s) \] - The complete form of the Eisenstein series: \[ E^*(z, s):=\pi^{-s}\Gamma(s)E(z, s)=\pi^{-s}\Gamma(s)\sum_{(m, n)\in\mathbb{Z}^2\setminus(0, 0)}\frac{y^s}{|mz + n|^{2s}} \] - The truncated integral \(I(T,\)