New explicit bounds for Mertens function and the reciprocal of the Riemann zeta-function

Ethan S. Lee,Nicol Leong
2024-07-26
Abstract:In this paper, we establish new explicit bounds for the Mertens function $M(x)$. In particular, we compare $M(x)$ against a short-sum over the non-trivial zeros of the Riemann zeta-function $\zeta(s)$, whose difference we can bound using recent computations and explicit bounds for the reciprocal of $\zeta(s)$. Using this relationship, we are able to prove explicit versions of $M(x) \ll x\exp\left(-\eta_1 \sqrt{\log{x}}\right)$ and $M(x) \ll x\exp\left(-\eta_2 (\log{x})^{3/5} (\log\log{x})^{-1/5}\right)$ for some $\eta_i > 0$. Our bounds with the latter form are the first explicit results of their kind. In the process of proving these, we establish another novel result, namely explicit bounds of the form $1/\zeta(\sigma + it) \ll (\log{t})^{2/3} (\log\log{t})^{1/4}$.
Number Theory
What problem does this paper attempt to address?
This paper attempts to solve the problem regarding the explicit bounds of the Mertens function \(M(x)\) and the reciprocal of the Riemann Zeta function \(\zeta(s)\). Specifically, the author aims to establish new explicit bounds to improve the existing results and provide more accurate estimates for \(M(x)\). ### Main objectives of the paper: 1. **Establish new explicit bounds**: The author hopes to establish new explicit bounds for the Mertens function \(M(x)\), especially in comparison with the short sum of non - trivial zeros, and limit the differences between them through recent calculations and the explicit bounds of the reciprocal of \(\zeta(s)\). 2. **Prove upper bounds of specific forms**: The goals of the paper are to prove explicit bounds in the following two forms: - \(M(x)\ll x\exp\left(-\eta_1\sqrt{\log x}\right)\) - \(M(x)\ll x\exp\left(-\eta_2(\log x)^{3/5}(\log\log x)^{-1/5}\right)\) where \(\eta_i > 0\) are constants. In particular, the second form is an explicitly given result for the first time. 3. **Provide specific numerical results**: The paper also provides specific numerical results. For example, for different \(x_0\), calculate specific \(c_1(x_0)\) and \(c_2(x_0)\) such that \(|M(x)| < c_1(x_0)x\exp\left(-c_2(x_0)\sqrt{\log x}\right)\). ### Methods for solving the problem: - **Combination of analytical and computational methods**: The author uses the truncated Perron formula to represent \(M(x)\) as an integral and compares it with the short sum of non - trivial zeros through the Cauchy residue theorem. - **Utilize zero - free regions**: By using the classical and Korobov - Vinogradov zero - free regions, the author obtains the explicit bounds of the reciprocal of \(\zeta(s)\), which helps to limit the differences between the integral and the short sum. - **Optimize parameter selection**: By optimizing the parameters \(T\), \(d_1\) and \(\omega\), the author obtains the optimal explicit bounds. ### Applications: These new results have multiple applications in the literature. For example, they can be used to study the error term of the number \(L(x)\) of square - full integers and improve the explicit descriptions of other related problems. In conclusion, this paper successfully establishes new explicit bounds for the Mertens function by introducing new techniques and optimizing existing methods, thus providing important tools and references for further research.