Kronecker second limit formula for real quadratic fields
YoungJu Choie, Rahul Kumar
TL;DR
The paper derives the second Kronecker limit formula for real quadratic fields by connecting Zagier's zeta function to zeta values of narrow ideal classes through the higher Herglotz-Zagier-Novikov function $\mathscr{F}_{k}(x; \alpha, \beta)$. It establishes analytic continuation and a suite of functional equations (two-, three-, six-term) for $\mathscr{F}_{k}$, along with a cohomological interpretation via SL$_2(\mathbb{Z})$-cocycles and a link to a generalized Dedekind eta function. The main results express $D^{k/2}$-values of zeta at natural integers in terms of $\mathscr{F}_{2k}$ evaluated at reduced forms, yielding explicit two-term limits for the $k=1$ case and higher analogues. The work also provides numerical verification and identifies rational zeta values emerging from combinations of zeta values, proposing a coherent arithmetic framework with potential connections to cohomology and Stark-type phenomena.
Abstract
In this paper, the second Kronecker ``limit" formula for a real quadratic field is established for the first time. More precisely, we obtain the second Kronecker limit formula of Zagier's zeta function. Using the reduction theory of Zagier, which connects Zagier's zeta function to the zeta function of real quadratic fields, we express the values of the zeta function of narrow ideal classes in real quadratic fields at natural arguments in terms of an analytic function which we call the \emph{higher Herglotz-Zagier-Novikov function} and denote it by $\mathscr{F}_k(x; α, β)$. This function plays a central role in our study. The function $\mathscr{F}_k(x; α, β)$ possesses elegant properties, for example, we prove that it satisfies the two, three and six-term functional equations. As a result of our Kronecker limit formula and functional equations, we provide another expression for the combinations of zeta values. Finally, we interpret our Kronecker ``limit" formula in terms of cohomological relations and establish a connection between $\mathscr{F}_k(x; α, β)$ and a generalized Dedekind-eta function.
