Generalizations of Thomae's formula for \(Z_n\) curves (Q609445)

In the book under review, the authors present the background necessary to understand and then prove Thomae's formula for \(Z_n\) curves. The renaissance of the research in this area is largely due to the interest of the mathematical physics community, but one should also mention applications of these formulae in explicit computations with curves over finite fields (see for instance [\textit{R. Lercier, D. Lubicz}, Ramanujan J. 12, No. 3, 399--423 (2006; Zbl 1166.11021)]). A {(\textit fully ramified) \(Z_n\) curve} is the compact Riemann surface associated to an equation \[ w^n = \prod_{i=1}^t (z-\lambda_i)^{\alpha_i} \] where \(\lambda_i \neq \lambda_j\) for \(i \neq j\) and for \(1 \leq i \leq t\) the powers \(\alpha_i\) satisfy \(1 \leq \alpha_i \leq n-1\), are coprime to \(n\) and their sum is divisible by \(n\). The genus of this Riemann surface is \(g=(n-1)(t-2)/2\). To any compact Riemann surface one can associate its Jacobian variety, which is a torus of the form \(\mathbb{C}^g/(\mathbb{Z}^g + \Pi \mathbb{Z}^g)\) where \(\Pi\) is a matrix depending on a choice of homology and regular differential bases and called a \textit{period matrix} (sometimes also called a Riemann matrix). With such a matrix in hand, one can define a holomorphic functions of \(g\) complex variables \(\theta \begin{bmatrix} \varepsilon \\ \varepsilon' \end{bmatrix}(z,\Pi)\) depending on two vectors \(\epsilon, \epsilon' \in \mathbb{R}^g\). The value at \(z=0\) of this function is called \textit{theta constant} (or by the German name Thetanullwert). The \textit{Thomae formulae} are a bridge between the analytic world of theta constants and the algebraic world of the equation of the curve. The main purpose of the authors is to show how to construct a finite set of \(\varepsilon,\epsilon' \in \mathbb{Q}^g\) and a polynomial \(P_{ \begin{bmatrix} \varepsilon \\ \epsilon' \end{bmatrix}}\) such that the value of the quotient \[ \frac{\theta^N \begin{bmatrix} \varepsilon \\ \varepsilon' \end{bmatrix}(0,\Pi)}{P_{ \begin{bmatrix} \epsilon \\ \epsilon' \end{bmatrix}}(\lambda_1,\ldots,\lambda_t)} \] is independent of \( \begin{bmatrix} \varepsilon \\ \varepsilon' \end{bmatrix}\) for some power \(N\). Such expressions are fundamental: for instance they give the equations of the Jacobian in terms of the equation of the curve and are also use to investigate solutions of differential equations in physics. The history of the formulae goes back to the 19th century and the work of \textit{J. Thomae} on \(Z_2\) curves (i.e. hyperelliptic curves) [Borchardt J. 71, 201--222 (1870; JFM 02.0244.01); J. Reine Angew. Math. 66, 92--96 (1866; ERAM 066.1717cj)]. See also [\textit{J. Guárdia}, Ann. Inst. Fourier 57, No. 4, 1253--1283 (2007; Zbl 1177.11052)] for some applications. One century later, in [\textit{M. Bershadsky, A. Radul}, Int. J. Mod. Phys. A 2, No. 1, 165--178 (1987; Zbl 1165.81373)], the so-called \textit{non-singular case} (i.e. \(\alpha_i=1\) for \(1 \leq i \leq t\)) was worked out using methods from string and quantum field theory and then reworked by \textit{A. Nakayashiki} [Publ. Res. Inst. Math. Sci. 33, No.6, 987--1015 (1997; Zbl 0921.14016)] replacing some of the ideas with the work of [\textit{J. D. Fay}, Theta functions on Riemann surfaces. Lecture Notes in Mathematics. 352. Berlin-Heidelberg-New York: Springer-Verlag. 137 p. (1973; Zbl 0281.30013)] on Szegő Kernel function. Modifying this method, `singular' cases were also handle in [\textit{V. Z. Enolski, T. Grava}, Lett. Math. Phys. 76, No. 2-3, 187--214 (2006; Zbl 1159.14018)] (note also the work of [\textit{K. Matsumoto}, Publ. Res. Inst. Math. Sci. 37, No.3, 419--440 (2001; Zbl 1006.32018)] in the case \(n=3\)). Finally, and this reference was apparently not available to the authors of the book at that time [Lett. Math. Phys. 94, No. 3, 313--333 (2010; Zbl 1213.14060)] \textit{Y. Kopeliovich} is able to give a formula for all \(Z_n\) curves generalizing Nakayashiki's method. Apart from \(Z_n\) curves, I am only aware of Thomae formula for genus \(3\) non hyperelliptic curves [\textit{H. Weber}, Preisschrift. Berlin. G. Reimer. (1876; JFM 08.0293.01); in: Algorithmic number theory. 6th international symposium, ANTS-VI, Burlington, VT, USA, June 13--18, 2004. Proceedings. Berlin: Springer. Lecture Notes in Computer Science 3076, 379--394 (2004; Zbl 1125.11331)]). The point of view of the book is to work out Thomae formulae for \(Z_n\) curves from `first principles', i.e. just using Riemann's theory of theta functions. The idea started with a seminar by Farkas, and then was developed after a note on the \(Z_2\) case [\textit{A. Eisenmann, H. M. Farkas}, Online J. Anal. Comb. 3, Article 2, 14 p., electronic only (2008; Zbl 1188.14018); and an article \textit{D. G. Ebin, H. M. Farkas}, J. Anal. Math. 111, 289--320 (2010; Zbl 1215.14030)] on the non-singular case. The book is divided into three main parts. The first one recalls the necessary results concerning compact Riemann surfaces and theta functions mainly without proofs which can be found for instance in [Riemann surfaces. 2nd ed. Graduate Texts in Mathematics. 71. New York etc.: Springer-Verlag. (1992; Zbl 0764.30001)]. It also summarizes the material one needs to know about \(Z_n\) curves. The second part deals with some didactic examples intended to help the reader assimilate the general cases. Non-singular \(Z_n\) curves and then singular cases with \(\alpha_i=1\) or \(n-1\) are addressed. Finally a last part is more a `work in progress' and proves special cases with different \(\alpha_i\) as no general theory is available yet. Note that the kind of Thomae formulae developed in the book are less precise than the original statement of Thomae or the ones found using Nakayashiki's method. In the latter the expression of the quotient is worked out in terms of \(\Pi\) up to an absolute constant. This constant is known for \(Z_2\) curves and in a preprint, Matsumoto and Terasoma are able to determine this constant in the case of non-singular \(Z_3\) curves. However the `elementary' approach which is chosen in the book makes it a nice development of Riemann's ideas and accessible to graduate students and young researchers.