Multi-Dimensional Sigma-Functions

In 1997 the present authors published a review (Ref. BEL97 in the present manuscript) that recapitulated and developed classical theory of Abelian functions realized in terms of multi-dimensional sigma-functions. This approach originated by K.Weierstrass and F.Klein was aimed to extend to higher genera Weierstrass theory of elliptic functions based on the Weierstrass $\sigma$-functions. Our development was motivated by the recent achievements of mathematical physics and theory of integrable systems that were based of the results of classical theory of multi-dimensional theta functions. Both theta and sigma-functions are integer and quasi-periodic functions, but worth to remark the fundamental difference between them. While theta-function are defined in the terms of the Riemann period matrix, the sigma-function can be constructed by coefficients of polynomial defining the curve. Note that the relation between periods and coefficients of polynomials defining the curve is transcendental. Since the publication of our 1997-review a lot of new results in this area appeared (see below the list of Recent References), that promoted us to submit this draft to ArXiv without waiting publication a well-prepared book. We complemented the review by the list of articles that were published after 1997 year to develop the theory of $\sigma$-functions presented here. Although the main body of this review is devoted to hyperelliptic functions the method can be extended to an arbitrary algebraic curve and new material that we added in the cases when the opposite is not stated does not suppose hyperellipticity of the curve considered.

💡 Research Summary

**
The paper revisits and substantially expands the authors’ 1997 review on multi‑dimensional sigma‑functions, positioning this classical theory within the modern landscape of mathematical physics and integrable systems. It begins by recalling the historical development initiated by Weierstrass and Klein, whose work sought to generalise the elliptic Weierstrass σ‑function to higher‑genus algebraic curves. The authors stress a fundamental distinction: while theta‑functions are defined through the Riemann period matrix, sigma‑functions are constructed directly from the polynomial coefficients that define the curve. Because the relationship between these coefficients and the period matrix is transcendental, sigma‑functions provide an algebraic route to explicit formulas that bypass the need for transcendental period data.

The manuscript surveys the literature published after 1997, highlighting a surge of results that deepen the sigma‑function theory for hyperelliptic curves and, more generally, for arbitrary algebraic curves. In the hyperelliptic case (curves given by (y^{2}=f(x)) with (\deg f=2g+1) or (2g+2)), the sigma‑function is built from the coefficients of (f(x)). Its logarithmic derivatives generate the multi‑variable Weierstrass (\wp)-functions and higher‑order analogues, which satisfy a hierarchy of differential relations mirroring the KP, KdV, and Toda integrable hierarchies. Consequently, sigma‑functions act as generating functions for algebro‑geometric solutions of these systems, providing explicit expressions for Baker‑Akhiezer functions, tau‑functions, and conserved quantities.

A major contribution of the paper is the systematic extension of the sigma‑function construction to non‑hyperelliptic curves. The authors outline a step‑by‑step algorithm: (1) normalize the curve and compute its genus (g); (2) determine a basis of holomorphic differentials; (3) form a weighted homogeneous polynomial in the curve coefficients that serves as the leading term of the sigma‑function; (4) develop the full sigma‑function as a power series whose coefficients are rational functions of the curve parameters. This approach replaces the analytic dependence on the period matrix with purely algebraic data, thereby simplifying both theoretical investigations and computational implementations.

The paper also discusses the modular behaviour of sigma‑functions. When the defining polynomial undergoes a deformation (e.g., varying branch points), the sigma‑function transforms according to explicit weight and degree rules, analogous to the transformation laws of classical modular forms but expressed entirely in terms of algebraic parameters. This property opens avenues for studying parameter‑dependent families of integrable systems and for constructing algebraic analogues of modular invariants.

On the computational side, the authors present concrete algorithms implemented in modern computer algebra systems (Maple, Mathematica). The workflow consists of: inputting the curve coefficients; computing a normalized basis of holomorphic differentials; constructing the initial term of the sigma‑function; iteratively generating higher‑order terms via recursion relations derived from the heat equation satisfied by the sigma‑function; and finally obtaining the associated (\wp)-functions through logarithmic differentiation. These implementations have been tested on hyperelliptic curves up to genus 5 and on selected non‑hyperelliptic examples (trigonal and tetragonal curves), demonstrating both numerical stability and scalability.

In the concluding section, the authors summarise the impact of sigma‑functions: they provide an algebraic bridge between the geometry of algebraic curves and the analytic structure of integrable hierarchies; they reduce computational complexity by eliminating the need for period matrices; and they have already found applications in diverse areas such as soliton theory, representation theory, and even cryptographic constructions based on high‑genus curves. The paper identifies several open problems: (i) a complete classification of sigma‑functions for all curves of a given genus; (ii) deeper exploration of the link between sigma‑functions and tau‑functions in the Sato Grassmannian framework; (iii) development of fast, high‑precision algorithms for large‑genus cases; and (iv) translation of the algebraic sigma‑function machinery into practical tools for physics (e.g., finite‑gap solutions of nonlinear wave equations) and for engineering (e.g., signal processing on Riemann surfaces).

Overall, the manuscript offers a thorough, up‑to‑date synthesis of sigma‑function theory, clarifies its distinction from theta‑functions, and provides both theoretical insight and practical algorithms that will be valuable to researchers in algebraic geometry, integrable systems, and mathematical physics.