Secants of Abelian Varieties, Theta Functions, and Soliton Equations1
I. A. Taimanov IntroductionThe main aim of the present article is to overview the applications of soliton equations to the geometry of Abelian varieties.Novikov was the first to indicate a possibility of this by conjecturing that Jacobi varieties are exactly principally-polarised Abelian varieties such that the Kadomtsev-Petviashvili equation integrates in their theta functions.This conjecture followed a stormy starting period of the development of finitezone integration of soliton equations (the Korteweg-de Vries equation, the Toda lattice, the sine-Gordon equation, etc.) (see [31,107]) which led to the method of Baker-Akhieser functions. This method was proposed by Krichever for constructing finite-zone solutions to the Kadomtsev-Petviashvili equation and describing commutative rings of ordinary differential operators of rank 1 ([61, 62]). The latter results inspired the Novikov conjecture. Its proof by Shiota ([98]) led to solution of one of the oldest and prominent problems of algebraic geometry, the RiemannSchottky problem.At the same time, Arbarello and De Concini ([2]) showed that the Novikov conjecture interweaves tightly with the ideas of Gunning who proposed to describe Jacobi varieties as admitting sufficiently many trisecants. The last condition amounts to validity of special theta function identities ([45]). In their research Arbarello and De Concini used the observation of Mumford that many soliton equations (the Korteweg-de Vries, Kadomtsev-Petviashvili, and sine-Gordon equations) are "hidden" in the Fay trisecant formula ([82]).The present article gives a survey of these papers on the Riemann-Schottky problem as well as of the papers on the analogue of the Novikov conjecture for Prym varieties. The analogue is abstracted further because the Veselov-Novikov, Landau-Lifschitz and BKP equations integrated in Prym theta functions are also "hidden" in the Fay and Beauville-Debarre quadrisecant formulae.Chapter 1 contains a brief introduction to the analytic theory of theta functions (see also [41,81]) which can be considered as self-contained if supplemented with information on cohomologies with coefficients in sheaves, for instance, in the amount of Chern's book [17].Chapter 2 provides necessary information on the theta functions of Jacobi and Prym (the detailed proofs are exposed, for instance, in [36,41]) and the inference of the trisecant and quadrisecant formulae.Chapters 3 and 4 are devoted to application of soliton equations to the theory of Jacobi and Prym varieties. Therewith necessary information on finite-zone solutions to non-linear equations and on Baker-Akhieser functions is exposed in brief in §8. This information is given in detail in the survey articles on finite-zone theory [31,62,66,27] (see also [8,64]). We hope that the present survey completes them in the part related to application of finite-zone theory to the geometry of Abelian varieties.