No label defined (Q4225232)

The question of characterizing solutions \(u(x,y,t)\) of the KP equation: \[ 3u_{yy}=(4u_t+6uu_x-u_{xxx})_x \] that are doubly-periodic (namely, elliptic functions) in \(x\in \mathbb{C}\) was posed more than 20 years ago by \textit{S. P. Novikov} [Funct. Anal. Appl. 8, 236-246 (1974); translation from Funkts. Anal. Prilozh. 8, No. 3, 54-66 (1974; Zbl 0299.35017)], essentially solved by \textit{I.M. Krichever} [Funkts. Anal. Prilozh. 14, No. 4, 45-54 (1980; Zbl 0462.35080)] and recast in an algebro-geometric framework by \textit{A. Treibich} and \textit{J.-L. Verdier} [Proceedings of the Indo-French conference on geometry, Bombay, 1989, Hindustan Book Agency, 187-232 (1993; Zbl 0837.14011)]. Treibich and Verdier demonstrated that the question of characterizing KdV solutions elliptic in \(x\), namely KP solutions elliptic in \(x\) that are \(y\)-independent, is even more subtle [\textit{A. Treibich} and \textit{J.-L. Verdier}, Prog. Math., 88, 437-480 (1990; Zbl 0726.14024)]. The present authors have developed an unrelated approach to the latter question. This paper is a compendium of their results to date, illustrated by the main methods and proofs, and set in context by introduction, generalizations and a lavish bibliography. By virtue of the fact [proved in the Appendix of Treibich and Verdier (1993; loc. cit.)] that an \(x\)-elliptic KP solution is algebro-geometric, the problem of characterizing elliptic KdV solutions becomes equivalent, via the Lax-pair property: \[ L_t=[L^{3/2},L] \] to the following: characterize the potentials \(u(x)\) of Hill's equation: \[ Ly(x)=Ey(x),\qquad L=d^2/dx^2+u(x), \qquad u(x)\;\text{ an\;elliptic\;function} \] which are algebro-geometric, namely such that the centralizer \({\mathcal C}\) of \(L\) in the ring of differential operators contains an element of odd order. By extension from the real-valued theory, such potentials are called ``finite-gap''. The present authors in [Acta Math. 176, No. 1, 73-107 (1996; Zbl 0927.37040)], were able to give an effective characterization of said potentials in terms of a basis of solutions of the eigenvalue problem \(Ly(x)=Ey(x)\). The result is as follows: \(u\) is an elliptic finite-gap potential if and only if \(u\) is a Picard potential (i.e., if and only if for each \(E\in \mathbb{C}\) every solution of \(\psi''(x)+u(x)\psi (x)=E\psi (x)\) is meromorphic with respect to \(x\)). For finite-gap potentials, the property follows from the expression of the Baker-Akhiezer function in terms of theta functions. The proof that the condition is sufficient is based on Picard's theorem [cf. \textit{E. L. Ince}, Ordinary Differential Equations, Dover, New York (1944; Zbl 0063.02971)], which gives periodicity properties of solutions of Hill's equation under the assumption that all solutions are meromorphic; by virtue of that theorem, Floquet theory can be used to derive certain properties of the Green function; this in turn can be used, with the help of explicit recursive equations derived by the authors, to construct solutions effectively. This theory is made intricate (although, as the authors remark, it remains explicit at every step) by the geometry of the complex eigenvalues of the problem. The recursive approach gives as a by-product equations for the `spectral curves'. Much remains to be done to extend this approach to elliptic solutions of higher reductions of the KP equation, as well as other hierarchies (multi-component KP, e.g.) and in the two final sections the authors report on their current progress again at constructing solutions in terms of the complex geometry of the spectrum for the elliptic linear operator of the Lax pair. Once completed, this approach will constitute both a beautiful mathematical theory and an effective tool for applications. For the entire collection see [Zbl 0903.00040].