Affine Lie algebras on Riemann surfaces (Q1337913): Difference between revisions

The theory of highest weight representations for affine Kac-Moody algebras is a well developed subject. These algebras are constructed as follows: Starting from a finite-dimensional simple Lie algebra \(\mathfrak g\) (or from the abelian Lie algebra \(\mathfrak{gl} (1)\)) one takes the ring \(A= \mathbb C [t, t^{-1}]\), constructs the current (or loop) algebra \(\mathfrak g\otimes A\) and takes its nontrivial (one-dimensional) central extension \(\widehat {\mathfrak g}\). For \(\mathfrak{gl} (1)\) one obtains the Heisenberg algebra. If one interprets the algebra \(A\) as the algebra of meromorphic functions on \(\mathbb P_ 1 (\mathbb C)\) (the Riemann surface of genus 0) which are holomorphic outside \(\{0,\infty\}\) it is possible to generalize the situation to higher genus Riemann surfaces \(X\). \(A\) is now the algebra of meromorphic functions on \(X\) which are holomorphic outside two fixed points \(z_ +, z_ -\). In this way one obtains higher genus affine Kac-Moody algebras. This goes back to \textit{I. M. Krichever} and \textit{S. P. Novikov} [Funkts. Anal. Prilozh. 21, 46--63 (1987; Zbl 0634.17010); Funkts. Anal. Prilozh. 21, 47--61 (1987; Zbl 0659.17012); Funkts. Anal. Prilozh 23, 24--40 (1989; Zbl 0684.17012)]. For a multi-point generalization, see also the reviewer [Differential operator algebras on compact Riemann surfaces, in: Generalized symmetries in physics, Clausthal, 1993, World Scientific, 425--434 (1994)]. The author of the article under review started the examination of these algebras in the case that the Riemann surface is an elliptic curve (e.g. a one-dimensional complex torus) and 2 points for allowed poles in [Funkts. Anal. Prilozh. 24, 51--61 (1990; Zbl 0715.17023)]. In this paper, he generalizes his approach to higher genus Riemann surfaces. The current algebra and certain central extensions (not necessarily one-dimensional ones) are studied. The current group is introduced and the orbits of the adjoint and coadjoint action are studied. One main tool is the generalization of the approach of Frenkel and Segal relating the orbits of Kac-Moody algebras to monodromy groups of certain differential equations. The author defines the Weyl groups for these higher genus Kac-Moody algebras. He shows that the space of ``principal orbits'' (orbits corresponding to commutative monodromy groups) of the adjoint representation is in 1-1 correspondence to a quotient of the first homology space of \(X\) with certain points removed (depending on the cocycles considered) with values in the Cartan subalgebra of \({\mathfrak g}\) under this Weyl group. A similar theorem is shown for the coadjoint action.