1 S ep 2 00 1 On orbifold elliptic genus
C. Dong,Kefeng Liu,M. Xiaonan
2002-01-01
Abstract:Elliptic genus was derived as the partition function in quantum field theory [26]. Mathematically it is the beautiful combination of topology of manifolds, index theory and modular forms (cf. [15], [10]). Elliptic genus for smooth manifolds has been well-studied. Recently, Borisov and Libgober ([3], [4]) proposed some definitions of elliptic genus for certain singular spaces, especially for complex orbifolds which is a global quotient M/G, here the finite group G acts holomorphically on complex variety M . Similar definitions have been introduced by string theorists in the 80s, in the study of orbifold string theory. One of their guiding principles is the modular invariance. More recently orbifold string theory has attracted the attention of geometers and topologists. For example Chen and Ruan (cf. [7], [22]) have defined orbifold cohomology and orbifold quantum cohomology groups. One of most important properties for elliptic genus is its rigidity property under compact Lie group action. For smooth manifolds, the rigidity and its generalizations have been well studied. Since orbifold elliptic genus is the partition function of orbifold string theory, it is natural to expect the rigidity property for orbifold elliptic genus. Although the global quotients from a very important class of orbifolds, many interesting orbifolds are not global quotients. For example, most of the Calabi-Yau hypersurfaces of weighted projective spaces are not global quotients. In this paper we define elliptic genus for general orbifolds which generalizes the definition of Borisov and Libgober, and prove their rigidity property. We actually introduce the more general elliptic genus involving twisted bundles and proved its rigidity. The idea of considering the weights in the definition of orbifold elliptic genus comes from [4] and our proof of the K-theory version of Witten rigidity theorems [21, §4]. The proof of the rigidity is again a combination of modular invariance and the index theory. Only more complicated combinatorics are involved in the definition and the proof. This paper is organized as follows: In Sections 1 and 2 we review the equivariant index theorem on orbifolds. We define orbifold elliptic genus and prove its rigidity for almost complex orbifolds in Section 3. Finally in Section 4 we introduce orbifold elliptic genus for spin orbifolds, and will study its rigidity property on a later occasion. The authors would like to than Jian Zhou for many interesting discussions regarding orbifold elliptic genus.