Atiyah–Bott fixed-point theorem

Atiyah–Bott fixed-point theorem
Name	Atiyah–Bott fixed-point theorem
Field	Algebraic topology; Differential geometry; Representation theory
Introduced	1960s
Authors	Michael Atiyah; Raoul Bott

Atiyah–Bott fixed-point theorem is a foundational result connecting fixed points of continuous group actions on manifolds with global topological invariants, synthesizing ideas from index theory, characteristic classes, and equivariant methods. The theorem, formulated by Michael Atiyah and Raoul Bott, provides a localization formula that expresses equivariant indices and cohomological integrals as sums of contributions from fixed-point loci associated to group elements, linking to earlier work of Lefschetz, Hirzebruch, and McKean–Singer. It played a central role in subsequent developments involving the Atiyah–Singer index theorem, Witten, and the study of moduli spaces in gauge theory and algebraic geometry.

Statement

The theorem concerns a compact smooth manifold M with an action of a compact Lie group G, typically specializing to G = S^1 or a torus T = (S^1)^n, and an elliptic complex or equivariant vector bundle E with an equivariant operator D. For a group element g in G the theorem equates the equivariant index ind_g(D) with a sum over connected components F of the fixed-point set M^g; each summand involves characteristic classes: the Chern character ch(E|_F), the Todd class td(TF) or Â-genus in the spin case, and the equivariant Euler class e_G(N_F) of the normal bundle N_F. In formulaic terms, ind_g(D) = Σ_F ∫_F (ch_g(E|_F) · Todd_g(TF)) / e_g(N_F), echoing structures from the Hirzebruch–Riemann–Roch theorem and the Lefschetz fixed-point theorem.

Historical context and motivations

Atiyah and Bott developed the result in the context of the 1960s surge in index theory initiated by Atiyah and Isadore Singer through the Atiyah–Singer index theorem, and influenced by fixed-point ideas in Siegfried Lefschetz and Friedrich Hirzebruch. Motivations came from problems in gauge theory studied by Simon Donaldson and Edward Witten, from moduli of bundles considered by Armand Borel and Jean-Pierre Serre, and from representation-theoretic questions linked to Hermann Weyl and Harish-Chandra. The localization perspective also resonated with work in symplectic geometry by Vladimir Guillemin and Shlomo Sternberg and with algebraic geometry trends led by Alexander Grothendieck and David Mumford.

Equivariant cohomology and localization

Equivariant cohomology H_G^*(M) developed by Bertram Kostant and Raoul Bott (in joint contexts) and formalized in the Cartan model supplies the natural language for the theorem; the Cartan model relates differential forms on M to polynomial functions on the Lie algebra of G, connecting to constructions in Henri Cartan and Élie Cartan. The localization principle used by Atiyah and Bott generalizes earlier localization phenomena by Berline–Vergne and integrates with the equivariant Chern character ch_G and equivariant characteristic classes including td_G and the equivariant Euler class e_G. These tools enabled translation between geometric fixed-point data and global indices, paralleling ideas appearing later in the study of Hamiltonian actions by Michèle Vergne and Frances Kirwan.

Proof outline

Atiyah and Bott's proof combines analytic and topological techniques. One begins with an elliptic complex and the construction of its equivariant index via heat kernel methods inspired by McKean–Singer and analytic localization methods reminiscent of Witten's supersymmetric localization. Using the equivariant deformation of the symbol and Thom isomorphism in equivariant K-theory, one reduces global index computations to contributions from tubular neighborhoods of fixed components, invoking the equivariant Chern character to pass to equivariant cohomology and applying residue-like evaluation against equivariant Euler classes. Along the way, tools from K-theory by Atiyah and Michael Karoubi, as well as spectral flow ideas connected to Graeme Segal and Daniel Quillen, play structural roles.

Examples and applications

Classical examples include recovering the holomorphic Lefschetz formula for automorphisms of complex projective varieties studied by Kunihiko Kodaira and Oscar Zariski, applications to the Riemann–Roch theorem for varieties in the style of Hirzebruch and Grothendieck, and calculations of characters of induced representations in the spirit of Weyl character formula and Harish-Chandra theory. In symplectic geometry, the theorem underpins the Duistermaat–Heckman formula studied by Hans Duistermaat and Gert Heckman and computations in equivariant cohomology of moduli spaces such as the moduli space of flat connections on Riemann surfaces analyzed by William Goldman and Nigel Hitchin. In mathematical physics it informs partition-function localization in supersymmetry and quantum field theory approaches of Edward Witten and computations in topological quantum field theory of Atiyah (Michael) style axioms.

Generalizations and related results

The Atiyah–Bott fixed-point theorem spawned generalizations in equivariant K-theory, algebraic geometry, and derived contexts. Notable extensions include the Berline–Vergne localization formula, the equivariant Riemann–Roch theorems of Fulton and MacPherson and of Edidin–Graham, and derived or motivic refinements connected to work by Maxim Kontsevich and Dennis Gaitsgory. Relations to Floer homology from Andreas Floer and to mirror symmetry influenced by Kontsevich and Alexander Givental reflect the theorem’s reach across Mumford-style moduli problems, representation theory of loop groups influenced by Pressley–Segal perspectives, and applications to index computations in noncompact and groupoid settings treated by Connes.

Category:Fixed-point theorems