Springer resolution

Springer resolution
Name	Springer resolution
Field	Algebraic geometry
Introduced	mid-20th century
Main contributors	T.A. Springer
Related	Flag variety, nilpotent cone, Grothendieck–Springer resolution

Springer resolution is a geometric construction linking the geometry of the nilpotent cone in a semisimple Lie algebra to the geometry of the flag variety and to representations of Weyl groups. The resolution provides a desingularization of the nilpotent variety by parametrizing pairs consisting of a nilpotent element and a compatible flag; it is central to the study of T.A. Springer's work on Weyl group actions, the geometry of nilpotent orbits, and connections with the Weyl group and flag variety. The construction underlies many developments in geometric representation theory, including the proof of the Springer correspondence and influences on the Kazhdan–Lusztig theory and the Deligne–Lusztig theory.

Definition and construction

Let G be a connected complex semisimple algebraic group with Lie algebra g and B a Borel subgroup with flag variety G/B. The Springer resolution is the variety { (x, F) ∈ g × G/B | x ∈ Lie(G) is nilpotent and x preserves the flag F } equipped with the projection to the nilpotent cone N ⊂ g given by (x, F) ↦ x. Equivalently, one constructs the cotangent bundle T*(G/B) and observes an embedding T*(G/B) → g × G/B; composing with projection to g yields a proper, G-equivariant morphism π: T*(G/B) → N that is birational onto each nilpotent orbit closure. In classical terms π is obtained via the moment map for the cotangent action of G on G/B; this uses the identification of T*(G/B) with {(v, F) | v ∈ g*, F ∈ G/B, v|Lie(B_F)=0} and invoking the Killing form to identify g ≅ g*.

Historical background and development

The construction arose from efforts by T.A. Springer in the 1970s to relate cohomology of varieties to representations of Weyl groups. Springer's original papers built on earlier geometry of flag varietys studied by A. Borel and the structure theory developed by Claude Chevalley and Armand Borel. Subsequent refinements and generalizations connected the resolution to the work of Grothendieck on simultaneous resolution of singularities and to advances by George Lusztig on character sheaves. The Grothendieck–Springer simultaneous resolution incorporates the Cartan and the adjoint quotient and links to results of H. Kraft and C. Procesi about nilpotent orbit closures. Later interactions with Kazhdan and Lusztig produced the Kazhdan–Lusztig conjectures, and connections to the Beilinson–Bernstein localization theorem and modules over enveloping algebras were clarified by Jean Bernstein and I.N. Bernstein's school.

Geometric and algebraic properties

The Springer resolution π: T*(G/B) → N is a proper, surjective, G-equivariant morphism that is an isomorphism over the regular nilpotent orbit and provides a resolution of singularities of closures of nilpotent orbits. Fibers π^{-1}(x) are projective varieties known as Springer fibers; for x nilpotent these fibers are unions of Schubert varieties in G/B and their topology reflects representation-theoretic data. The cohomology H^*(π^{-1}(x), Q) carries a natural action of the Weyl group W of G (the Springer action), producing representations that realize the Springer correspondence between irreducible W-representations and pairs (O, L) where O is a nilpotent orbit and L an irreducible local system on O. The map interacts with the Bruhat decomposition of G/B, the Borel–de Siebenthal theory of parabolic subgroups, and purity statements coming from the work of Pierre Deligne on weights.

Examples and computations

In type A (G = SL_n), the nilpotent cone corresponds to nilpotent matrices and the Springer fiber over a nilpotent matrix with Jordan blocks of sizes given by a partition λ of n is isomorphic to the variety of flags stabilized by that matrix; these are described combinatorially via standard Young tableaux. The cohomology of these fibers yields the classical correspondence between irreducible representations of the symmetric group S_n and partitions, recovering the Specht modules through geometry studied by Richard Stanley and Jean-Yves Thibon. For G of type B, C, D (orthogonal and symplectic groups), Springer fibers relate to signed Young diagrams and to representations of hyperoctahedral groups; explicit computations use work of Spaltenstein and T.X. Springer's further analysis. Low-rank examples (SL_2, SL_3, G_2) provide explicit descriptions: for SL_2 the resolution is trivial and fibers are points or projective lines, while for SL_3 fibers over subregular nilpotents are unions of projective lines intersecting according to Dynkin diagrams as in work by Brieskorn and Slodowy.

Applications in representation theory

The Springer resolution is foundational for constructing the Springer correspondence, which assigns Weyl group representations to nilpotent orbits and lends geometric interpretation to character formulas. It plays a key role in proving the existence of W-actions on intersection cohomology groups and in the realization of Hecke algebra modules via perverse sheaves on G/B as developed by G. Lusztig and Kazhdan–Lusztig. The geometry underpins the proof of multiplicity formulas in category O by I.N. Bernstein, J. Bernstein, and J. Humphreys, and feeds into the construction of character sheaves and the study of unipotent representations of finite reductive groups via the Deligne–Lusztig framework. Connections to affine Springer fibers link the resolution to the theory of automorphic forms and the study of orbital integrals in the trace formula by R.P. Langlands and Kazhdan.

Generalizations and variants

Variants include the Grothendieck–Springer simultaneous resolution relating g × G/B to the adjoint quotient and the parabolic Springer resolution replacing B by a parabolic subgroup P, yielding maps from T*(G/P) to closures of Richardson orbits. Affine and global analogues produce affine Springer fibers in the context of loop groups and the Hitchin fibration investigated by Nigel Hitchin; these are central to the geometric Langlands program developed by Edward Frenkel and collaborators. Other extensions involve modular and mixed-characteristic settings considered by T. Haines and Ivan Mirković, and categorical enhancements using derived and dg-schemes in the work of Dennis Gaitsgory and Jacob Lurie.

Category:Algebraic geometry