Complex algebraic stacks and morphisms of intersection complexes
Matthew Huynh
TL;DR
This work extends the Barthel–Brasselet–Fieseler–Gabber–Kaup construction of associated morphisms to finite-type complex algebraic stacks with affine stabilizers, proving the existence of $\\mu^{f}: f^{*}IC_{\\mathcal{Y}} o IC_{\\mathcal{X}}$ (equivalently $\\lambda_f: IC_{\\mathcal{Y}} o f_{*}IC_{\\mathcal{X}}$) for any morphism $f: \\mathcal{X} o \\mathcal{Y}$. The proof leverages a stack version of the Decomposition Theorem (via Sun’s result) and a resolution-based construction to establish the required commutative diagrams, while addressing non-uniqueness of associated morphisms. As an application, the Borel–Moore fundamental class of a closed substack of a Deligne–Mumford stack with affine stabilizers lifts to a class in the intersection cohomology, providing a robust mechanism to lift cycle data to intersection cohomology in the stack context. Overall, the paper broadens intersection-theoretic tools to the realm of algebraic stacks, with implications for singularities and cohomology in moduli problems.
Abstract
We generalize a construction of Barthel-Brasselet-Fieseler-Gabber-Kaup in the setting of complex varieties to the setting of finite type, complex algebraic stacks. Given two such stacks $\mathcal{X},\mathcal{Y}$ with affine stabilizers, and a morphism between them, we construct a morphism from the pullback of the intersection complex of $\mathcal{Y}$ to the intersection complex of $\mathcal{X}$. As an application, we show that the Borel-Moore fundamental class of a closed substack $\mathcal{Z}$ in a Deligne-Mumford stack $\mathcal{X}$ lifts to a class in the intersection cohomology of $\mathcal{X}$.
