On construction of differential $\mathbb Z$-graded varieties
Aliaksandr Hancharuk, Ruben Louis
TL;DR
The paper develops a computable framework for constructing $ abla$-graded $Q$-varieties that encode both the Koszul–Tate resolution of a quotient $ abla ext{O}/ abla ext{I}$ and a positively graded $Q$-variety associated with Lie–Rinehart algebras. The central innovation is an arborescent Koszul–Tate resolution that feeds into an explicit arborescent extension, with a refined homological perturbation algorithm that terminates in finite steps under finite projective resolutions. The main results establish existence and explicit descriptions of these extensions over $ abla ext{O}/ abla ext{I}$ and over $ abla ext{O}$ preserving $ abla ext{I}$, including a robust set of geometric applications to universal Lie $ abla$-∞ algebroids and singular foliations. Concrete examples demonstrate the methodology on Lie algebroids, vector fields vanishing on subspaces, and quadratic-ideal preserving derivations, highlighting new multipart algebraic structures and multiplications arising from hook maps. This framework provides a computable bridge between singular spaces, Lie–Rinehart theories, and higher graded geometric structures with potential impact on BV/BFV formalisms and deformation theory.
Abstract
Given a commutative unital algebra $\mathcal O$, a proper ideal $\mathcal I$ in $\mathcal O$, and a positively graded differential variety over $\mathcal O/\mathcal I$, we provide a $\mathbb Z$-graded extension, whose negative part is an arborescent Koszul-Tate resolution of $\mathcal O/ \mathcal I$. This extension is obtained through an algorithm exploiting the explicit homotopy retract data of the arborescent Koszul-Tate resolution, so that the number of homological computations in the construction is significantly reduced. For a positively graded differential variety over $\mathcal O$ that preserves the ideal $\mathcal I$, the extension admits a manifest description in terms of decorated trees and computed data. As a by-product, to every Lie--Rinehart algebra over the coordinate ring of an affine variety $ W \subseteq M = \mathbb{C}^d $, one associates an explicit differential $\mathbb{Z}$-graded variety over $M$ whose negative component is the arborescent Koszul--Tate resolution of the coordinate ring $\mathbb C[x_1, \ldots, x_d]/\mathcal I_W$ of $W$, and whose positive component is the universal dg-variety of the given Lie--Rinehart algebra. Concrete examples are given.
