A classical model for derived critical loci (Q892749)

From MaRDI portal
scientific article
Language Label Description Also known as
English
A classical model for derived critical loci
scientific article

    Statements

    A classical model for derived critical loci (English)
    0 references
    12 November 2015
    0 references
    There are three typical main examples of \(-1\)-shifted symplectic derived schemes introduced by Pantev, Toën, Vaquié and Vezzosi [\textit{T. Pantev} et al., Publ. Math., Inst. Hautes Étud. Sci. 117, 271--328 (2013; Zbl 1328.14027)]: (a) The critical locus of a regular function \(f : U \to \mathbb{A}^1\) on a smooth \(K\)-scheme \(U\). (b) The intersection \(L \cap M\) of smooth Lagrangians \(L, M\) in an algebraic symplectic manifold. (c) A moduli scheme \(\mathcal{M}\) of stable coherent sheaves on a Calabi-Yau 3-fold. These structures have applications in symplectic geometry and in Donaldson-Thomas theory of Calabi-Yau 3-folds. The paper under review introduces and develops a new class of geometric objects called \(d\)-critical loci \((X, s)\) which are entirely classical. Here \(s\) is a global section of a certain sheaf on a \(K\)-scheme \(X\), satisfying some local conditions. This paper also develops a parallel theory in which \(X\) is a complex analytic space. The paper under review, moreover defines and studies the notions of canonical bundle \(K_{X,s}\), orientations, group actions and equivariant structures etc. for a \(d\)-critical locus \((X,s)\), and finally extends all these notions to the case of Artin stacks. This is a central example of a \(d\)-critical locus \((X,s)\): Let \(U\) be a smooth \(\mathbb{C}\)-scheme and, \(f : U \to \mathbb{A}^1\) a regular function and \(X\) be the critical locus of \(f\). Write \(i: X \to U\) for the inclusion, and \(I \subset i^{-1}(\mathcal{O}_U)\) for the sheaf of ideals vanishing on \(X \subset U\). Then \(i^{-1}(f) \in H^0(i^{-1}(\mathcal{O}_U))\) with \(d(i^{-1}(f)) \in H^0(i^{-1}(T^*_U))\) so \(i^{-1}(f)+I^2 \in H^0(i^{-1}(\mathcal{O}_U)/I^2)\) with \(d(i^{-1}(f)+I^2)=0 \in H^0(i^{-1}(T^*_U)/I\cdot i^{-1}(T^*_U))\). Thus, there is a unique section \(s \in H^0(\mathcal{S}_X)\) with \(\iota(s)=i^{-1}(f)+I^2\), where \[ 0\to \mathcal{S}_X\overset{\iota}\rightarrow\frac{i^{-1}(\mathcal{O}_U)}{I^2}\overset {d}\rightarrow\frac{i^{-1}(T^*_U)}{I\cdot i^{-1}(T^*_U)}. \] \(s\) remembers all of the information about \(f\) which makes sense intrinsically on \(X\), rather than on the ambient space \(U\). There are significant applications of \(d\)-critical loci that are studied by the author of the paper under review and his collaborators. To mention three: 1) Suppose that \(K\) is algebraically closed with zero characteristic. Let \(X\) be the associated classical \(K\)-scheme of a \(-1\)-shifted symplectic derived scheme. Then \(X\) extends naturally to a \(d\)-critical locus \((X,s)\) in such a way that \(K_{X,s}\) is naturally isomorphic to the determinant line bundle of the cotangent complex of the derived scheme restricted to \(X^{\mathrm{red}}\). In other words, there is a (non-full) truncation functor from \(-1\)-shifted symplectic derived schemes to \(d\)-critical loci. In particular, examples (a)-(c) above have the structure of \(d\)-critical loci. 2) Let \((X, s)\) be a \(d\)-critical locus over \(K\) with an orientation. Then there exists a certain motive \(MF_{X,s}\) in a ring of motives over \(X\) that can be used for defining motivic Donaldson-Thomas invariants. 3) Let \((X, s)\) be a \(d\)-critical locus over \(\mathbb{C}\) with an orientation. Then there exist certain algebraic \(\mathbb{Z}\)-perverse sheaf \(P^\bullet_{X,s}\), \(\mathscr{D}\)-module \(D_{X,s}\) and mixed Hodge module \(H^\bullet_{ X,s}\) on \(X\). This is important for categorification of Donaldson-Thomas invariants.
    0 references
    0 references
    0 references
    0 references
    0 references
    \(-1\)-shifted symplectic derived schemes
    0 references
    \(d\)-critical loci
    0 references
    0 references
    0 references
    0 references