\(\mathcal D\)-modules on analytic spaces (Q1176253): Difference between revisions

The author gives a definition of the category \(M(X,{\mathcal D})\) of ``\(\mathcal D\)-modules'' on a complex analytic space \(X\), without using any structure sheaf on \(X\). In the smooth case, this category agrees with the category of usual \({\mathcal D}_X\)-modules. He extends the notions and results of the smooth case to the singular case: coherence, holonomy, regularity, local algebraic cohomology, \(\otimes\), duality, direct and inverse images. De Rham function, Riemann-Hilbert correspondence. The key points are: 1) Kashiwara's theorem, giving the equivalence between the category of \(\mathcal D\)-modules over a complex submanifold \(M\) over \(Z\) satisfying \(\Gamma_{[Y]}(M)\simeq M\). 2) The fact that every complex analytic space is locally embeddable in a complex manifold. 3) The construction of a Grothendieck topology \({\mathcal C}(X)\) over \(X\) [cf. `Grothendieck topologies' by \textit{M. Artin}, Cambridge, Mass.: Harvard University (1962; Zbl 0208.48701) and `Notes on crystalline cohomology' by \textit{P. Berthelot} and \textit{A. Ogus}, Princeton, N.J.: Princeton University Press (1978; Zbl 0383.14010)], whose objects are the closed embeddings \(U\hookrightarrow V\) such that \(U\subset X\) is an open set and \(V\) is smooth, and whose morphisms are the morphisms of \(V\) such that their restrictions to \(U\) are the inclusions as subsets of \(X\). The objects of \(M(X,{\mathcal D})\) are defined locally for \({\mathcal C}(X)\), and the glueing is accomplished by the !-direct image functor. The proofs of the results in the paper depend only on the corresponding results in the smooth case, and the compatibility with the glueing data. Therefore, the main author's contribution is a convenient technical manipulation of these compatibilities. It seems possible to take the point of view of `Le théorème de comparaison entre cohomologies de De Rham d'une variété algébrique complexe et le théorème d'existence de Riemann' by \textit{Z. Mebkhout}, Publ. Math., Inst. Hautes Etud. Sci. 69, 47-89 (1989; Zbl 0709.14015), to be able to define the ``irregularity complexes'', and to avoid the use of the general theorem of desingularization of Hironaka.