2
$\begingroup$

Is there a simple description for $D$ modules over $\text{Spec}\left(k\left[x,y \right] / \left(xy\right) \right)$ (say k is algebraically closed of characteristic 0)? Is there a description of the $D$-module pushforward from its normalization?

$\endgroup$
1
  • 3
    $\begingroup$ The ($\infty$-categorical) theory of D-modules satisfies descent along proper surjective maps (see Gaitsgory-Rozenblyum “Crystals and D-modules”). For a nodal curve this implies that a D-module is equivalently a D-module on the normalization with identifications on the nodes $\endgroup$
    – Exit path
    Commented Feb 27, 2023 at 21:18

1 Answer 1

1
$\begingroup$

If $C$ and $D$ are the two branches of the nodal curve, and $p\in C$ and $q\in D$ are the points that are identified to form $X=C\cup D$, then I think that a $\mathcal D$-module on $X$ consists of:

• A vector bundle $V$ with flat connection over $C$.
• A vector bundle $W$ with flat connection over $D$.
• A vector space $U$.

• A injective linear map $V\otimes \Omega^1_C|_{\{p\}}\to U$.
• A injective linear map $W\otimes \Omega^1_D|_{\{q\}}\to U$.

$\endgroup$
2
  • 2
    $\begingroup$ Why not general $\mathcal D$-modules on $C$ and $D$? $\endgroup$
    – Will Sawin
    Commented Feb 27, 2023 at 20:56
  • 2
    $\begingroup$ As @WillSawin points out, I think this is incorrect as stated. For example, how would one obtain a skyscraper D-module at a smooth point using this description? $\endgroup$
    – Exit path
    Commented Feb 27, 2023 at 21:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .