Let $G$ be a commutative algebraic group stack over $\mathbb{F}_q$ (I don't really care about the precise definition: I'm secretly thinking about the Picard stack of a projective curve). To what extent do the standard results on the Lang isogeny $x \mapsto \text{Fr}_q(x) \cdot x^{-1}$ carry over from the theory of group schemes over $\mathbb{F}_q$? For example, I don't see any reason for this map to be representable. But I would like to have a sense in which it is "etale with group $G(\mathbb{F}_q)$." Can this be done?

Edit: I think that for the Picard stack $\text{Pic}$ of a projective curve, the fiber of the Lang map $\text{Pic} \to \text{Pic}$ over the trivial line bundle is identified with $$\coprod_{\text{Pic}(X)} \bullet / \mathbb{F}_q^{\times}$$ (here $\bullet = \text{Spec } \mathbb{F}_q$ and the disjoint union is taken over the Picard group of isomorphism classes). So the Lang map indeed fails to be representable, but my intuition suggests that it is "smooth of relative dimension zero." Unfortunately, my technical facility with stacks is inadequate to formulate a statement along these lines.

  • $\begingroup$ Have you tried looking at the various proofs for algebraic groups? If so, where to they go wrong? $\endgroup$ – anon May 20 '13 at 16:22
  • $\begingroup$ Well, one problem is that if this map is not representable then I don't know what it means for it to be etale. I think it is possible to define "smooth of relative dimension 0" for non-representable maps, but I am an amateur in this area and have no idea how to prove something like this. $\endgroup$ – Justin Campbell May 20 '13 at 16:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.