1
$\begingroup$

Assume you have a sheaf morphism $f:\mathcal{F} \rightarrow \mathcal{G}$ and consider the sheafification of the presheaf $\textrm{im}(f)$. I want to prove that it's isomorphic to a sub sheaf of $\mathcal{G}$. I basically do this by showing that whenever there is an injective morphism of presheaves $\mathcal{F} \rightarrow \mathcal{G}$, the morphism induced on the sheafifications is still injective, hence I can conclude. It's basically Exercise 2.1.4 in Hartshorne. The inclusion $i :\textrm{im}(f)(U) \rightarrow \mathcal{G}(U)$ is injective for all $U$ so the argument applies to my situation. The thing is that the proof is pretty long in this way and I was wondering if there is a more direct way of proving it exploiting the specifics of this context rather than prove it for every injective morphism of presheaves. Thanks in advance.

$\endgroup$
  • $\begingroup$ Did you show part (a) in exercise 2.1.4 ? Because then you have a morphism of presheaves $im(f) \to G$ which is injective on each open set $U$, so by (a) you get an injective morphism of sheaves $im(f)^+ \to G^+ \cong G$. $\endgroup$ – Watson Oct 27 '18 at 9:12
  • $\begingroup$ Yes, but this is the path that I feel makes it too long. Proving part a relies on other exercises as well. Also, it requires proving that the stalks of the presheaf and the stalks of the sheafification are isomorphic. I was wondering if there's a way to exploit the fact that I'm talking about the image sheaf and an inclusion here, rather than a random presheaf and injective presheaf morphism $\endgroup$ – Dalamar Oct 27 '18 at 9:14
1
$\begingroup$

If you showed part (a) in exercise 2.1.4, then you are almost done for (b) ! And I don't think there is any easier way that this. It might look too long for you now, but in a few months, I believe that you will actually find this proof rather short !

Indeed, you have a morphism of presheaves $$\mathrm{im}(f) \to \mathscr G,$$ which is injective on each open set $U$, so by (a) you get an injective morphism of sheaves $$\mathrm{im}(f)^+ \to \scr G^+.$$

But $\scr G$ is a sheaf by assumption, so that $\scr G^+ \cong G$. Therefore, the image sheaf $\mathrm{im}(f)^+$ is isomorphic to a subsheaf of $\scr G$

$\endgroup$
  • $\begingroup$ Well then maybe there's a quicker way to prove part a)? Do you think it's worth it for me to ask another question about it? $\endgroup$ – Dalamar Oct 27 '18 at 9:25
  • $\begingroup$ @Dalamar : it is completely normal to find these kind of things a bit difficult at the beginning, while learning sheaf theory. But I really believe that to solve part a), you need to use that sheafification preserves the stalks. This is actually one of the main properties of sheafification ! :-) $\endgroup$ – Watson Oct 27 '18 at 9:30
  • $\begingroup$ I already proved that :). However after that there's a small problem for me. How can I say that a morphism on the stalks induces a morphism on the whole sheaf and take the final step? Is it just obvious? $\endgroup$ – Dalamar Oct 27 '18 at 9:51
  • 1
    $\begingroup$ If $F \to G$ is a morphism of presheaves, then composing with the natural map $G \to G^+$, you get a morphism $F \to G^+$ from a presheaf to a sheaf. By the universal property of sheafification, this morphism factors through a morphism of sheaves $F^+ \to G^+$. Is that answering your comment? $\endgroup$ – Watson Oct 27 '18 at 9:56
  • 1
    $\begingroup$ @Dalamar : see here ; if all the induced maps on the stalks are injective, then the morphism of sheaves is injective. $\endgroup$ – Watson Oct 27 '18 at 10:10

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.