Given a sheaf $\mathscr F$ on $X$, how does one show that its sheafification (in the sense of Hartshorne) is isomorphic to it? The most obvious thing to do is a universal property argument: since $\mathscr F$ is already a sheaf, to the identity $id_{\mathscr F}:\mathscr{F\to F}$ there corresponds a unique morphism of presheaves (and thus sheaves) $\psi:\mathscr{F^+\to F}$ such that $$\psi\circ\theta=id_{\mathscr F}\tag{1}$$ where $\theta:\mathscr{F\to F^+}$ is canonical. So $\psi$ has a left-inverse. In the category of sheaves do I need to do extra work to show that $\psi$ is right-invertible, or do I get this for free (that is, without referencing the explicit construction) from $(1)$?


$\theta \circ \psi\colon \mathscr{F}^+ \to \mathscr{F}^+$ is a morphism such that $(\theta \circ \psi) \circ \theta = \theta \circ (\psi \circ \theta) = \theta \circ \operatorname{id}_\mathscr{F} = \theta$. $\operatorname{id}_{\mathscr{F}'}$ is another such morphism and you can apply uniqueness.

  • $\begingroup$ Thanks. So the conclusion is that $f\circ g=id$ implies $g\circ f=id$ in any category? $\endgroup$ – theage Aug 5 '15 at 19:06
  • 1
    $\begingroup$ @theage That's definitely not what I'm saying! Think about the usual characterizations of injective/surjective functions in the category of sets. $\endgroup$ – Hoot Aug 5 '15 at 19:12
  • $\begingroup$ Oh, yes I see. Sorry for the rather stupid generalization. $\endgroup$ – theage Aug 5 '15 at 19:27
  • $\begingroup$ I think the takeaway is just that initial objects (what's the category?) are unique up to unique isomorphism. $\endgroup$ – Hoot Aug 5 '15 at 23:59

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.