2
$\begingroup$

I've had problem working on this exercise from the July 31, 2023 version of Vakil's The Rising Sea:

15.6.G. Exercise. Show that the map of graded rings $S_\bullet \to \Gamma_\bullet\widetilde{S_\bullet}$ induces an isomorphism $\operatorname{Proj} \Gamma_\bullet\widetilde{S_\bullet} \cong \operatorname{Proj} S_\bullet$, and under this isomorphism, the respective $\mathscr{O}(1)$'s are identified.

It could be easily seen that for homogeneous $f \in S_n (n>0)$, we have isomorphisms on $D_+(f)$ in these two projective schemes that glue nicely. However, I could not prove $\Gamma_\bullet\widetilde{S_\bullet}$ is finitely generated in degree 1, thus I could not show $\mathscr{O}(1)$ on $\operatorname{Proj} \Gamma_\bullet\widetilde{S_\bullet}$ is an invertible sheaf. Nor could I prove this map of graded rings induces a map of projective schemes defined on all of $\operatorname{Proj} \Gamma_\bullet\widetilde{S_\bullet}$, since a general map of graded rings only induce a morphism on an open subscheme. How could I prove these two statements?

$\endgroup$
2
  • 1
    $\begingroup$ I've made a number of typographical improvements to your post, like using $\operatorname{Proj} R$ to produce $\operatorname{Proj} R$ instead of the \mathrm and \space version you used before, and using \bullet for the subscript instead of a period. Please also note that since Vakil's text is still in preparation with new versions coming all the time, it's helpful to add a link to the version you're using in the post - the first time I went to look at the surrounding material, I got an older version which didn't even have an exercise 15.6.G. $\endgroup$
    – KReiser
    Commented Oct 26, 2023 at 23:48
  • 1
    $\begingroup$ Sorry, I had an answer up but it fell short and I couldn't really find a good way to fix it. Your concern about showing that $V(\phi(S_+))=\emptyset$ is good and valid, and I can't think of a way to show that without quite a bit of work, some of which is done later (or not at all) in the text. See Hartshorne theorem II.5.19 and exercise II.5.9 for the result in the case when $S_0$ is a field, but it's not exactly easy - it requires most of the content of Serre's FAC to do this. $\endgroup$
    – KReiser
    Commented Oct 30, 2023 at 6:17

0

You must log in to answer this question.