5
$\begingroup$

Show that the map of graded rings $k[w, x, y, z] \rightarrow k[s, t]$ given by $(w, x, y, z) \mapsto (s^3, s^2t, st^2, t^3)$ induces a closed embedding $\mathbb{P}_k^1 \rightarrow \mathbb{P}_k^3$, which yields an isomorphism of $\mathbb{P}_k^1$ with the twisted cubic.

I think I need to use the following exercise: If $S \rightarrow R$ is a surjection of graded rings, then the domain of the induced morphism is $Proj (R)$, and the induced morphism $Proj (R) \rightarrow Proj (S)$ is a closed embedding.

However, I don't see how the map $(w, x, y, z) \mapsto (s^3, s^2t, st^2, t^3)$ is surjective.

$\endgroup$
  • $\begingroup$ Did you take a look at this post? $\endgroup$ – Liam Mar 29 at 2:36
2
$\begingroup$

You're correct that the map isn't surjective. The point is that this map has image in degrees divisible by $3$, and so factors through the associated Veronese subring of $k[s, t]$ composed of the pieces of degree divisible by $3$, onto which it is surjective. In precise language:

Let $S = k[x, y, z, w], R = k[s, t]$, and let $\varphi \colon S \to R$ be the morphism of graded rings described above. Let $R' := \bigoplus_{j = 0}^{\infty} R_{3j}$ be the subring of $R$ consisting of the components of $R$ with degree divisible by $3$. The morphism of graded rings $i \colon R' \to R$ induces an isomorphism $\tilde{i} \colon \mathrm{Proj}(R) \to \mathrm{Proj}(R')$; this is, for instance, Exercise 6.4H of Vakil's Foundations of Algebraic Geometry.

Note that $\varphi$ takes image in $R'$, so letting $\varphi'$ be the map $S \to R'$ defined exactly as $\varphi$ is, we have that $\varphi = i \circ \varphi'$. Since $\varphi'$ is surjective, the induced map $\tilde{\varphi'} \colon \mathrm{Proj}(R') \to \mathrm{Proj}(S)$ is a closed embedding. Hence, since $\tilde{\varphi} \colon \mathrm{Proj}(R) \to \mathrm{Proj}(S)$ is the composition of an isomorphism ($\tilde{i}$) with a closed embedding ($\tilde{\varphi'}$), it is itself a closed embedding, as desired.

$\endgroup$
  • $\begingroup$ When you say "Since $\varphi$ is surjective", should it be $\varphi'$? $\endgroup$ – Smash Mar 29 at 14:27
  • $\begingroup$ @Smash: yes, thank you! $\endgroup$ – Alex Wertheim Mar 29 at 20:00

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.