7
$\begingroup$

$\mathbb{A}^2\setminus (0,0)$ is often given as an example of a variety that is not affine. I am trying to understand this example better by seeing it as a special case of a natural general claim.

Claim: Let $X$ be an affine variety over an algebraically closed field $k$ such that $k[X]$ is a UFD. Let $Y$ be any closed subset of codimension at least 2. Then $X\setminus Y$ is not affine.

Is this the right general claim?

If so, can I drop the assumption that $k$ is algebraically closed, or is that needed?

$\endgroup$
  • 2
    $\begingroup$ If you drop the assumption that $k$ is algebraically closed, then you have to be more clear about what you mean by "affine variety", what you mean by morphism, etc. $\endgroup$ – Zhen Lin Feb 14 '14 at 17:04
  • $\begingroup$ You would probably be better off doing this in terms of schemes if you want to get rid of $k$ being algebraically closed. $\endgroup$ – Tobias Kildetoft Feb 14 '14 at 18:12
  • $\begingroup$ @ZhenLin - good point. $\endgroup$ – Ben Blum-Smith Feb 16 '14 at 23:43
7
$\begingroup$

I think the right generalization to you your example is the following.

Let $X$ be an integral affine, Notherian normal scheme, and $F$ a (EDIT: Non-empty--thanks Cantlog) closed subset with $\text{codim}_X(F)\geqslant 2$. Then, $X-F$ is a non-affine open subscheme.

Indeed, using "Algebraic Hartog's Lemma" one can show that the inclusion $(X-F)\hookrightarrow X$ induces an isomorphism $\mathcal{O}_X(X)\to\mathcal\to\mathcal{O}_{X}(X-F)$. If $X-F$ were affine, this would contradict the equivalence of categories $\mathsf{Sch}^{\text{op}}\xrightarrow{\approx}\mathsf{Ring}$.

$\endgroup$
  • 1
    $\begingroup$ One must suppose $F$ is non-empty. The statement is true without assuming $X$ is affine. $\endgroup$ – Cantlog Feb 14 '14 at 21:42
  • $\begingroup$ @Cantlog I see that you're a scholar of the empty set :) For the general case one must make cohomological considerations, right? I was trying to avoid making the reasoning too sophisticated--or is there a more elementary way to prove this? $\endgroup$ – Alex Youcis Feb 14 '14 at 23:08
  • $\begingroup$ This is because I often forget to check with such a special case myself :). The general case is an easy consequence of the affine case, just consider the codimension at the generic points of $F$. $\endgroup$ – Cantlog Feb 15 '14 at 18:40

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.