The Theorem 1.9.5 in T. A. Springer's Linear Algebraic Groups (page 19) states that:

Let $\phi: X \rightarrow Y$ be a morphism of varieties. Then $\phi X$ contains a non-empty open subset of its closure $\overline{\phi X}$.

(Here, a variety is defined as a quasi-compact ringed space to have an open cover by affine subsets along with being separated.)

The first line of Springer's proof wrotes:

Using a covering of $Y$ by affine open sets, we reduce the proof to the case that $Y$ is affine.

I wonder how such reduction proceeds, particularly whether the preimage of a affine open subset of $Y$ is a subvariety of $X$. (Initially I wrongly claim that the preimage of a affine open subset of $Y$ is not necessarily quasi-compact.)

Hope for an answer, thanks in advance!


Actually, the map $\phi:X\to Y$ is quasi-compact, so the preimage of any quasi-compact set (like an affine open set of $Y$) is again quasi-compact. The key result is the following:

Lemma (Stacks Project 03GI): Let $f:X\to Y$ and $g:Y\to Z$ be morphisms of schemes. If $g\circ f$ is quasi-compact and $g$ is quasi-separated, then $f$ is quasi-compact.

In our situation, we take $Z=\operatorname{Spec} k$, $g$ the structure morphism, $f=\phi$, and then we have that $g\circ\phi$ is equal to the structure morphism $X\to \operatorname{Spec} k$ by assumption that $\phi$ is a map of $k$-schemes. As a separated morphism is quasi-separated, we have that $g$ is quasi-separated, and because $X$ is quasi-compact and $\operatorname{Spec} k$ is a point, we have that $g\circ f$ is quasi-compact. Thus your situation satisfies the assumptions of the lemma and we may conclude that $\phi$ is quasi-compact.

(I do not have a copy of this book in front of me right now, but I remember that some of these algebraic groups texts are notorious for not doing things correctly from a modern algebraic-geometry perspective. If it's doing some wacky thing like only taking closed points, do not fear: these geometric objects embed as subspaces with the induced topology of the schemes in question, and the maps between the not-quite-schemey versions are exactly the restrictions of the scheme maps to these subspaces and everything is okay.)

  • $\begingroup$ May I ask that utilizing the method above, whether we can prove the preimage of a subvariety is still a subvariety? $\endgroup$ – Wembley Inter Oct 23 '20 at 8:32
  • $\begingroup$ @WembleyInter What do you mean? What parts of that claim are tripping you up? This shows quasi-compactness, and any locally closed subscheme of a separated scheme is again separated, so that would seem to do it by your definitions. $\endgroup$ – KReiser Oct 23 '20 at 9:39
  • $\begingroup$ Ah... Now I get it. Thanks! $\endgroup$ – Wembley Inter Oct 23 '20 at 11:01
  • $\begingroup$ @WembleyInter if this resolves your issues, please consider accepting the answer. If not, is there something you feel is missing from the answer? $\endgroup$ – KReiser Oct 23 '20 at 22:12

Actually, $\phi$ is a quasi-compact map.

Indeed, let $B$ be an affine base scheme, so that $X,Y$ are separated quasicompact over $B$. Then $\phi$ is the composition of $\phi_1: X= X \times_Y Y \rightarrow X \times_B Y$ and $\phi_2: X \times_B Y \rightarrow Y$.

$Y \rightarrow B$ is separated so $\phi_1$ is a closed immersion so is quasi-compact. But $\phi_2$ is a base change of $X \rightarrow B$ which is a quasi-compact map so is quasi-compact. So the composition $\phi$ of $\phi_1$ and $\phi_2$ is quasi-compact.


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.