Let $f : X \to Y$ be an open faithfully flat morphism of schemes. In the text I'm reading (Angelo Vistoli's notes on descent) it is claimed that then every point $x \in X$ admits an open neighorhood $U$ such that $f(U)$ is open and the morphism $U \to f(U)$ is quasi-compact. The latter is one of the possible definitons of a fpqc morphism. However, I don't understand at all why $U \to f(U)$ should be quasi-compact.

This is needed to prove that the fpqc topology is finer than the fppf topology.

  • $\begingroup$ Well, if you take for $U$ any open affine neighbourhood of $x$ then $f(U)$ is open because $f$ is, and $U\to f(U)$ is quasi-compact because $U$ is... $\endgroup$ – Matthieu Romagny Sep 18 '10 at 11:46
  • $\begingroup$ I don't understand. There are morphisms $f : X \to Y$, which are not quasi-compact, although $X$ is affine and $Y$ is quasi-compact. For example, take some affine $X$ which has an open subset $U$ which is not quasi-compact, and glue two copies of $X$ along $U$ to get $Y$. Then the, say first, inclusion $X \to Y$ is not quasi-compact. $\endgroup$ – Martin Brandenburg Sep 18 '10 at 12:03

[Edit] (add some details).

Replacing $Y$ with an affine open neighborhood $V$ of $f(x)$ (and $X$ with $f^{-1}(V)$), one can suppose that $Y$ is affine. Cover $X$ by affine open subsets {$U_i$}$_i$. As $Y$ is quasi-compact, a finite number of the $f(U_i)$ cover $Y$. If necessarily, we can add one more $U_i$ so $x$ belong to one of these $U_i$'s. The union $U$ of these (finitely many) $U_i$ is quasi-compact, and we have $f(U)=Y$, $x\in U$. The morphism $f|_U : U\to Y$ is a morphism from a quasi-compact scheme to an affine scheme, so it is quasi-compact because for any affine open subset $V$ of $Y$, $(f|_U)^{-1}(V)\cap U_i= V\times_Y U_i$ is affine.

  • $\begingroup$ hopefully this answers 1,2; though it's mostly just a re-hash of Liu's answer 1.if $x \not \in U$ then as Liu says just add pick a open affine $U_0$ containing $x$ and replace $U$ with $U \cup U_0$. 2.Any q-compact open $V$ in $Y$ is covered by a finite number of affine opens $V_i$ so $f|_U^{-1}(V)$is covered by the finitely many $f|_U^{-1}(V_i)$ so it suffices to show $f|_U^{-1}(V_i)$ is q-compact; Liu shows $f|_U^{-1}(V_i)$ is covered by a finite number of affines hence quasi-compact. $\endgroup$ – solbap Sep 18 '10 at 20:54

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.