6
$\begingroup$

I would like to understand this notion better; where could I find some examples? In particular, I am interested in the following questions (and references for the answers).

  1. Is a universal homeomorphism of connected regular (excellent finite dimensional) schemes an isomorphism if these schemes are not positive characterstic ones?

  2. Suppose that a finite morphism $f:X\to Y$ of connected regular (excellent finite dimensional) schemes is generically purely inseparable. Does $f$ restrict to a universal homeomorphism of some open (non-empty) subschemes of $X$ and $Y$?

$\endgroup$

1 Answer 1

8
$\begingroup$
  1. Yes. Let $f\colon X \to Y$ be a universal homeomorphism of locally noetherian schemes. Assume that $X$ and $Y$ are integral and $Y$ is normal, and the function field $k(Y)$ has characteristic 0. Then $k(X) = k(Y)$, and $f$ is an isomorphism by Zariski's main theorem.

  2. Under these conditions $f$ is a universal homemorphism. In fact $X$ is the normalization of $Y$ in $k(X)$, again by Zariski's main theorem. But $k(X)$ is purely inseparable over $Y$, so it is obtained by a successive extraction of $p^{\rm th}$ roots of 1. If $F \colon Y \to Y$ is the absolute Frobenius, some power of $F$ will factor through $X$, and then it is easy to see that $X$ is universally isomorphic to $Y$, since $F$ is a universal isomorphism.

$\endgroup$
2
  • $\begingroup$ For your answer to (1), would you mind giving a reference for the version of Zariski's main theorem that you're using? $\endgroup$
    – Will Chen
    Commented Dec 18, 2020 at 1:05
  • 1
    $\begingroup$ @WillChen It will be the version you can find in the Liu's book "Alg.Geo and Arith.Curves" as what Angelo wrote is basically the exercise 3.9 at the end of chapter on Zariski's theorem and such therein. $\endgroup$ Commented Apr 12, 2022 at 14:15

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .