6
$\begingroup$

I have the following Grothendieck pretopologies on the category of schemes.

The first one, the covers are families of morphisms $\{ U_i \to X \}$ such that for every point $x \in X$ there exists some $i$ and some $u \in U_i$ over $x$ such that $[k(u):k(x)] = 1$ (notice I haven't asked anything else of the morphisms $U_i \to X$; they don't have to be étale, or even of finite type).

The second one, the covers are finite families of morphisms $\{ V_i \to X\}$ of finite type such that $\amalg V_i \to X$ is surjective and each $V_i \to X$ is an immersion (i.e. a composition of closed and open immersions).

If the schemes are noetherian they should give rise to the same category of sheaves.

Obvious choices in my mind for naming the first one is "completely decomposed" (for obvious reasons) or "discrete" (because every cover is refinable by the cover $\{ x \to X \}_{x \in X}$), but both of these exist already. The second I would call the "open-closed topology" but this exists as well.

Any suggestions? Does there already exists a name in the literature?

Improve this question
$\endgroup$
6
  • $\begingroup$ There is already a "completely decomposed" topology floating out there: en.wikipedia.org/wiki/Nisnevich_topology $\endgroup$
    – stankewicz
    Commented Apr 6, 2012 at 12:38
  • $\begingroup$ I guess I should make a note about Nisnevich's topology vs. your pretopology (and I'm not any sort of expert in Grothendieck topologies so caveat lector). Nisnevich explicitly requires his morphisms be etale so that his topology a) can be used for studying class groups of tori (or so I was told by Cristian Gonzalez-Aviles) and b) lies strictly between the etale topology and the zariski topology. Do you have any particular reason for not placing restrictions on your morphisms? $\endgroup$
    – stankewicz
    Commented Apr 6, 2012 at 12:44
  • 1
    $\begingroup$ The motivation for my definition is the following: I have a particular functor $F$ in mind. I can show that for every scheme $X$ in the category of schemes that I am working, the canonical morphism $F(X) \to \prod_{x \in X} F(x)$ is injective. The group on the right is precisely the sheafification of $F$ for the topology described in the question, and this language makes the statements and proofs much cleaner. For example stating this morphism is a monomorphism is just saying that $F$ is separated for the topology described in the question. $\endgroup$
    – name
    Commented Apr 7, 2012 at 8:56
  • $\begingroup$ In the five years since stating the question: Did you settle on a name? Which functor $F$ did you have in mind? Are there some notes where I can read up on your observations? $\endgroup$
    – Ingo Blechschmidt
    Commented Nov 22, 2016 at 19:14
  • $\begingroup$ @IngoBlechschmidt (i) See arxiv.org/abs/1305.5349 Section 3.5. (ii) See Lemma 3.6.1 and Example 3.6.6. (iii) See (i). $\endgroup$
    – name
    Commented Jun 1, 2018 at 11:18

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.