14
$\begingroup$

Recall that an open cover $\mathfrak{U} = \{ U_\alpha \}$ of a manifold $M$ is called a good cover if all possible finite intersections $U_{\alpha_1} \cap ... \cap U_{\alpha_n}$ are contractible.

Question: What is the minimum number of open sets required for a good cover of the 2-dimensional torus?

The picture below provides a good cover of the torus (i.e. opposite sides of the parallelogram identified as usual) using 7 open sets (i.e. take sufficiently small open neighbourhoods of the hexagons). Can one do any better than 7? If not, how does one prove that 7 is optimal?

Good cover of the torus with 7 open sets (neighbourhoods of the hexagons)

$\endgroup$
1
  • 2
    $\begingroup$ Would be nice to see a 3d rendering of that ... $\endgroup$
    – J.J. Green
    Commented Apr 16, 2020 at 10:12

1 Answer 1

22
$\begingroup$

You can't do any better than $7$. This follows from

Karoubi, Max; Weibel, Charles A., On the covering type of a space, Enseign. Math. (2) 62, No. 3-4, 457-474 (2016). ZBL1378.55002.

in particular Theorem 5.3 in the arXiv version.

The strict covering type of a space $X$ is the minimal cardinality of a good cover, denoted $\operatorname{sct}(X)$. This is not a homotopy invariant, so Karoubi and Weibel introduce the covering type, defined by $$ \operatorname{ct}(X)=\min\{\operatorname{sct}(X')\mid X'\simeq X\} $$ Obviously $\operatorname{ct}(X)\leq \operatorname{sct}(X)$. In Theorem 5.3 they use cohomological arguments to show that $\operatorname{ct}(T^2)=7$.

$\endgroup$
4
  • $\begingroup$ Great, thanks for the reference! $\endgroup$
    – Dennis
    Commented Apr 16, 2020 at 10:57
  • 6
    $\begingroup$ Noob question: is this at all related to the fact that all maps on a torus are $7$-colorable but some are not $6$-colorable? $\endgroup$ Commented Apr 16, 2020 at 21:19
  • $\begingroup$ @GregMartin: Good question. I don't know of a relationship between covering type and chromatic numbers of embedded graphs, but equally I don't know enough to say that such a thing doesn't exist. $\endgroup$
    – Mark Grant
    Commented Apr 21, 2020 at 9:07
  • $\begingroup$ Doesn't the linked paper state the relation of chromatic numbers and covering types on the first page? Or is there some subtle technical detail I missed? $\endgroup$
    – Trebor
    Commented Oct 31, 2023 at 10:13

You must log in to answer this question.

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