nLab
good open cover

Context

Topology

Topos theory

topos theory

Background

Toposes

Internal Logic

Topos morphisms

Extra stuff, structure, properties

Cohomology and homotopy

In higher category theory

Theorems

Contents

Definition

Definition

A cover {U iX}\{U_i \to X\} of a topological space XX is called a good cover – or good open cover if it is

  1. an open cover;

  2. such that all the U iU_i and all their nonempty finite intersections are contractible topological spaces.

Remark

For XX a topological manifold one often needs that the non-empty finite intersections are homeomorphic to an open ball and for smooth manifolds often one needs that the finite non-empty intersections are diffeomorphic to an open ball.

In the literature this is traditionally glossed over, but it is in fact a subtle point, see the discussion at open ball and see below at Existence on paracompact manifolds.

Therefore it makes sense to explicitly say:

Definition

Given a smooth manifold XX a differentiably good open cover is a good open cover one all whose finite non-empty intersections are in fact diffeomorphic to an open ball, hence to a Cartesian space.

Properties

Exsistence on paracompact manifold

Proposition

Every paracompact manifold admits a good open cover, def. 1.

Proof

Every paracompact manifold admits a Riemannian metric, and for any point in a Riemannian manifold there is a geodesically convex neighborhood (any two points in the neighborhood are connected by a unique geodesic in the neighborhood, one whose length is the distance between the points; see for example the remark after lemma 10.3 (Milnor) page 59). It is immediate that a nonempty intersection of finitely many such geodesically convex neighborhoods is also geodesically convex, hence contractible.

Remark

It is apparently a folk theorem that every geodesically convex neighbourhood in a Riemannian manifold is diffeomorphic to a Cartesian space (for instance this is asserted on page 42 of (BottTu). This implies the following strengthening of the above result, which appears stated as theorem 5.1 in BottTu. But a complete proof in the literature is hard to find (see also the discussion of the references at ball). We give a complete proof below.

Proposition

Every paracompact manifold of dimension dd admits a differentiably good open cover, def. 2, hence an open cover such that every non-empty finite intersection is diffeomorphic to the Cartesian space d\mathbb{R}^d.

Proof

By (Greene) every paracompact manifold admits a Riemannian metric with positive convexity radius r convr_{\mathrm{conv}} \in \mathbb{R}. Choose such a metric and choose an open cover consisting for each point pXp\in X of the geodesically convex open subset U p:=B p(r conv)U_p := B_p(r_{conv}) given by the geodesic r convr_{conv}-ball at pp. Since the injectivity radius of any metric is at least 2r conv2r_{\mathrm{conv}} it follows from the minimality of the geodesics in a geodesically convex region that inside every finite nonempty intersection U p 1U p nU_{p_1} \cap \cdots \cap U_{p_n} the geodesic flow around any point uu is of radius less than or equal the injectivity radius and is therefore a diffeomorphism onto its image.

Moreover, the preimage of the intersection region under the geometric flow is a star-shaped region in the tangent space T uXT_u X: because the intersection of geodesically convex regions is itself geodesically convex, so that for any vT uXv \in T_u X with exp(v)U p 1U p n\exp(v) \in U_{p_1} \cap \cdots \cap U_{p_n} the whole geodesic segment texp(tv)t \mapsto \exp(t v) for t[0,1]t \in [0,1] is also in the region.

So we have that every finite non-empty intersection of the U pU_p is diffeomorphic to a star-shaped region in a vector space. By the results cited at ball (e.g. theorem 237 of (Ferus)) this star-shaped region is diffeomorphic to an n\mathbb{R}^n.

Coverages of good open covers

Corollary

The category ParaMfdParaMfd of paracompact manifolds admits a coverage whose covering families are good open covers.

The same holds true for subcategories such as

Proof

It is sufficient to check this in ParaMfdParaMfd. We need to check that for {U iU}\{U_i \to U\} a good open cover and f:VUf : V \to U any morphism, we get commuting squares

V j U i(j) V f U \array{ V_j &\to& U_{i(j)} \\ \downarrow && \downarrow \\ V &\stackrel{f}{\to}& U }

such that the {V iV}\{V_i \to V\} form a good open cover of VV.

Now, while ParaMfdParaMfd does not have all pullbacks, the pullback of an open cover does exist, and since ff is necessarily a continuous function this is an open cover {f *U iV}\{f^* U_i \to V\}. The f *U if^* U_i need not be contractible, but being open subsets of a paracompact manifold, they are themselves paracompact manifolds and hence admit themselves good open covers {W i,jf *U i}\{W_{i,j} \to f^* U_i\}.

Then the family of composites {W i,jf *U iV}\{W_{i,j} \to f^* U_i \to V\} is clearly a good open cover of VV.

Proposition

Every CW complex admits a good open cover.

Proof

It suffices to prove that if XX admits a good open cover and ϕ:S nX\phi: S^n \to X is an attaching map, then the pushout

S n ϕ X i D n+1 Y\array{ S^n & \stackrel{\phi}{\to} & X \\ i \downarrow & & \downarrow \\ D^{n+1} & \to & Y }

also admits a good open cover. Let {U α}\{U_\alpha\} be a good open cover of XX closed under nonempty finite intersections, and choose a contracting homotopy h α:I×U αU αh_\alpha: I \times U_\alpha \to U_\alpha such that h α(0,)=idh_\alpha(0, -) = id and h α(1,)h_\alpha(1, -) is constant. For any subset SD n+1S \subseteq D^{n+1}, let Hull(S)Hull(S) denote the convex hull of SS. Then, if VV is relatively open in the boundary S nS^n, Hull(V)Hull(V) is open in D n+1D^{n+1}. It follows that the image in YY of

V αU αHull(ϕ 1(U α))XD n+1V_\alpha \coloneqq U_\alpha \cup Hull(\phi^{-1}(U_\alpha)) \subseteq X \cup D^{n+1}

is open in YY, and it is contractible: define a contracting homotopy

H α:I×V αV αH_\alpha: I \times V_\alpha \to V_\alpha

by

H α(t,v)={(12t)v+2tvv 0t<12,vint(D n+1)V α v 0t12,vU α h α(2t1,ϕ(vv)) 12t1,vint(D n+1)V α h α(2t1,v) 12t1,vU α H_\alpha(t, v) = \left\{ \array{ (1 - 2t)v + 2t \frac{v}{|v|} & 0 \leq t \lt \frac1{2}, v \in int(D^{n+1}) \cap V_\alpha \\ v & 0 \leq t \leq \frac1{2}, v \in U_\alpha \\ h_\alpha(2t - 1, \phi(\frac{v}{|v|})) & \frac1{2} \leq t \leq 1, v \in int(D^{n+1}) \cap V_\alpha \\ h_\alpha(2t - 1, v) & \frac1{2} \leq t \leq 1, v \in U_\alpha } \right.

These sets V αV_\alpha together with int(D n+1)int(D^{n+1}) form a good open cover of YY.

nnPOV

The following nPOV perspective on good open covers gives a useful general “explanation” for their relevance, which also explains the role of good covers in Cech cohomology generally and abelian sheaf cohomology in particular.

Proposition

Let sPSh(CartSp) projsPSh(CartSp)_{proj} be the category of simplicial presheaves on the category CartSp equipped with the projective model structure on simplicial presheaves.

Let XX be a smooth manifold, regarded as a 0-truncated object of sPSh(C)sPSh(C).

Let {U iX}\{U_i \to X\} be a good open cover by open balls in the strong sense: such that every finite non-empty intersection is diffeomorphic to an d\mathbb{R}^d.

Then: the Cech nerve C({U})sPSh(C)C(\{U\}) \in sPSh(C) is a cofibrant resolution of XX in the local model structure on simplicial presheaves.

Proof

By assumption we have that C(U)C(U) is degreewise a coproduct of representables. It is also evidently a split hypercover.

This implies the statement by the characterization of cofibrant objects in the projective structure.

This has a useful application in the nerve theorem.

Notice that the descent condition for simplicial presheaves on CartSp at (good) covers is very weak, since all Cartesian spaces are topologically contractible, so it is easy to find the fibrant objects AsPSh(C) proj,locA \in sPSh(C)_{proj, loc} in the topological localization of sPSh(C) projsPSh(C)_{proj} for the canonical coverage of CartSp. The above observation says that for computing the AA-valued cohomology of a diffeological space XX, it is sufficient to evaluate AA on (the Cech nerve of) a good cover of XX.

We can turn this around and speak for any site CC of a covering family {U iX}\{U_i \to X\} as being good if the corresponding Cech nerve is degreewise a coproduct of representables. In the projective model structure on simplicial presheaves on CC such good covers will enjoy the central properties of good covers of topological spaces.

References

  • John Milnor, Morse theory , Princeton University Press (1963)
  • R. Greene, Complete metrics of bounded curvature on noncompact manifolds Archiv der Mathematik Volume 31, Number 1 (1978)

Revised on January 16, 2014 14:15:10 by David Roberts (129.127.252.10)