4
$\begingroup$

Let $(X,d)$ be an $n$-dimensional $(n< \infty)$ complete geodesic metric space, where any two points in $X$ are joined by a unique shortest geodesic. Let $S$ be a sufficiently small metric $(n-1)$-sphere in $X$. Let $\epsilon>0$ be the radius of $S$. Pick a point $p \in S$. Find a metric $(n-1)$-sphere $S'$ of radius $0<r < \epsilon/2$ around $p$. A $k$-dimensional metric sphere $S^k$ in $(X,d)$ of radius $r$ centered at $p$ is given by a $k$-dimensional subset of $(X,d)$: $S^k \subset \{x\in X|d(p,x) = r\}$. Here dimension I'm referring to covering dimension.

Is the intersection $S \cap S'$ always an $(n-2)$-dimensional metric sphere?

$\endgroup$
12
  • 2
    $\begingroup$ Just to clarify - what is a k-dimensional metric sphere in a complete geodesic metric space? I believe the answer is "no" regardless of anything, but this clarification is needed. $\endgroup$ May 29, 2020 at 15:29
  • $\begingroup$ If you have some notion of k-dimensional sphere in mind such that there is a family with ((n-k)k + n + 1) parameters (say, emitting geodesic segments of length r along some k-dimensional subspace in the tangent space to some point), then the answer is no by dimension counting. $\endgroup$ May 29, 2020 at 15:32
  • 1
    $\begingroup$ You still haven't defined what you mean by sphere, so that it's meaningful for the intersection. A sphere in a metric space (which metric space)? A topological sphere? $\endgroup$
    – YCor
    May 29, 2020 at 16:30
  • 1
    $\begingroup$ @Lev: I think that it is a typo, the OP probably meant $S^k \subset \ldots$ instead of $S^k =$. // But as stated: obviously $S\cap S'$ is a subset of $S$ and so is a subset of some sphere. So the OP's question essentially boils down to whether the intersection has dimension $(n-2)$.... $\endgroup$ May 29, 2020 at 17:03
  • 2
    $\begingroup$ It's not at all obvious (to me at least) why a metric sphere in a geodesically complete metric space of Lebesue dimension $n$ must always have Lebesgue dimension $n-1$. $\endgroup$
    – Mark Grant
    May 29, 2020 at 20:19

2 Answers 2

3
$\begingroup$

Let $C_\alpha$ denote the $n$-dimensional closed Euclidean solid cone with the cone angle $\alpha\in (0, \pi)$. Let $X_\alpha$ be the metric space obtained by gluing two copies $C^\pm_\alpha$ of $C_\alpha$ at their tips, and equipped with the natural path-metric. Let $o\in X_\alpha$ denote the common tip of the cones. I will use it as the center of the first sphere $S=S(o,1)$; I will use $\epsilon=1$. For every point $x\in S$, the sphere $S'=S(x,r)$ is $n-1$-dimensional. However, for each $r\in (0,1)$, for all sufficiently small $\alpha$, for every $x\in S\subset X_\alpha$, the intersection $S'\cap S$ is empty, hence, has dimension $-1$, not $n-2$. By modifying this construction one can build uniquely geodesic spaces where spheres as in your question are $n-1$-dimensional but have arbitrary dimension of their intersection, between $-1$ and $n-1$.

Edit. Here is a generalization, to ensure that the spheres $S, S'$ are connected (and locally path-connected) and the intersection $S\cap S'$ is nonempty.

Let $Y$ be a closed solid cone in the Euclidean space $E^k, 1\le k<n$, with the tip $o$. I will glue $Y$ to the space $X_\alpha$ as above so that $o$ is identified with the common tip $o\in X$ and two boundary rays of $Y$ are identified with geodesic rays in the two cones $C^\pm_\alpha \subset X$. Let $Z$ denote the resulting path-metric space. It is easy to check (say, using Reshetnyak gluing theorem) that $Z$ is a $CAT(0)$ space, hence, is uniquely geodesic.

Moreover, both spheres $S, S'\subset Z$ (defined as before) are connected and locally path-connected. Furthermore, for every $r$, there is $\alpha$ such that $S\cap S'$ is $(k-2)$-dimensional. (Note that $k-2<n-2$.)

By working a bit harder, one can modify this construction so that spheres are still connected and locally path-connected, while $S\cap S'$ is $(n-1)$-dimensional, where $n$ is the dimension of the ambient space.

$\endgroup$
2
  • $\begingroup$ What if the metric spheres are connected and locally path-connected? $\endgroup$
    – Shijie Gu
    May 30, 2020 at 13:53
  • $\begingroup$ @Bobbrown: See the edit. $\endgroup$ May 31, 2020 at 21:43
0
$\begingroup$

Even for generic Riemannian metrics, the answer to the title question is no.

Consider the metric $ds^2=dx^2 + (2-x^2)dy^2+ dz^2$.

The geodesic equations for this metric are \begin{align} x'' &= -x y'^2 \\ y'' &= 2x x' y'/(2-x^2) \\ z'' &= 0 \end{align}

The geodesic that begins at $(-1,0,0)$ with initial time-derivative $(1.00,1,0)$ arrives at $(0,0.54,0)$ after distance $1.21$.

The geodesic that begins at $(-1,0,0)$ with initial time-derivative $(1.47,0,1)$ arrives at $(0,0,0.68)$ after distance $1.21$.

So the spheres of radius $1.21$ from $(\pm1,0,0)$ intersect in a topological circle that includes the four points $(0,\pm 0.54,0)$ and $(0,0,\pm 0.68)$.

The first two points are diametrically opposite on that topological circle, and they are $0.76$ units from the origin.

The second two points are also diametrically opposite on that topological circle, and they are $0.68$ units from the origin.

So the topological circle of intersection is not a metric circle, and the same would be true even for the intersections of spheres with smaller radii.

$\endgroup$
2
  • $\begingroup$ Is there a similar example when one metric sphere centered on the other like I described in the post? $\endgroup$
    – Shijie Gu
    May 30, 2020 at 14:24
  • $\begingroup$ Surely yes, but I’d figure that out more easily if the description can be updated with some line breaks or a diagram or more parallel notation. $\endgroup$
    – user44143
    May 30, 2020 at 22:02

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.