0

A smooth manifold $M$ can be defined as a pair (topological manifold $X$, smooth, as in $C^{\infty}$, atlas $\mathscr M$), where a topological manifold is defined as a locally Euclidean Hausdorff 2nd countable space $X$ without regard to any continuous, as in $C^0$, atlas.

I want to know if I can define a smooth manifold $M$ instead as a pair (locally Euclidean Hausdorff 2nd countable space $X$, smooth atlas $\mathscr M$), and then a topological manifold $Y$ as a pair (locally Euclidean Hausdorff 2nd countable space $X$, $C^0$ atlas $\mathscr X$).

I mean, if I define a smooth manifold $M$ as a pair (topological manifold $Y$, smooth atlas $\mathscr M$), where $Y$ in turn is (locally Euclidean Hausdorff 2nd countable space $X$, continuous atlas $\mathscr X$) ... then does $M$ really does care about $\mathscr X$ ?

If so, then please give an example of smooth manifolds with the same smooth atlas (and topological space) but different continuous atlases and then they're indeed not identical or diffeomorphic or whatever? (I forgot the term, but I think it's like ... they're not considered smooth submanifolds of each other. Oh I think the term is 'distinct' or not 'identical' in the sense that the identity map is not the diffeomorphism or something. I think that's the 'equal underlying set' version of 'if the inclusion map is not a smooth immersion & topological embedding or something'.)

Like $M_1$ = ((locally Euclidean Hausdorff 2nd countable space $X$, continuous atlas $\mathscr X_1$), smooth atlas $\mathscr M$)

and

$M_2$ = ((locally Euclidean Hausdorff 2nd countable space $X$, continuous atlas $\mathscr X_2$), smooth atlas $\mathscr M$)

with $\mathscr X_1 \ne \mathscr X_2$ but then $M_1$ & $M_2$ are not diffeomorphic or whatever.

BCLC
  • 14,197

1 Answers1

5

Part of the definition of a smooth manifold is that the topology induced by the smooth atlas must agree with that of the underlying continuous atlas. Moreover, unlike differential structures, a topological space can't have multiple non-isomorphic structures as a topological manifold, essentially by definition. In fancier terms, the category of topological manifolds embeds fully faithfully into the category of topological spaces since morphisms between topological manifolds are simply the continuous maps.

CJ Dowd
  • 2,898
  • Oh right yeah... thanks. ok got it as well from here. as for unique.... next question : here but i think i kinda get it now... – BCLC Jun 13 '24 at 18:30