There is one point I don't understand about G-torsor.

  • Thread starter Thread starter kakarukeys
  • Start date Start date
  • Tags Tags
    Point
kakarukeys
Messages
187
Reaction score
0
There is one point I don't understand about G-torsor.

A Lie group G acts freely and transitively on a manifold F.
F x G -> F
(f, g) -> fg, f(g1g2) = (fg1)g2
is a smooth map.

fix an element f of F
then the map h
F -> G
fg -> g
is a homeomorphism.

I know h is open from the continuity of the map
{f} x G -> F
g -> fg

How to see h is continuous?
 
Last edited:
Physics news on Phys.org
please help! I'm desperate. I will help out in the Homework and coursework questions section if anybody can help me.
 
Your description of the map h: F -> G seems to have a typo. I guess you meant it to be f -> fg.

h is factored through Stab(g), right? Now Stab(g) is trivial and h is surjective since F is a G-torsor. Thus h is one-to-one.

By definition h is continuous, so h is bijective and continuous. Since in addition F and G are locally compact and Hausdorff, h is a homeomorphism.
 
kakarukeys said:
There is one point I don't understand about G-torsor.

A Lie group G acts freely and transitively on a manifold F.
F x G -> F
(f, g) -> fg, f(g1g2) = (fg1)g2
is a smooth map.

fix an element f of F
then the map h
F -> G
fg -> g
is a homeomorphism.

I know h is open from the continuity of the map
{f} x G -> F
g -> fg

How to see h is continuous?

No, there is no typo, I have typed a little too fast. Let me use Latex and state my question clearer.

There is one point I don't understand about G-torsor.

A Lie group G acts freely and transitively on a manifold F.
\rho: F \times G \longrightarrow F
\rho(f, g) = fg
f(g_1g_2) = (fg_1)g_2

fix an element f of F
then the map
h_f: \{fg | \forall g\in G\} \longrightarrow G
h_f(fg) = g
is a homeomorphism.

I know h_f is open from the continuity of the map
\rho_f = h_f^{-1}
\rho_f: \{f\} \times G \longrightarrow F
\rho_f(g) = fg

How to see h is continuous?

Your h is my \rho_f. Were you saying \rho_f is open because F, G are (required to be) locally compact and Hausdorff?
 
I couldn't find any theorem which guarantees that.

closests two are:

(1) if G is compact and F is Hausdorff, \rho_f is open
(2) if G is locally compact, F is locally compact and Hausdorff, F is a topological group under the induced group operations, \rho_f is open
 

Similar threads

Replies
4
Views
1K
Replies
9
Views
2K
Replies
1
Views
2K
Replies
13
Views
595
Replies
1
Views
3K
Replies
5
Views
2K
Back
Top