Problem 1.7 (Stereographic projection)

Let N denote the north pole ( 0 , , 0 , 1 ) 𝕊 n n + 1 , and let S denote the south pole ( 0 , , 0 , 1 ) . Define the stereographic projection σ : 𝕊 n { N } n by

σ ( x 1 , , x n + 1 ) = ( x 1 , , x n ) 1 x n + 1 .

Let σ ~ ( x ) = σ ( x ) for x 𝕊 n { S } .

(a)
For any x 𝕊 n { N } , show that σ ( x ) = u , where ( u , 0 ) is the point where the line through N and x intersects the linear subspace where x n + 1 = 0 (Fig. 1.13). Similarly, show that σ ~ ( x ) is the point where the line through S and x intersects the same subspace. (For this reason, σ ~ is called stereographic projection from the south pole.)
(b)
Show that σ is bijective, and σ 1 ( u 1 , , u n ) = ( 2 u 1 , , 2 u n , | u | 2 1 ) | u | 2 + 1 .

(c)
Compute the transition map σ ~ σ 1 and verify that the atlas consisting of the two charts ( 𝕊 n { N } , σ ) and ( 𝕊 n { S } , σ ~ ) defines a smooth structure on 𝕊 n . (The coordinates defined by σ or σ ~ are called stereographic coordinates.)
(d)
Show that this smooth structure is the same as the one defined in Example 1.31.

Answers

(a)
Fix an arbitrary x 𝕊 n { N } . A line passing through N and x is given by
L ( N , x ) : n + 1 , t N + t ( N x )

or in vector notation,

L ( N , x ) ( t ) = [ 0 0 1 ] + t ( [ 0 0 1 ] [ x 1 x n x n + 1 ] ) = [ t x 1 t x n 1 + t ( 1 x n + 1 ) ]
Figure 1: Stereographic projection from the north pole in case of 𝕊 1 .

Where does L exactly cross the x n + 1 -axis? This happens at the value t 0 for which the n + 1 th component of L ( t 0 ) is zero. To derive it, we set

1 + t 0 ( 1 x n + 1 ) = ! 0 t 0 = 1 1 x n + 1

which results in

L ( t 0 ) = [ x 1 1 x n + 1 , , x n 1 x n + 1 , 0 ] = [ σ ( x ) , 0 ] .

Therefore, the intersection point is indeed ( σ ( x ) , 0 ) .

(b)
We compute the formula for σ σ 1 and σ 1 σ . Before we start, we derive a formula for σ 2 :

| σ ( x ) | 2 = ( x 1 ) 2 + + ( x n ) 2 ) ( 1 x n + 1 ) 2 = ( x 1 ) 2 + + ( x n ) 2 + ( x n + 1 ) 2 ( x n + 1 ) 2 ( 1 x n + 1 ) 2 = | x | ( x n + 1 ) 2 ( 1 x n + 1 ) 2 = 1 ( x n + 1 ) 2 ( 1 x n + 1 ) 2 .

Using the above, we compute σ 1 σ ( x ) :

σ 1 σ ( x ) = σ 1 ( ( x 1 , x n ) 1 x n + 1 ) = ( 2 x 1 , , 2 x n , ( | σ ( x ) | 2 1 ) ( 1 x n + 1 ) ) ( | σ ( x ) | 2 + 1 ) ( 1 x n + 1 ) = ( 2 x 1 , , 2 x n , ( 1 x n + 1 ) ( 1 x n + 1 ) ) ( 1 ( x n + 1 ) 2 ( 1 x n + 1 ) 2 + 1 ) ( 1 x n + 1 ) = ( 2 x 1 , 2 x n , 2 x n + 1 1 ( x n + 1 ) 2 1 x n + 1 + 1 x n + 1 = ( 2 x 1 , 2 x n , 2 x n + 1 ( 1 ( x n + 1 ) 2 + ( 1 x n + 1 ) 2 1 x n + 1 ) = ( 2 x 1 , 2 x n + 1 ) ( 2 2 x n + 1 1 x n + 1 ) = ( 2 x 1 , 2 x n + 1 ) 2 = ( x 1 , , x n + 1 ) = x .

Similarly, σ σ 1 is given by

σ σ 1 ( x ) = σ ( ( 2 x 1 , , 2 x n , | x | 2 1 ) | x | 2 + 1 ) = ( 2 x 1 , , 2 x n ) ( | x | 2 + 1 ) ( 1 | x | 2 1 | x | 2 + 1 ) = ( 2 x 1 , , 2 x n ) | x | 2 + 1 ( | x | 2 1 ) = ( 2 x 1 , , 2 x n ) 2 ) = ( x 1 , x n ) = x .

Thus, σ is a bijection.

(c)
The transition map is given by:
σ ~ σ 1 : σ ( 𝕊 n { N , S } ) σ ~ ( 𝕊 n { N , S } ) , ( u 1 , , u n ) ( u 1 , , u n ) | u |

(since σ ( S ) = 0 , the point u = 0 is not in the domain of σ ~ σ 1 ). To see why, compute

σ ~ σ 1 ( u 1 , u n ) = σ ~ ( ( 2 u 1 , , 2 u n , | u | 2 1 ) | u | 2 + 1 ) = σ ( ( 1 ) ( ( 2 u 1 , , 2 u n , | u | 2 1 ) | u | 2 + 1 ) ) = σ ( ( 2 u 1 , , 2 u n , | u | 2 1 ) | u | 2 1 ) = ( 2 u 1 , , 2 u n ) ( | u | 2 + 1 ) + ( | u | 2 1 ) = ( 2 u 1 , , 2 u n ) 2 | u | 2 = u | u | 2

We can skip a similar computation for its inverse by noticing that it is its own inverse:

( σ ~ σ 1 ) ( σ ~ σ 1 ) ( u ) = σ ~ σ 1 ( u | u | 2 ) = u | u | 2 | u | u | 2 | 2 = u | u | 2 1 | u | 2 = u

In other words, the transition map is a diffeomorphism. Same goes for σ σ ~ 1 . Since 𝕊 n { N } and 𝕊 n { S } form an open cover of 𝕊 n , we conclude that ( 𝕊 n { N } , σ ) and ( 𝕊 n { S } , σ ~ ) define a smooth structure on 𝕊 n .

(d)
The charts from Example 1.31 can be grouped into three categories:
  • ( U n + 1 + , ϕ n + 1 + ) which contain N .
  • ( U n + 1 , ϕ n + 1 ) which contain S .
  • ( U i ± , ϕ i ± ) for i = 1 , , n which do not contain N and S .

We argue that each type of chart is smoothly compatible with σ ; the case for σ ~ is similar. In the first and in the second case, we have

ϕ n + 1 ± σ 1 ( u 1 , , u n ) = ( 2 u 1 , , 2 u n ) | u | 2 + 1 ,

Both of these functions are smooth. The inverses are given by

σ ( ϕ n + 1 ± ) 1 ( u 1 , , u n ) = ( u 1 , , u n ) 1 1 | u | 2

Thus, σ ( ϕ n + 1 ) 1 is smooth. Since the domain ( ϕ n + 1 + ) ( U n + 1 + { N } ) does not include 0 , σ ( ϕ n + 1 + ) 1 is smooth.

For the third case, we have

ϕ i ± σ 1 ( u 1 , , u n ) = ( 2 u 1 , , 2 u i ^ , , 2 u n , | u | 2 1 ) | u | 2 + 1 ,

which is smooth, and

σ ( ϕ i ± ) 1 ( u 1 , , u n ) = ( u 1 , , u i 1 , 1 | u | 2 , u i , , u n 1 ) 1 u n ,

which is also smooth, since u n 1 in U i ± . In other words, the smooth atlas from Example 1.31 is smoothly compatible with the smooth atlas of stereographic projection.

User profile picture
2023-08-24 12:12
Comments