0% found this document useful (0 votes)
25 views

Solutions 2

This document provides solutions to homework problems involving stereographic projection and Riemannian metrics on spheres. It first gives explicit formulas for the stereographic projection of n-dimensional spheres onto planes in cases where n=1, 2, 3. It then shows that for the unit 2-sphere, points with rational Cartesian coordinates have rational stereographic coordinates and vice versa. Finally, it expresses the Riemannian metric on circles and spheres using both standard angular coordinates and stereographic coordinates obtained via projection.

Uploaded by

Jose Luis Giri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
25 views

Solutions 2

This document provides solutions to homework problems involving stereographic projection and Riemannian metrics on spheres. It first gives explicit formulas for the stereographic projection of n-dimensional spheres onto planes in cases where n=1, 2, 3. It then shows that for the unit 2-sphere, points with rational Cartesian coordinates have rational stereographic coordinates and vice versa. Finally, it expresses the Riemannian metric on circles and spheres using both standard angular coordinates and stereographic coordinates obtained via projection.

Uploaded by

Jose Luis Giri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Homework 2.

Solutions
1 a) Write down explicit formulae expressing stereographic coordinates for n-dimensional sphere (x1 )2 +
. . . + (xn+1 )2 = R2 of radius R via coordinates x1 , . . . , xn+1 and vice versa.
(For simplicity you may consider cases n = 2, 3.)
b)† Check that for unit sphere S 2 (x2 + y 2 + z 2 = 1) all the points with rational cartesian coordinates
x, y, z have rational stereographic coordinates u, v and vice versa.
a) Write down the stereographic projection from the North pole of the sphere–point N = (0, 0, . . . , R)
on the plane xn+1 = 0. Consider the segment N D which intersects the sphere at the point (x1 , . . . , xn+1 )
((x1 )2 + (x2 )2 + . . . + (xn+1 )2 = R2 ). This segment intersects the plane xn+1 = 0 at the point D with
coordinates xi ui for i = 1, . . . , n. Then comparing similar triangles we have

R ui Rxi
n+1
= i, i.e. ui = (i = 1, . . . , n)
R−x x R − xn+1
and
ui (R − xn+1 )
xi = , (i = 1, . . . , n) .
R
Using the fact that (x1 )2 + . . . + (xn+1 )2 = R2 we come to
¡ ¢2 n
1 2 n 2 (R − xn+1 ) X ¡ i ¢2
(x ) + . . . + (x ) = u = (R − xn+1 )(R + xn+1 ) .
R2 i=1

Dividing by R − xn+1 (xn+1 6= R since North pole is removed) we come to


Pn
n+1 (ui )2 − R2 i 2ui R2
x = Pni=1 i 2 2
R , x = Pn i 2 2
(i = 1, 2, . . .)
i=1 (u ) + R i=1 (u ) + R

For projection with centre in South pole we have to change xn+1 7→ −xn+1 .
Write down these formulae for cases n = 1, 2, 3,
Case n = 1: Circle x2 + y 2 = R2 . Stereographic coordinate t. Centre of projection (0, R):
( 2
Rx x = R2tR2 +t2
t= , 2 2 (1)
R−y y = tt2 −R
+R2 R

Case n = 2: Sphere x2 + y 2 + z 2 = R2 . Stereographic coordinates u, v. Centre of projection (0, 0, R):


 2uR2
(
u= Rx x =
 R2 +u2 +v 2
R−z
, 2vR2 (2)
Ry y= R2 +u2 +v 2
v= R−z

 u +v 2 −R2
2
z= u2 +v 2 +R2 R

Case n = 3: 3-dimensional sphere x2 + y 2 + z 2 + t2 = R2 . Stereographic coordinates u, v, w. Centre of


projection (0, 0, 0, R2 ):  2
 
 x = R2 +u2uR
Rx  2 +v 2 +w 2
 u = R−t
 
y = 2vR2
Ry R2 +u2 +v 2 +w2
v = R−t , 2 (2)

 w = Rz 
 z = R2 +u2wR


2 +v 2 +w 2
R−t  2 2
+w2 −R2
z = uu2 +v
+v 2 +w2 +R2 R

b)† We see that from explicit formulae. This is rational transformation of conic surfaces.
2 Consider the Riemannian metric on the circle of the radius R induced by the Euclidean metric on the
ambient plane.

1
a) Express it using polar angle as a coordinate on the circle.
b) Express the same metric using stereographic coordinate t obtained by stereographic projection of the
circle on the line, passing through its centre.
Riemannian metric of Euclidean space is G = dx2 + dy 2 . ½
x = R cos ϕ
a) using the angle: In this case parametric equation of circle is . Then
y = R sin ϕ
¯
G = (dx2 + dy 2 )¯x=R cos ϕ,y=R sin ϕ = (d cos ϕ) + (d sin ϕ) = R2 dϕ2 .
2 2

b) In stereographic coordinate using (1) we have:


µ µ ¶¶2 µ µ 2 ¶¶2
¯ 2tR2 t − R2
G = (dx + dy )¯x=x(t),y=y(t) =
2 2
d + d R =
R 2 + t2 R2 + t2
µ ¶2 µ ¶2 µ ¶2 "µ ¶2 µ ¶2 #
2R2 dt 4t2 R2 dt 2tRdt 2t(t2 − R2 )Rdt 2Rdt 2t2 R t(t2 − R2 )
2 2
− 2 + − = R− 2 + t−
R +t (R + t2 )2 R 2 + t2 (R2 + t2 )2 R 2 + t2 (R + t2 ) (R2 + t2 )
µ ¶2 µ ¶
2Rdt R2 (R2 − t2 )2 4R4 t2 4R4 dt2
= + =
R 2 + t2 (R2 + t2 )2 (R2 + t2 )2 (R2 + t2 )2

3 Consider the Riemannian metric on the sphere of the radius R induced by the Euclidean metric on
the ambient 3-dimensional space.
a) Express it using spherical coordinates on the sphere.
b) Express the same metric using stereographic coordinates u, v obtained by stereographic projection of
the sphere on the plane, passing through its centre.
Solution
Riemannian metric of Euclidean space is G = dx2 + dy 2 + dz 2 . (
x = R sin θ cos ϕ
a) using the spherical coordinates: In this case parametric equation of sphere is y = R sin θ sin ϕ .
z = R cos θ
Then
¯
G = (dx2 +dy 2 +dz 2 )¯x=R sin θ cos ϕ,y=R sin θ sin ϕ,z=R cos θ = R2 ((d sin θ cos ϕ)) +R2 ((d sin θ sin ϕ)) +R2 ((d cos θ)) =
2 2 2

2 2 2
R2 (cos θ cos ϕdθ − sin θ sin ϕdϕ) + R2 (cos θ sin ϕdθ + sin θ cos ϕdϕ) + R2 (− sin θdθ) = dθ2 + sin2 θdϕ2 .
¯
b) in stereographic coordinates using (2) we have G = (dx2 + dy 2 + dz 2 )¯x=x(u,v),y=y(u,v),z=z(u,v) =

µ µ ¶¶2 µ µ ¶¶2 µ µ 2 ¶¶2


2uR2 2vR2 u + v 2 − R2
d + d + d R =
R2 + u2 + v 2 R2 + u2 + v 2 R2 + u2 + v 2
µ ¶2 µ ¶2
2du 2u(2udu + 2vdv) 2dv 2v(2udu + 2vdv)
R4 − + R 2
− +
R2 + u2 + v 2 (R2 + u2 + v 2 )2 R2 + u2 + v 2 (R2 + u2 + v 2 )2
µ ¶2
2 2udu + 2vdv (u2 + v 2 − R2 )(2udu + 2vdv)
+R − =
R2 + u2 + v 2 (R2 + u2 + v 2 )2
4R4 n£ ¤2 £ ¤2 o
2 2 2 4
(R2 − u2 + v 2 )du − 2uvdv + (R2 − v 2 + u2 )dv − 2uvdu + 4R2 (udu + vdv)2 =
(R + u + v )
4R4 n¡ ¢2 o 4R4 (du2 + dv 2 )
2 2 2 4
R2 + u2 + v 2 (du2 + dv 2 ) =
(R + u + v ) (R2 + u2 + v 2 )2

2
Remark
In the case of n-dimensional sphere S n of radius R in (n + 1)-dimensional Euclidean space En+1 (it can
be defined by the equation (x1 )2 + . . . + (xn+1 )2 = 1 in cartesian coordinates x1 , . . . , xn , xn+1 ) Riemannian
metric on this sphere induced by the Euclidean metric in the ambient space in stereographic coordinates has
following appearance:
 
n µ µ
X ¶¶ 2 µ µ Pn ¶¶2
¡ ¢ ¯ 2R u 2 j
i=1 (u
i 2
) − R2
G = (dx1 )2 + . . . + (dxn+1 )2 ¯xµ =xi (ui ) =  d Pn
 + d R P n =
j=1
R2 + i=1 (ui )2 R2 + i=1 (ui )2

Pn
4R4 i=1 (dui )2
= Pn
(R2 + i=1 (ui )2 )2

4 Consider the surface L which is the upper sheet of two-sheeted hyperboloid in R3 :

L: z 2 − x2 − y 2 = 1, z > 0.

a) Find parametric equation of the surface L using hyperbolic functions cosh, sinh following an analogy
with spherical coordinates on the sphere.
(The surface L sometimes is called pseudo-sphere.)

b) Consider the stereographic projection of the surface L on the plane OXY , i.e. the central projection
on the plane z = 0 with the centre at the point (0, 0, −1).
Show that the image of projection
( of the surface L is the open disc x2 + y 2 < 1 in the plane OXY .
x = sinh θ cos ϕ
a) Parametric equation is y = sinh θ sin ϕ We see that the condition z 2 − x2 − y 2 = 1 is fulfilled.
z = cosh θθ cos ϕ
(Compare with equation of sphere in spheric coordinates.)

b) Calculations are very similar to the case of stereographic coordinates for 2-sphere x2 + y 2 + z 2 = 1
of the radius R = 1. Stereographic coordinates u, v. Centre of projection (0, 0, −1): We have ux = yv = 1+z1
.
½ x
u = 1+z
Hence v = y . Since x = u(1 + z), y = v(1 + z) then z 2 − 1 = x2 + y 2 and z 2 − 1 = (u2 + v 2 )(1 + z)2 ,
1+z
1+u2 +v 2
i.e. z = 1−u2 −v 2 . We come to
 2u
½
u= x x =
 1−u2 −v 2
2v
1+z
y , y= 1−u2 −v 2 , |u| < 1, |v| < 1 . (4)
v= 1+z 
z = u2 +v 2 +1
1−u2 −v 2

x2 +y 2 z 2 −1 z−1
The image of upper-sheet is an open disc u2 + v 2 = 1 since u2 + v 2 = (1+z)2 = (1+z)2 = z+1 . Since for upper
sheet z > 1 then 0 ≤ z−1
z+1 < 1.

5 Consider the pseudo-Riemannian, pseudo-Euclidean metric on R3 given by the formula

ds2 = dx2 + dy 2 − dz 2 .

Calculate the induced metric on the surface L considered in the Exercise 4, and show that it is a
Riemannian metric (it is positive-definite).
Perform calculations in spherical-like coordinates (see Exercise 4a) above) and in stereographic coordi-
nates (see exercise 4b) above)

3
Remark The surface L sometimes is called pseudosphere. The Riemannian metric on this surface
sometimes is called Lobachevsky (hyperbolic) metric.
The surface L with this metric realises Lobachevsky (hyperbolic) geometry, where Euclid‘s 5-th Axiom
fails. This Riemannian manifold (manifold+Riemannian metric) we call Lobachevsky (hyperbolic) plane.
In stereographic coordinates we come to realisation of Lobachevsky plane on the disc in E2 . It is so
called Poincare model of Lobachevsky geometry.
Solution. The calculations will be very similar to the calculations performed in the exercise 3 above.
Just we need consider cosh θ, sinh θ instead cos θ, sin θ and and sometimes changes the signs.
First of all consider spherical-like coordinates:
(
x = sinh θ cos ϕ
Equation of two-sheeted hyperboloid is y = sinh θ sin ϕ . Then
z = cosh θ
¯
G = (dx2 +dy 2 −dz 2 )¯x=sinh θ cos ϕ,y=sinh θ sin ϕ,z=cosh θ = ((d sinh θ cos ϕ)) +((d sinh θ sin ϕ)) −((d cosh θ)) =
2 2 2

2 2 2
(cosh θ cos ϕdθ − sinh θ sin ϕdϕ) + (cosh θ sin ϕdθ + sinh θ cos ϕdϕ) + (sinh θdθ) = dθ2 + sinh2 θdϕ2 .
µ ¶
1 0
matrix of Riemannian metric is G = . In the same way as for sphere these coordinates are
0 sinh2 θ
2
well-defined in all points except z = ±1, where sin θ = 0.
Now express Riemannian metric in stereographic coordinates (4):
µ µ ¶¶2 µ µ ¶¶2 µ µ 2 ¶¶2
¯ 2u 2v u + v2 + 1
G = (dx2 +dy 2 −dz 2 )¯x=x(u,v),y=y(u,v),z=z(u,v) = d + d − d =
1 − u2 − v 2 1 − u2 − v 2 1 − u2 − v 2

(Compare with calculations for sphere x2 + y 2 + z 2 = 1). We have G = dx2 + dy 2 − dz 2 =


µ ¶2 µ ¶2
2du 2u(2udu + 2vdv) 2dv 2v(2udu + 2vdv)
+ + + −
1 − u2 − v 2 (1 − u2 − v 2 )2 1 − u2 − v 2 (1 − u2 − v 2 )2
µ ¶2
2udu + 2vdv (u2 + v 2 + 1)(2udu + 2vdv) 4(du)2 + 4(dv)2
− + = .
1 − u2 − v 2 (1 − u2 − v 2 )2 (1 − u2 − v 2 )2
(To perform these calculations it is convenient to denote by s = 1 − u2 − v 2 .)
Resume: We come to the induced Riemannian metric on the surface from the pseudo-Riemannian metric
in ambient space.
6∗ Lobachevsky plane (hyperbolic plane) L in stereographic coordinates can be considered as an open disc
u + v 2 < 1 in the plane. In the previous exercise in particularly we calculated Riemannian metric of L in
2

these coordinates.
Find new coordinates x, y such that in these coordinates Lobachevsky plane (hyperbolic plane) can be
considered as an upper half plane y > 0} and write down explicitly Riemannian metric in these coordinates.
Hint: You may use complex coordinates:

z = x + iy, z̄ = x − iy, w = u + iv, w̄ = u − iv

and find an holomorphic transformation w = w(z) of the open disc ww̄ < 1 onto the upper plane Imz > 0.
Solution.
Recall that in the previous exercise we calculated expression for Lobachevsky metric in stereographic
4du2 +4dv 2
coordinates u, v, u2 + v 2 < 1. We come to the answer: G = (1−u 2 −v 2 )2 (see the previous exercise). (It was

4
realisation of Lobachevsky plane on the Euclidean disc. Sometimes it called Poincare model of Lobachevsky
(hyperbolic) geometry.)
4du2 +4dv 2
In complex coordinates w = u + iv, w̄ = u − iv the metric G = (1−u 2 −v 2 )2 obtained in the exercise 8 can

4dwdw̄2
be rewritten G = (1−ww̄) . Indeed

4dwdw̄ 4d(u + iv)d(u − iv) 4du2 + 4dv 2


G= 2
=G= 2
= .
(1 − ww̄) (1 − (u + iv)(u − iv)) (1 − u2 − v 2 )2

It is a beautiful problem in complex analysis: find Mobius transformation w = az+b


cz+d transformation which
transforms the interior of circle ww̄ = 1 into upper half plane Imz > 0. One can see that
1 + iz 1−w
w= , z=i
1 − iz 1+w
is the transformation which we need (Can you find all Mobius transformations which transform upper half
plane to the interior of unit circle?.)
Now calculate G in coordinates z, z̄. i.e. in coordinates (x, y):
4du2 + 4dv 2 4dwdw̄
G= =
(1 − u2 − v 2 )2 (1 − ww̄)2
We have µ ¶
1 + iz 2idz −2idz̄
dw = d = , dw̄ = ,
1 − iz (1 − iz)2 (1 + iz̄)2
1 + iz 1 − iz̄ 2i(z̄ − z)
1 − ww̄ = 1 − =
1 − iz 1 + iz̄ (1 − iz)(1 + iz̄)
Hence ³ ´³ ´
2idz −2idz̄
4dwdw̄ 4 (1−iz)2 (1+iz̄)2 −4ddz̄ dx2 + dy 2
G= = −4(z̄−z)2
= = ,
(1 − ww̄)2 (z̄ − z)2 y2
(1−iz)2 (1+iz̄)2

since z = x + iy and z̄ − z = −2iy.


We come to the very useful interpretation of hyperbolic geometry: upper half plane in E2 with metric
2 2
G = dx y+dy
2 . Later by default we will call ”Lobachevsky (hyperbolic) plane” the realisation of Lobachevsky
dx2 +dy 2
plane as an half-upper plane in E2 with these coordinates x, y (y > 0) with metric G = y2 .
7 ∗ Consider the metric induced on one-sheeted hyperboloid x2 + y 2 − z 2 = 1 embedded in R3 with the
pseudo-Euclidean metric dx2 + dy 2 − dz 2 (see the exercise 5). Show that this metric is not Riemannian one.
Solution. One can ( perform straightforward calculations in spherical-like coordinates: Equation of one-
x = cosh θ cos ϕ
sheeted hyperboloid is y = cosh θ sin ϕ . Then
z = sinh θ
¯
G = (dx2 +dy 2 −dz 2 )¯x=cosh θ cos ϕ,y=cosh θ sin ϕ,z=sinh θ = ((d cosh θ cos ϕ)) +((d cosh θ sin ϕ)) −((d sinh θ)) =
2 2 2

2 2 2
(sinh θ cos ϕdθ − cosh θ sin ϕdϕ) + (sinh θ sin ϕdθ + cosh θ cos ϕdϕ) − (cosh θdθ) = −dθ2 + cosh2 θdϕ2 .
µ ¶
−1 0
matrix is G = . The condition of positive-definiteness is not fulfilled. This is not Riemannian
0 cosh2 θ
metric.
∂ ∂
Another solution Consider the vectors e = ∂y and f = ∂z attached at the point (1, 0, 0). One can see
that these vectors are tangent to the hyperboloid, but they have the ”length” of different sign. (One of these
vectors is space-like vector, another time like vector.) We have pseudoriemannian metric at the tangent
space spanned by these two vectors.

You might also like