Previous Up Next
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Lecture 3 Homogeneous Spaces from the Group SL2(ℝ)

Now we adopt the previous theoretical constructions for the particular case of the group SL2(ℝ). It is common to present SL2(ℝ) solely as the transformation group of the Lobachevsky half-plane. We take a wider viewpoint considering SL2(ℝ) as an abstract group and wish to classify all its realizations as a transformation group of certain sets. More specifically, we describe all homogeneous spaces SL2(ℝ)/H, where H is a closed continuous subgroup of SL2(ℝ), see Section 2.2.2. To begin with, we start from the two-dimensional subgroup.

3.1 The Affine Group and the Real Line

The affine group of the real line, also known as the ax+b group , can be identified with either subgroup of lower- or upper-triangular matrices of the form:

  F=





1
a


        a0
c1


, a>0





,   F′=





1
a


        ab
01


, a>0





.

These subgroups are obviously conjugates of each other and we can consider only the subgroup F here. We are following the construction from Section 2.2.2 and using its notations.

Firstly, we address the simpler case of the subgroup F of all lower triangular matrices in SL2(ℝ), which admits both positive and negative entries on the main diagonal. The corresponding homogeneous space X=SL2(ℝ)/F′ is one-di­men­sio­nal and can be parametrised by real numbers. We define the natural projection p as:

     
    p:SL2(ℝ)→ ℝ:


    ab
cd


↦ 
b
d
,     for  d≠ 0,
(1)

since there is the following decomposition: 1



    ab
cd


=





    1
b
d
01










    
1
d
0
cd





,     where      





    
1
d
0
cd





∈ F′. (a)

The decomposition shows that each coset in SL2(ℝ)/F′ either

Then, we define the smooth map s to be a right inverse of p:

s: ℝ→ SL2(ℝ): u ↦


    1u
01


. (1)

The corresponding map r(g)=s(p(g))−1g is calculated to be:

     
    r:SL2(ℝ)→ F′:


    ab
cd


↦ 


    d−10 
cd


.
(2)

Consequently, the decomposition g=s(p(g))r(g) is given by (a).

Therefore the action of SL2(ℝ) on SL2(ℝ)/F′ is the Möbius (linear-fractional) transformations of the projective real line $ℝ^_$=ℝ∪{∞}:

g:u↦ p(g*s(u)) =
au+b
cu+d
,    where g=  


    ab
cd


. (3)
Exercise 1
  1. Check that the derived actions, see Exercise 30, associated with the one-parameter subgroups a(t), b(t) and z(t) from Exercise 24, are, respectively:
          AF=x
    d
    dx
    ,   BF=
    x2−1
    2
    d
    dx
    ,    ZF=−(x2+1)
    d
    dx
    . 
  2. Verify that the above operators satisfy the commutator relations for the Lie algebra sl2, cf. (13):
    [ZF,AF]=2BF,   [ZF,BF]=−2AF,   [AF,BF]= −
    1
    2
    ZF. (4)

In Section 4.4, we will see the relevance of this action to projective spaces. 4

Exercise 1(a) Formula (3) uses division by cu+d, how it shall be understood (or modified) if cu+d=0?
Hint: A similar issue is addressed in Sec. 
8.1 and can be consistently resolved through the projective coordinates. Another common approach is to extend arithmetic of real numbers to the projective real line $ℝ^_$ by the following rules involving infinity:
a
0
=∞,   
a
=0,   
=1,   where  a≠ 0. (a)
As a consequence we also note that a·∞+b=a·∞ for a≠ 0. But we do not define (and, actually, will not need) values of 0/0, 0· ∞ or ∞−∞, etc. Check, that formulae (a) together with ordinary arithmetic define in (3) a bijection $ℝ^_$→ $ℝ^_$ for gSL2(ℝ).

To describe the homogeneous space SL2(ℝ)/F and SL2(ℝ)-action on it, we rewrite the identity (a) for d≠ 0 as:



    ab
cd


=





    1
b
d
01







    χ(d)0
0χ(d)









    
1

d
0
χ(d)c

d







, (b)

where χ is the Heaviside function (10). The last matrix in the right-hand side of (b) belongs to F and is uniquely defined by this identity. Thus, cosets representing matrices with d≠ 0 are parametrised by ℝ×{−1,+1}. The respective realisations of maps from Section 2.2.2 are:

     
    p:SL2(ℝ)→ ℝ×{−1,+1}:


    ab
cd


↦ 


b
d
, χ(d)


,
(c)
    s:ℝ×{−1,+1}→ SL2(ℝ):
(u,σ) ↦


      σσ u
0σ 


,
(d)
      r:SL2(ℝ)→ F:


      ab
cd









    
1

d
0
χ(d)c

d







.
(e)
Exercise 1(b) Extend the above maps (c)–(e) for the matrices with d=0.
Hint: Note, that all matrices (
      ab
b−10
)∈SL2(ℝ) belong to two different cosets in SL2(ℝ)/F depending from the value of χ(b). These cosets can be denoted by (∞,+1) and (∞,−1).

Thus, the homogeneous space SL2(ℝ)/F can be identified with the double cover of the projective real line. SL2(ℝ)-action on SL2(ℝ)/F in terms of the above maps (c)–(e) is:

g:(u,σ)↦ p(g*s(u,σ )) =


au+b
cu+d
,  χ(cu+d)σ 


,  where g=  


    ab
cd


. (f)
Exercise 1(c) Extend the arithmetic rules (a) and the Heaviside function in (f) for the exceptional cases of either u=∞ or cu+d=0 to produce a bijection SL2(ℝ)/FSL2(ℝ)/F.

Here is some properties of the Möbius transformations of the real line:

Exercise 1(d)
  1. For a non-constant Möbius transformation, there may exist only two, one or none fixed point(s) on the real line.
  2. Let x1, x2, x3 be arbitrary pairwise distinct points of $ℝ^_$ with x2 being between x1 and x3. Then there is the unique element in SL2(ℝ) such that the associated Möbius transformation maps 0, 1, to x1, x2, x3 respectively. Therefore, SL2(ℝ) acts simply transitively on ordered triples of pair-wise distinct elements of $ℝ^_$. Hint: Check that the matrix (
          x3
    x2x1
    x3x2
    x1
          
    x2x1
    x3x2
    1
    ) provides the required map and normalize its determinant.

3.2 One-dimensional Subgroups of SL2(ℝ)

Any element X of the Lie algebra sl2 defines a one-parameter continuous subgroup gX(t)=etX, t∈ℝ of SL2(ℝ). Two such subgroups gX(t) and gY(t) with X=aY for a non-zero a∈ ℝ coincide as sets and we can easily re-parameterise them: gY(t)=gX(at). Therefore we will not distinguish such subgroups in the following.

We already listed four one-parameter continuous subgroups in Exercise 3 and can provide further examples, e.g. the subgroup of lower-triangular matrices. However, there are only three different types of subgroups under the matrix similarity AMAM−1.

Proposition 2 Any continuous one-parameter subgroup of SL2(ℝ) is conjugate to one of the following subgroups:
     
      A=




et/20
0et/2


=exp


t/20
0t/2


,  t∈ℝ

,
(1)
      N=




1t
01


=exp


0t
00


,
  t∈ℝ

,
(2)
      K=




          costsint
          −sintcost


=   exp


0t
t0


, t∈(−π,π]

.
(3)

Proof. Any one-parameter subgroup is obtained through the exponential map, see Section 2.3,

etX=
n=0
tn
n!
Xn (4)

of an element X of the Lie algebra sl2 of SL2(ℝ). Such an X is a 2× 2 matrix with zero trace. The behaviour of the Taylor expansion (4) depends on the properties of the powers Xn. This can be classified by a straightforward calculation:

Lemma 3 The square X2 of a traceless matrix X= (
      ab
ca
) is the identity matrix times a2+bc=−detX. The factor can be negative, zero or positive, which corresponds to the three different types of the Taylor expansion (4) of etX.

It is a simple exercise in characteristic polynomials to see that, through the matrix similarity, we can obtain from X a generator

The determinant is invariant under the similarity. Thus, these cases are distinct.


Exercise 4 Find matrix conjugations of the following two subgroups to A and N respectively:
     
      =




coshtsinht
sinhtcosht


=exp


0t
t0


,  t∈ℝ

,
(5)
      N=




10
t1


=exp


00
t0


,
  t∈ℝ

.
(6)
Exercise 4(a)
  1. Determine which gSL2(ℝ) belongs to a one-partial­rightarrow­me­ter subgroup, that is find the range of the exponential map XeX, where Xsl2 and eXSL2(ℝ). Hint: The image excludes matrices with real, negative eigenvalues, other than I [125]*Ex. 3.22.
  2. Show that g=eXSL2(ℝ) belongs to a subgroup conjugated to A, N or K if and only if | tr(g) | is bigger, equal or less than 2 respectively. Hint: Use the invariance of trace under matrix conjugation and the values of trace for matrices in A, N and K.
  3. Let a Möbius transformation of the real line has two, one or none fixed points as described in Exercise 1, then it is generated by a matrix which is conjugated to an element of the subgroups A, N and K respectively. Why we do not mention either the matrix belongs to the image of exponential map in this part?

We will often use subgroups and N′ as representatives of the corresponding equivalence classes under matrix conjugation.

An interesting property of the subgroups A, N and K is their appearance in the Iwasawa decomposition [240]*§ III.1 of SL2(ℝ)=ANK in the following sense. Any element of SL2(ℝ) can be represented as the product:



    ab
cd


= 


α0
0α−1




1ν 
01




      cosφ −sinφ
      sinφcosφ


. (7)
Exercise 5 Check that the values of parameters in the above decomposition are as follows:
     α=(c2+d2)−1/2,    ν=ac+bd,   φ = arctan
c
d
.
Consequently, cosφ=d/√c2+d2 and sinφ=−c/√c2+d2.

The Iwasawa decomposition shows once more that SL2(ℝ) is a three-dimensional manifold. A similar decomposition G=ANK is possible for any semisimple Lie group G, where K is the maximal compact group, N is nilpotent and A normalises N. Although the Iwasawa decomposition will be used here on several occasions, it does not play a crucial role in the present consideration. Instead, Proposition 2 will be the cornerstone of our construction.

3.3 Two-dimensional Homogeneous Spaces

Here, we calculate the action of SL2(ℝ) (3) (see Section 2.2.2) on X=SL2(ℝ)/H for all three possible one-dimensional subgroups H=, N′ or K. Counting dimensions (3−1=2) suggests that the corresponding homogeneous spaces are two-dimensional manifolds. In fact, we identify X in each case with a subset of ℝ2 as follows. First, for every equivalence class of SL2(ℝ)/H we chose a representative, which is an upper-triangular matrix.

Exercise 6 Show that
  1. For any matrix gSL2(ℝ) and for each value σ=−1, 0, 1 there is a factorisation g=gu (
          dσ c
    cd
    ) for some upper-triangular 2× 2 matrix gu. Hint: See (9), (12) and (15).
  2. There is at most one upper-triangular matrix in every equivalence class SL2(ℝ)/H, where H=, N or K, where, in the last case, uniqueness is up to the constant factor ± 1. For the subgroup , there may be not such upper triangular matrix.
    Hint: The identity matrix is the only upper-triangular matrix in these three subgroups, where, again, the uniqueness for the subgroup K is understood up to the scalar factor ±1.

The existence of such a triangular matrix will be demonstrated in each case separately. Now, we define the projection p:SL2(ℝ)→ X, assigning p(g)=(a1b1,a12), where (

a1b1
0a1−1

) is the upper-triangular matrix1 representing the equivalence class of g. We also choose [170]*p. 108 the map s: XG in the form:

s: (u,v) ↦
1
v


    vu
01


,    (u,v)∈ℝ2,  v>0. (8)

This formula will be used for all three possible subgroups H.

3.3.1 From the Subgroup K

The homogeneous space SL2(ℝ)/K is the most traditional case in representation theory. The maps p and s defined above produce the following decomposition g=s(p(g))r(g):



    ab
cd


=
1
d2+c2


    1 bd+ ac
0c2+d2




    dc
cd


. (9)

Then, the SL2(ℝ)-action defined by the formula g· x=p(g*s(x)) (3) takes the form:



    ab
cd


:  (u,v)↦


(au+b)(cu+d) +cav2
( cu+d)2 +(cv)2
, 
v
( cu+d)2 +(cv)2



. (10)
Exercise 7Use CAS to check the above formula, as well as analogous formulae (13) and (1) below. See Appendix C.3 for CAS usage.

Obviously, it preserves the upper half-plane v>0. The expression (10) is very cumbersome and it can be simplified by the complex imaginary unit i2=−1, which reduces (10) to the Möbius transformation



    ab
cd


:  w↦  
aw+b
cw+d
,    where  w=u+iv. (11)

We need to assign a meaning to the case cw+d=0 and this can be done by the addition of an infinite point ∞ to the set of complex numbers—see, for example, [26]*Definition 13.1.3 for details.

In this case, complex numbers appeared naturally.

3.3.2 From the Subgroup N

We consider the subgroup of lower-triangular matrices N′ (6). For this subgroup, the representative of cosets among the upper-triangular matrices will be different. Therefore, we receive an apparently different decomposition g=s(p(g))r(g), cf. (9)



    ab
cd


=
1
d2


    1 bd
0d2




    d0
cd


,    where  d≠ 0. (12)

We postpone the treatment of the exceptional case d=0 until Section 8.1. The SL2(ℝ)-action (3) now takes the form:



    ab
cd


:  (u,v)↦


au+b
cu+d
, 
v
( cu+d)2



. (13)

This map preserves the upper half-plane v>0 just like the case of the subgroup K. The expression (13) is simpler than (10), yet we can again rewrite it as a linear-fractional transformation with the help of the dual numbers unit ε 2=0:



    ab
cd


:  w↦  
aw+b
cw+d
,    where  w=u+ε v. (14)

We briefly review the algebra of dual numbers in Appendix B.1. Since they have zero divisors, the fraction is not properly defined for all cu+d=0. The proper treatment will be considered in Section 8.1 since it is not as simple as in the case of complex numbers.

3.3.3 From the Subgroup

In the last case of the subgroup , we still can obtain the decomposition



    ab
cd


=
1
d2c2


    1 bdac
0d2c2




    dc
cd


,  where  d≠ ± c. (15)

However, the new aspect is here that (15) presents the decomposition g=s(p(g))r(g) if and only if d2c2>0. Otherwise the matrix (

dc
cd

) is not in . If d2c2<0 we use the decomposition 15



    ab
cd


=
1
c2d2


    acbd−1 
c2d20




    cd
dc


,  where  d≠ ± c. (a)

The modified maps p(g): GG/H is:

p(g)=





      (a1b1,a12),
if  g∼ 


a1b1
0a1−1


;
      (a1b1,−a12),
if  g∼ 


b1a1
a1−10


.
(b)

The respective modification of the map s: G/HG is:

s(u,v)= 











      
1
v


        vu
01


,
if  v>0;
      
1
v


        uv
10


,
if  v<0.
(c)

The geometrical meaning of this modification is that the homogeneous spaces G/K and G/N are parametrised by points of the upper half-plane, while G/ shall be parametrised by the union of the upper and lower half-planes. We will further discuss this difference Section 8.2. The exceptional situation dc will be treated in Section 8.1.

The SL2(ℝ)-action (3) takes the form



    ab
cd


:  (u,v)↦


(au+b)(cu+d) −cav2
( cu+d)2 −(cv)2
, 
v
( cu+d)2 −(cv)2



. (1)

Notably, this time the map does not preserve the upper half-plane v>0: the sign of ( c u+d)2 −(cv)2 is not determined. To express this map as a Möbius transformation, we require the double numbers (also known as split-complex numbers) unit є 2=1:

  


    ab
cd


:  w↦  
aw+b
cw+d
,    where  w=u+є v.

The algebra of double numbers is briefly introduced in Appendix B.1.

3.3.4 Unifying All Three Cases

There is an obvious similarity in the formulae obtained in each of the above cases. To present them in a unified way, we introduce the parameter σ which is equal to −1, 0 or 1 for the subgroups K, N′ or , respectively. Then, decompositions (9), (12) and (15) are



    ab
cd


=
1
d2−σ c2


    1 bd−σ  ac
0d2−σ c2




    dσ c
cd


,   where  d2−σ c2≠ 0. (2)

The respective SL2(ℝ)-actions on the homogeneous space SL2(ℝ)/H, where H=, N′ or K, are given by



    ab
cd


:  (u,v)↦


(au+b)(cu+d) −σ cav2
( cu+d)2 −σ (cv)2
,
v
( cu+d)2 −σ (cv)2



. (3)

Finally, this action becomes the linear-fractional (Möbius) transformation for hypercomplex numbers in two-dimensional commutative associative algebra (see Appendix B.1) spanned by 1 and ι:



    ab
cd


:  w↦  
aw+b
cw+d
,    where  w=u+ι v,   ι2=σ. (4)

Thus, a comprehensive study of SL2(ℝ)-homogeneous spaces naturally introduces three number systems. Obviously, only one case (complex numbers) belongs to mainstream mathematics. We start to discover empty cells in our periodic table.

Remark 8 As we can now see, the dual and double numbers naturally appear in relation to the group SL2(ℝ) and, thus, their introduction in [185, 190] was not “a purely generalistic attempt”, cf. the remark on quaternions of [288]*p. 4.
Remark 9 A different choice of the map s: G/HG will produce different (but isomorphic) geometric models. In this way, we will obtain three types of “unit disks” in Chapter 10.

3.4 Elliptic, Parabolic and Hyperbolic Cases

As we have seen in the previous section, there is no need to be restricted to the traditional route of complex numbers only. The arithmetic of dual and double numbers is different from complex numbers mainly in the following aspects:

  1. They have zero divisors. However, we are still able to define their transforms by (4) in most cases. The proper treatment of zero divisors will be done through corresponding compactification—see Section 8.1.
  2. They are not algebraically closed. However, this property of complex numbers is not used very often in analysis.

We have agreed in Section 1.1 that three possible values −1, 0 and 1 of σ:=ι2 will be referred to here as elliptic, parabolic and hyperbolic cases, respectively. This separation into three cases will be referred to as the EPH classification. Unfortunately, there is a clash here with the already established label for the Lobachevsky geometry. It is often called hyperbolic geometry because it can be realised as a Riemann geometry on a two-sheet hyperboloid. However, within our framework, the Lobachevsky geometry should be called elliptic and it will have a true hyperbolic counterpart.

Notation 10 We denote the space 2 of vectors u+v ι by e, p or h to highlight which number system (complex, dual or double, respectively) is used. The notation σ is used for a generic case. The use of E, P, H or e, p, h (for example, in labelling the different sections of an exercise) corresponds to the elliptic, parabolic, hyperbolic cases.
Remark 11 In introducing the parabolic objects on a common ground with elliptic and hyperbolic ones, we should warn against some common prejudices suggested by the diagram (2):
  1. The parabolic case is unimportant (has “zero measure”) in comparison to the elliptic and hyperbolic cases. As we shall see (e.g. Remark 8 and 2), the parabolic case presents some richer geometrical features.
  2. The parabolic case is a limiting situation or an intermediate position between the elliptic and hyperbolic cases. All properties of the former can be guessed or obtained as a limit or an average from the latter two. In particular, this point of view is implicitly supposed in [241].

    Although there are some confirmations of this (e.g. Fig. 10.3(E)–(H)), we shall see (e.g. Remark 21) that some properties of the parabolic case cannot be guessed in a straightforward manner from a combination of elliptic and hyperbolic cases.

  3. All three EPH cases are even less disjoint than is usually thought. For example, there are meaningful notions of the centre of a parabola (3) or the focus of a circle (2).
  4. A (co-)invariant geometry is believed to be “coordinate-free”, which sometimes is pushed to an absolute mantra. However, our study within the Erlangen programme framework reveals two useful notions (Definition 3 and (2)), mentioned above, which are defined by coordinate expressions and look very “non-invariant” at first glance.

3.5 Orbits of the Subgroup Actions

We start our investigation of the Möbius transformations (4)

  


    ab
cd


: w↦  
aw+b
cw+d

on the hypercomplex numbers w=uv from a description of orbits produced by the subgroups A, N and K. Due to the Iwasawa decomposition (7), any Möbius transformation can be represented as a superposition of these three actions.

The actions of subgroups A and N for any kind of hypercomplex numbers on the plane are the same as on the real line: A dilates and N shifts—see Fig. 1.1 for illustrations. Thin traversal lines in Fig. 1.1 join points of orbits obtained from the vertical axis by the same values of t and grey arrows represent “local velocities”—vector fields of derived representations.

Exercise 12 Check that:
  1. The matrix (
          et0
    0et
    )=exp(
          −t0
    0t
    ) from A makes a dilation by e−2t, i.e. ze−2t z. The respective derived action, see Example 30, is twice the Euler operator uu+vv.
  2. The matrix (
          1t
    01
    )=exp(
          0t
    00
    ) from N shifts points horizontally by t, i.e. zz+t=(u+t)+ι v. The respective derived action is u.
  3. The subgroup of SL2(ℝ) generated by A and N is isomorphic to the ax+b group, which acts transitively on the upper half-plane.

    Hint: Note that the matrix (

            
    a
    b/
    a
    0
    1/
    a

    )= (

            1b
    01

    ) (

            
    a
    0
    0
    1/
    a

    ) maps ι to aι +b and use Exercise 2.

By contrast, the action of the third matrix from the subgroup K sharply depends on σ=ι2, as illustrated by Fig. 1.2. In elliptic, parabolic and hyperbolic cases, K-orbits are circles, parabolas and (equilateral) hyperbolas, respectively. The meaning of traversal lines and vector fields is the same as on the previous figure.

Exercise 13 The following properties characterise K-orbits:
  1. The derived action of the subgroup K is given by:
    Kσd(u,v)=(1+u2+σ v2)∂u+2uvv,    σ=ι2.  (5)
    Hint: Use the explicit formula for Möbius transformation of the components (3). An alternative with CAS is provided as well, see Appendix C.3 for usage.
  2. A K-orbit in σ passing the point (0,s) has the following equation:
    (u2−σ v2)−2v
    s−1−σ s
    2
    +1=0. (6)
    Hint: Note that the equation (6) defines contour lines of the function F(u,v)=(u2−σ v2+1)/v, that is, solve the equations F(u,v)=const. Then, apply the operator (5) to obtain KσdF=0.
  3. The curvature of a K-orbit at point (0,s) is equal to
    κ=
    2s
    1+σ s2
    . (7)
  4. The transverse line obtained from the vertical axis has the equations:
    (u2−σ v2)+2cot(2φ) u−1=0,   for g=


            cosφsinφ
    −sinφ cosφ


    ∈ K. (8)
    Hint: A direct calculation for a point (0,s) in the formula (3) is possible but demanding. A computer symbolic calculation is provided as well.

Much more efficient proofs will be given later (see Exercise 2), when suitable tools will be at our disposal. It will also explain why K-orbits, which are circles, parabolas and hyperbolas, are defined by the same equation (6). Meanwhile, these formulae allow us to produce geometric characterisation of K-orbits in terms of classical notions of conic sections, cf. Appendix B.2.

Exercise 14 Check the following properties of K-orbits (6):
  1. For the elliptic case, the orbits of K are circles. A circle with centre at (0, (s+s−1)/2) passing through two points (0,s) and (0,s−1).
  2. For the parabolic case, the orbits of K are parabolas with the vertical axis V. A parabola passing through (0,s) has horizontal directrix passing through (0, s−1/(4s)) and focus at (0,s+1/(4s)).
  3. For the hyperbolic case, the orbits of K are hyperbolas with asymptotes parallel to lines uv. A hyperbola passing the point (0,s) has the second branch passing (0,−s−1) and asymptotes crossing at the point (0,(ss−1)/2). Foci of this hyperbola are:
          f1,2=


    0, 
    1
    2


    (1±
    2
    )s−(1∓
    2
    )s−1




    . 

The amount of similarities between orbits in the three EPH cases suggests that they should be unified one way or another. We start such attempts in the next section.

3.6 Unifying EPH Cases: The First Attempt

It is well known that any K-orbit above is a conic section and an interesting observation is that corresponding K-orbits are, in fact, sections of the same two-sided right-angle cone. More precisely, we define the family of double-sided right-angle cones to be parameterized by s>0:

x2+(y
1
2
(s+s−1))2−(z
1
2
(ss−1))2=0. (9)

Therefore, vertices of cones belong to the hyperbola {x=0, y2z2=1}—see Fig. 1.3.

Exercise 15 Derive equations for the K-orbits described in Exercise 14 by calculation of intersection of a cone (9) with the following planes:
  1. Elliptic K-orbits are sections of cones (9) by the plane z=0 (EE on Fig. 1.3).
  2. Parabolic K-orbits are sections of (9) by the plane yz (PP on Fig. 1.3).
  3. Hyperbolic K-orbits are sections of (9) by the plane y=0 (HH on Fig. 1.3);

Moreover, each straight line generating the cone, see Fig. 1.3(b), is crossing corresponding EPH K-orbits at points with the same value of parameter φ from (7). In other words, all three types of orbits are generated by the rotations of this generator along the cone.

Exercise 16Verify that the rotation of a cone’s generator corresponds to the Möbius transformations in three planes.

Hint: I do not know a smart way to check this, so a CAS solution is provided.

From the above algebraic and geometric descriptions of the orbits we can make several observations.

Remark 17
  1. The values of all three vector fields dKe, dKp and dKh coincide on the “real” U-axis (v=0), i.e. they are three different extensions into the domain of the same boundary condition. Another origin of this: the axis U is the intersection of planes EE, PP and HH on Fig. 1.3.
  2. The hyperbola passing through the point (0,1) has the shortest focal length2 among all other hyperbolic orbits since it is the section of the cone x2+(y−1)2+z2=0 closest from the family to the plane HH.
  3. Two hyperbolas passing through (0,v) and (0,v−1) have the same focal length since they are sections of two cones with the same distance from HH. Moreover, two such hyperbolas in the lower and upper half-planes passing the points (0,v) and (0,−v−1) are sections of the same double-sided cone. They are related to each other as explained in Remark 4.

We make a generalisation to all EPH cases of the following notion, which is well known for circles [71]*§ 2.3 and parabolas [339]*§ 10:

Definition 18 A power p of a point (u,v) with respect to a conic section given by the equation x2−σ y2−2lx−2ny+c=0 is defined by the identity
p=u2−σ v2−2lu−2nv+c. (10)
Exercise 19 Check the following properties:
  1. A conic section is the collection of points having zero power with respect to the section.
  2. The collection of points having the same power with respect to two given conic sections of the above type is either empty or the straight line. This line is called radical axis of the two sections.
  3. All K-orbits are coaxial [71]*§ 2.3 with the real line being their joint radical axis, that is, for a given point on the real line, its power with respect to any K-orbit is the same.
  4. All transverse lines (8) are coaxial, with the vertical line u=0 being the respective radial axis.

In the case of circles the power of a point is known as Steiner power.

3.7 Isotropy Subgroups

In Section 2.2 we described the two-sided connection between homogeneous spaces and subgroups. Section 3.3 uses it in one direction: from subgroups to homogeneous spaces. The following exercise does it in the opposite way: from the group action on a homogeneous space to the corresponding subgroup, which fixes the certain point.


Figure 3.1: Actions of isotropy subgroups K, N′ and , which fix point ι in three EPH cases.

Exercise 20 Let SL2(ℝ) act by Möbius transformations (4) on the three number systems. Show that the isotropy subgroups of the point ι are:
  1. The subgroup K in the elliptic case. Thus, the elliptic upper half-plane is a model for the homogeneous space SL2(ℝ)/K.
  2. The subgroup N (6) of matrices


            10
             ν1


    =


            0−1
            10




            1ν 
             01




            01
            −10


    (11)
    in the parabolic case. It also fixes any point ε v on the vertical axis, which is the set of zero divisors in dual numbers. The subgroup N is conjugate to subgroup N, thus the parabolic upper half-plane is a model for the homogeneous space SL2(ℝ)/N.
  3. The subgroup  (5) of matrices


            coshτsinhτ
            sinhτcoshτ 


    =
    1
    2


            1−1
            11




            eτ0
            0e−τ




            11
            −11


    (12)
    in the hyperbolic case. These transformations also fix the light cone centred at є, that is, consisting of є+zero divisors. The subgroup is conjugate to the subgroup A, thus two copies of the upper half-plane (see Section 8.2) are a model for SL2(ℝ)/A.

Figure 3.1 shows actions of the above isotropic subgroups on the respective numbers, we call them rotations around ι. Note, that in parabolic and hyperbolic cases they fix larger sets connected with zero divisors.

It is inspiring to compare the action of subgroups K, N′ and on three number systems, this is presented on Fig. 3.2. Some features are preserved if we move from top to bottom along the same column, that is, keep the subgroup and change the metric of the space. We also note the same system of a gradual transition if we compare pictures from left to right along a particular row. Note, that Fig. 3.1 extracts diagonal images from Fig. 3.2, this puts three images from Fig. 3.1 into a context, which is not obvious from Fig. 3.2. Even greater similarity in the respective analytic expressions is presented by the next exercise.




Figure 3.2: Actions of the subgroups K, N′, are shown in the first, middle and last columns respectively. The elliptic, parabolic and hyperbolic spaces are presented in the first, middle and last rows respectively. The diagonal drawings comprise Fig. 3.1 and the first column Fig. 1.2.


Exercise 21 Using the parameter τ=−1, 0, 1 for the subgroups K, N and respectively, check the following properties of the actions of the subgroups K, N and :
  1. Vector fields of the respective actions are
          (u2+σ v2−τ)∂u+ 2uvv,
    where σ=ι2 represent the metric of the space .
  2. Orbits of the isotropy subgroups , N and K satisfy the equation
    (u2−σ v2)−2nv−τ=0,    where  n∈ℝ. (13)

    Hint: See method used in Exercise 2. An alternative derivation will be available in Exercise 24.

  3. The isotropy subgroups of ι in all EPH cases are uniformly expressed by matrices of the form 13


            aσ b
            ba


    ,    where  a2−σ b2=1. (a)
  4. The isotropy subgroup of a point uv consists of matrices
        






          
    1+σ v2c2
    +uc
    c(σ v2u2)
          c
    1+σ v2c2
    uc






    SL2(ℝ)
    and describe admissible values of the parameter c.

    Hint: Use the previous item and the transitive action of the ax+b from Exercise 3.

  5. The transverse lines on Fig. 3.2 in all nine cases are, cf. (13):
    (u2−σ v2)−lu+τ=0,     where  l∈ℝ. (1)
Definition 22 In the hyperbolic case, we extend the subgroup to a subgroup A by the element (
    01
−1 0
).

This additional elements flips the upper and lower half-planes of double numbers—see Section 8.2. Therefore, the subgroup Ah fixes the set {ι,−ι}.

Lemma 23 Möbius action of SL2(ℝ) in each EPH case is generated by action of the corresponding isotropy subgroup (Ah in the hyperbolic case) and actions of the ax+b group, e.g. subgroups A and N.

Proof. The ax+b group acts transitively on the upper or lower half-planes. Thus, for any gSL2(ℝ), there is an h in the ax+b group such that h−1g either fixes ι or sends it to −ι. Thus, h−1g is in the corresponding isotropy subgroup.


3.7.1 Trigonometric Functions

The actions of isotropy subgroups can be viewed as “rotations around ι”. In the Euclidean geometry rotations are transformations preserving angles. Thus, we can use isotropy subgroups to define angles in all EPH geometries as certain invariants of Möbius transformations.

Definition 24 The σ-tangent (denoted tanσ) between two vectors P=(u,v) and P′=(u′,v′) in σ2 is defined by:
tanσ(P,P′)=
uv′−uv
uu′−σ vv
 . (2)
Then, the respective σ-cosine and σ-sine functions are connected to σ-tangent by:
     
    cosσ2(P,P′)
=
1
1−σ tanσ2(P,P′)
         
 
=
(uu′−σ vv′)2
(uu′−σ vv′)2 −σ(u  v′− v u′)2
=
(uu′−σ vv′)2
(u2−σ v2)(u2−σ v2)
 ,
        (3)
sinσ2(P,P′)
=
tanσ2(P,P′)
1−σ tanσ2(P,P′)
   
         
 
=
(u  v′− vu′)2
(uu′−σ vv′)2 −σ(u  v′− vu′)2
=
(u  v′− vu′)2
(u2−σ v2)(u2−σ v2)
 .
        (4)

As a consequence we have the following EPH form of the Pythagoras identity, cf. the identity at (a):

cosσ2(P,P′)−σ sinσ2(P,P′)=1. (5)
Exercise 25Let P and P be two vectors tangent to σ2 at ι. Denote by P and P′′ images of P and P under the action of the matrix (
    aσ b
    ba
) from the corresponding isotropy subgroup, see (a). Then:
  1. The angle between a vector and its image is independent from the vector:
          tanσ(P,P′)=
    2ab
    a2+σ b2
     .   
  2. The angle between vectors is preserved by the transformation: tanσ(P,P′)=tanσ(P′,P′′). Since cosine and sine are functions of the tangent they are preserved as well

Exercise 26 Show that the angle between any two tangent vectors at any point of σ2 is preserved by an arbitrary Möbius transformation from SL2(ℝ). Hint: Use a combination of Lem. 23 and Exercise 2.

We continue to use the standard notation tan and tanh for the respective elliptic tane and hyperbolic tanh functions, similarly for sines and cosines.


1
For the subgroup A′ this will be refined in the subsection 3.3.3.
site search by freefind advanced

Last modified: October 28, 2024.
Previous Up Next