ANNALS OF PHYSICS 99, 188-195 (1976)

Transformation Laws of the Newman-Penrose Field Variables*

M. CARMELI AND M. KAYE

Department of Physics, Ben Gurion University of the Negev, Beer Sheva 84120, Israel
Received October 1, 1975

The transformation laws of all the Newman-Penrose gravitational field variables

under the six parameter proper, orthochronous, Lorentz group are given. The method
used here utilizes results obtained by Carmeli in his gauge theoretic formulation of the
gravitational field equations.

I. INTR~DU~T~~N

The null tetrad formalism of Newman and Penrose [I] (hereafter abreviated
to NP) has proven to be of extreme value in theoretical investigations in the theory
of general relativity. These investigations are carried out in coordinate systems
adapted to the particular problem. However, the null tetrad vectors are defined
only up to the three (complex) parameter proper, orthochronous, Lorentz group
or X(2, C), its covering group. This freedom of choice of the null tetrad is exploited
in order to simplify the field equations and is therefore of considerable value.
Not all the transformation laws of the NP field variables under the group
SL(2, C) have been presented, so far, in the literature [2]. In Section II we derive
these transformation laws by utilizing results obtained by Carmeli [3] in his gauge
theoretic formulation of the gravitational field equations. In Section III the resulting
transformation laws are written out explicitly. For completeness, and for the
convenience of the reader, we give the transformation laws for the total set of the
NP variables, including those that have been previously found. It is worthwhile
pointing out that the method utilized in the present work proved to lead to a large
reduction in the calculations as compared with the usual approach.

II. THE NP FIELD VARIABLESAND THEIR TRANSFORMATIONPROPERTIES

The NP field variables were defined in a gauge theoretic framework by Carmeli

in the following way [2, 31. At each point of spacetime two 2-component spinors
* This research was supported in part by the Israel National Academy of Sciences, Grant
No. 197 (B) P.
188
Copyright 0 1976 by Academic Press, Inc.
All rights of reproduction in any form reserved.
 NEWMAN-PENROSE FIELD VARIABLES 189

5” a , where a = 0, 1, are introduced to define a spin frame. Here A = 0, 1 are

spinor indices whereas lower case Latin letters are dyad indices. They are nor-
malized such that caBeBAcbA= E,~ , where the B’S are the skew symmetric Levi-
Civita symbols. The vectors (IS,): are now defined by taking the covariant
derivatives of caA,
VuSaA = @,>a” 5bA, (2.1)
or, in matrix notation
V,t = B,L (2.2)
where B, and 5 are 2 x 2 complex matrices whose elements are (B,J,b and caA,
respectively. The normalization condition imposed on the spinors SaAthen implies
that B, be traceless and 5 be unimodular.
Next, a 2 x 2 complex traceless matrix Fpy whose elements are skew-symmetric
tensors is defined by applying the commutator of the covariant derivative
(V,V, - V,V,,) on 5 giving Fuy(, where
Fuv = Vu - UA + P, ,&I, (2.3)
and [B,, , B,] = B,B, - B,B, . Finally, the correspondence between tensors and
spinors is obtained by means of four 2 x 2 hermitian matrices, denoted by Z& ,
which satisfy the relation
g,,5~**6~Dt = EAcEB’I)’ . (2.4)
Another set of hermitian matrices, u p, is now defined by the following relation
u” = &wy, (2.5)
where t denotes hermitian conjugation.
The elements of the matrix uU define a null tetrad of vectors where a&,, and urI,
are real, whereas u&, and a& are complex and conjugate to each other. Moreover,
they satisfy the orthogonality relation u$,,uUcd, = ~~~~~~~~and thus are the null
vectors introduced by NP, where ut;o, = 111,a&, = mu, u;b, = w and u&, = nu.
The dyad components of the matrices B, and F,, are given by
Bab’ = &,tBu , (2.6)
F ab’cd’ = u:bdk~w , (2.7)

where, the four matrices Bab, have the form

(2.8)
190 CARMELI AN!3 KAYE

the twelve complex functions E, K, T, etc., are the NP spin coefficients. The six
matrices Fagod,have the form

F01'00' =
54 -1cr,
[ 42+2A -41' I
F11'10' = ;; -+:;32A],
l

F10'00' =
[ g:: $j~

F11'01' =
IIif 22a :$;jg
$2 + 411- A
F11'00' =
[ +3 + 941
-A
-A-$ll+A'
- $01
1
F10'01' $1 - do1
=
[ -Y 3 k$+ 21 A A - 411 - A I '

where tie ,..., & are the complex tetrad components of the Weyl tensor, &,,,
are the complex tetrad components of the trace free Ricci tensor, and A is pro-
portional to the Ricci scalar.
Under a change of spin frame

5 = SC’ (2.10)

where S is an element of the group SL(2, C), the B, and Fuytransform as follows:

B,' = S-lB,S - S-%,S, (2.11)

FL, = S-1F$3. (2.12)

These are well-known results in gauge theory and were shown by Carmeli to
be valid in gravitation too. For convenience we introduce the matrix g = S-l.
From the definition of the null tetrad vectors and Eq. (2.10) we find that the null
tetrad transforms as follows under an SL(2, C) transformation:
cl;;’ = g,“a;f,g+f d’ . (2.13)
Similarly

(BEdab = gcegtf’d,[gaVb,)~* k-l>,b - gaD@,,*g-l),bl (2.14)

(F;b’cd’); = &“gtf’b’ .dktn’& gp’V’efwLS (flhq. (2.15)

 NEWMAN-PENROSE FIELD VARIABLES 191

If the general matrix g of the group SL(2, C) is denoted by

A B
ga b- 4 CD 1 3
(2.16a)
then
W’),” = [-“c -A”], (2.16b)
and
(g+)“’ b’ =
L 1
; ; . (2.16~)

The matrix g can be written as a product of three matrices of the form [4, 51

&>= [z0 z-1

O1' (2.17)

where z is a complex parameter. The transformation gl(z) leaves the spinor tOA,
and hence the null vector 1, , invariant. It is called a one- (complex) parameter null
rotation about Z, . The transformation g3(z) is also a one-parameter null rotation,
but about the vector n, . The transformation gZ(z) corresponds to an ordinary
Lorentz transformation (boost) in the 1, - n, plane, along with a spatial rotation
in the m, - FE, plane.

III. EXPLICIT EXPRESSIONS FOR THE TRANSFORMATION LAWS

Using the transformation equations (2.13)-(2.15) and the explicit forms of the
matrices B and F as given in (2.8) and (2.9), we find the following results. In the
following we have put 5,” = IA and &” = n A, and have used the accepted notation
for the intrinsic derivatives D = aoO,, 6 = a,,,, , 8 = a,,r and d = a,,* , where
a,,’ = ~$,?a, .
A. Null Rotation about I,,
Here one has 5’ = g,(z)c. Hence

(3.1)

and the null tetrad transforms as follows:

1,’ = I, ,
mu‘=Zu+mrr, (3.2)
n,’ = z.51, + zm, + Eii, •k n, .

595/99/I-13
192 CARMELI AND KAYE

The spin coefficients transform as follows:

p’ = p + ZK,
CII’ = 01 + “(p + C) + Z’K, (3.3a)
A’ = h + z(r + 201) + z”(p + 2~) + Z3K + 8~ + zDz,
K’ = K,

E’ = E + ZK, (3.3b)
TT’ = r + ~ZE + Z2K + Dz,
CT’ = CT + ZK,

,8’ = ,8 + ZU + FE + z%K, (3.3c)

/.i = p + 2zp + z7r + z2u + 2~2~ + Z2i?K + 6.%’ + ZDz,
7’ = 7 + ZU + -zp + zs?K,

y’ = y + Z(’ + ,k?) + %X + Z2U + Zz@ + E) + Z’ZK,

(3.3d)
v’ = v + z(p + 27) + a + 22(2/l + T) + zZ(2a + T) + Z2P(2E + p)
+ z3u + Z3i!K + AZ + z 6z + Z 8z + zn?!Dz.

The tetrad components of the Weyl tensor transform as follows:

$2’ = z2*o + 24 + $2 ) (3.4)

*3’ = z3*ll + 3% + 3z$b, + $h3)
#a’ = Z”&I + 4z3#1 + 6z2#, + 4z& + a,h4,

and the components of the trace-free Ricci tensor transform as follows:

 NEWMAN-PENROSE FIELD VARIABLES 193

B. Boost in l”-n” Plane and Spatial Rotation in m+E” Plane

Here one has 5’ = g,(z)[. The complex parameter z can be written in the form
z = A1lze~Ojz,where A and 0 are real. Then A is the boost parameter and 0 the
spatial rotation parameter. Hence

I*’ = zl* )
(3.6)
nA’ = z-k*,

and the null tetrad transforms as follows:

I,’ = z.21, = Al, ,

m,’ = .+lm bl = &em (3.7)
n,’ = z-lpn LL= A-In: .

The spin coefficients transform as follows:

p’ = zzp,
a’ = z-lZ(cY - zBz-l), (3.8a)
A’ = z-?a,
Id = z%lc 9
E’ = ZZ(E - zDz-l), (3.8b)
& = z-lZrr
u’ = zSI-la,
8 = zz-y/c? - zsz-l), (3.8~)
p’ = z-15-l/J,
7’ = ZF+T,
y’ = z-?i?l(y - zdz-‘), (3.8d)
y) = z"~-1"

The tetrad components of the Weyl tensor transform as follows:

(3.9)
194 CARMELI AND KAYE

and the tetrad components of the trace-free Ricci tensor transform as follows:

(3.10)

C. Null Rotation about n,

Here one has 5’ = g3(z)<. Hence
IA’ = IA + znA ,
(3.11)
nA’ = nA ,

and the null tetrad transforms as follows:

I,’ = I, + c-l, + ziii, + zzn, )

mu’ = m, f zn, , (3.12)

~1,’ = n, .
The spin coefficients transform as follows:
,,’ = ,.I + 2zor + 2-r + z2h + 2zzy + z2Zv - 8, - TAZ,
a’ = (II + zx + zy + zzv, (3.13a)
A’ = A + Iv,
K’ = Z(fJ + 2E) + ik + z2(2cu.+ 77) + d(2p
K + + T) + z22(2y + ,..‘)
+ z3h + 231~ - Dg i z8z - 2%~ - zzdz,
E’ = E + z( 7r + a) + z/3 + z2h + zqp + y) + z2zv, (3.13b)
d = n + 2x + zp + dv,
u’ i u -t: Z(T + 2/3) + ‘z”(/.L + 2y) + z3v - 6z - zdz,
p’ = p + 4r.L + Y) + z2v, (3.13c)
p’ = p + zv,
T’ = T + 2zy + zzv - AZ,
y’ = y + zv, (3.13d)
v’ = v,
 NEWMAN-PENROSE FIELD VARIABLES 195

The tetrad components of the Weyl tensor transform as follows:

#o’ = &.I+ 4Z& + 6Z2& + 4~~43 + ~~$4,

9%’ = $I+ 3Z#, + 3Z2#, + z~$~,

$2’ = $2 + 2% + z2#4, (3.14)

#31 = $3 + q54,

$4) = #4,

and the tetrad components of the trace-free Ricci tensor transform as follows:

&I = hl + 2% + 2Z&l + 48z&, + 22+,, + z”$20 + 2222c$,,

t 222&, + 2222+,, )

(3.15)

In each case the Ricci scalar remains, of course, unchanged, A’ = A.

REFERENCES

1. E. NEWMAN AND R. PENROSE,J. Math. Phys. 3 (1962), 566; 4 (1963), 998.

2. M. CARMELI, SL(2, C) Symmetry of the gravitational field, in “Group Theory in Nonlinear
Problems,” Proceedings of the 1972 NATO Advanced Study Institute on Mathematical
Physics, held in Istanbul Turkey (A. 0. BARUT, Ed.), p. 59, D. Reidel, Dordrecht, 1974;
B. ARONSON AND E. T. NEWMAN, J. Math. Phys. 13 (1972), 1847.
3. M. CARMELI, General Relativity and Gravitation 5 (1974), 287.
4. M. A. NAIMARK, “Linear Representations of the Lorentz Group,” Pergamon Press, New
York, 1964.
5. I. M. GEL’FAND, M. I. CRAW, AND N. YA VILENKIN, “Generalized Functions, Vol. 5: Integral
Geometry and Representation Theory,” Academic Press, New York and London, 1966.