Documente Academic
Documente Profesional
Documente Cultură
104003-2
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
This allows us to give a covariant definition to the function the metric perturbation. The tensorial nature of the spheri-
f that appears in Eqs. (2.2), (2.3), and (2.4): cal harmonics refers to the unit two-sphere, and in this
2M section we use the metric AB and its inverse AB to lower
ra ra : f 1 : (2.6) and raise all upper-case Latin indices. All tensorial opera-
r tions (including covariant differentiation) shall refer to this
We also introduce "ab , the (antisymmetric) Levi-Civita metric.
tensor on M2 ; in the coordinates of Eqs. (2.2), (2.3), and The scalar harmonics are the usual spherical-harmonic
(2.4), "tr "ur "vr 1. The timelike Killing vector of functions Y lm A . They satisfy the eigenvalue equation
the Schwarzschild spacetime is tangent to M2 and is given AB DA DB ll 1Y lm 0.
by Vectorial spherical harmonics come in two types. The
even-parity harmonics are defined by
ta "ab rb ; (2.7)
a
in the coordinates of Eqs. (2.2), (2.3), and (2.4), t 1; 0. YAlm : DA Y lm ; (3.1)
We have ta ta f and ta ra 0, and the vectors ra , ta while the odd-parity harmonics are
form a basis on M2 . In terms of this basis we have gab
f1 ta tb ra rb and "ab f1 ta rb ra tb . XAlm : "A B DB Y lm : (3.2)
The covariant derivative operator compatible with gab is
denoted ra ; we thus have ra gbc
0. It is easy to show Their components are listed explicitly in Appendix A. The
that for the Schwarzschild solution, vectorial harmonics satisfy the orthogonality relations
Z 0 0
M Y Alm YAl m d ll 1ll0 mm0 (3.3)
ra rb r
gab ; (2.8)
r2
so that 䊐r 2M=r2 , where 䊐 : gab ra rb is the and
Laplacian operator on M2 . We also have ra tb Z 0 0
M=r2 "ab , which confirms that ta is a Killing vector. X Alm XAl m d ll 1ll0 mm0 ; (3.4)
The Riemann tensor on M2 is Rabcd 2M=r3 gac gbd
gad gbc . in which an overbar indicates complex conjugation and
We let AB be the inverse to AB , the metric on the unit d : sindd is an element of solid angle. We also
two-sphere. The covariant derivative operator compatible have
with AB is denoted DA ; we thus have DA BC
0. The Z 0 0
Levi-Civita tensor on the unit two-sphere is denoted "AB , Y Alm XAl m d 0; (3.5)
and " sin. The Riemann tensor on the unit-sphere is
RABCD AC BD AD BD . which states that the even-parity and odd-parity harmonics
Covariant differentiation in the Schwarzschild space- are always orthogonal. The definitions (3.1) and (3.2) for
time can be defined in terms of covariant differentiation the vectorial spherical harmonics are identical to those
in the submanifolds M2 and S2 . If a bc is the connection provided by Regge and Wheeler [1].
associated with ra , and if A BC is the connection associ- Tensorial spherical harmonics come in the same two
ated with DA , then it is easy to show that the nonvanishing types. The even-parity harmonics are AB Y lm and
components of the spacetime connection are given by
4 a a 4 a rra , 4 A 1 A 1
bc bc , Bc r rc B , YAB DA DB ll 1AB Y lm ;
lm :
BC BC (3.6)
4 A A
and BC BC . Using these rules we find that the 2
conservation identities for a stress-energy tensor T !" in while the odd-parity harmonics are
the Schwarzschild spacetime take the form
lm : 1
2 XAB "A C DB "B C DA DC Y lm : (3.7)
rb T ab DB T aB rb T ab rra AB T AB 0 (2.9) 2
r
Their components are listed explicitly in Appendix A. The
and tensorial harmonics satisfy the orthogonality relations
4 Z
ra T aA DB T AB ra T aA 0 (2.10) l0 m0 1
r Y AB
lm YAB d l 1ll 1l 2ll0 mm0 (3.8)
2
when expressed in terms of the submanifold connections.
and
Z 1
III. SPHERICAL HARMONICS X AB l0 m0
lm XAB d l 1ll 1l 2ll0 mm0 : (3.9)
2
In this section we introduce the scalar, vector, and tensor
spherical harmonics that are used in the decomposition of We also have
104003-3
KARL MARTEL AND ERIC POISSON PHYSICAL REVIEW D 71, 104003 (2005)
Z X
Y AB l0 m0 a )lm lm
a Y ; (4.4)
lm XAB d 0 (3.10)
lm
and X
A )lm YAlm ; (4.5)
lm
AB YAB
lm 0 AB XAB
lm : (3.11)
the fields )lm
a and )
lm depend on the coordinates xa only.
The definition (3.6) for the even-parity harmonics does not Under such a transformation the perturbation quantities
agree with that of Regge and Wheeler [1], who work change according to (see Appendix B)
instead with the set AB Y lm and DA DB Y lm . We find it
more convenient to form the tracefree combinations YAB lm
, hab ! h0ab hab ra )b rb )a ; (4.6)
which have the property of being (pointwise) orthogonal to
AB Y lm . The definition (3.7) for the odd-parity harmonics 2
also differs from Regge and Wheeler’s, but only by an ja ! j0a ja )a ra ) ra ); (4.7)
overall minus sign, which we find convenient to introduce. r
lm lm
The tensorial harmonics YAB and XAB can be related to
the spherical-harmonic functions of spin-weight s
2 ll 1 2
K ! K0 K ) ra ) a ; (4.8)
[28], and to the pure-spin harmonics used by Thorne [29]. r2 r
These relations are explored in Appendix A.
2
G ! G0 G ); (4.9)
IV. EVEN-PARITY SECTOR r2
A. Perturbation fields and gauge transformations where we have discarded the spherical-harmonic labels for
The even-parity sector refers to those components of the brevity (we shall continue with this practice until the end of
metric perturbation that can be expanded in terms of the the section). It is easy to show that the combinations
even-parity spherical harmonics Y lm , YAlm , AB Y lm , and
lm
YAB . Introducing the notation 4 gab gab pab , 4 gaB h~ ab : hab ra "b rb "a (4.10)
paB , and 4 gAB r2 AB pAB for the perturbed metric,
the even-parity sector of the metric perturbation is and
X
pab hlm lm
ab Y ; (4.1) ~ : K 1 ll 1G 2 ra "a
K (4.11)
lm 2 r
104003-4
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
then be extracted by involving the orthonormality relations allows us to make the substitutions hab ! h~ab , K ! K
~ and
(3.3) and (3.8) satisfied by the spherical harmonics. At the therefore to express the gauge-invariant Einstein tensor in
end of this calculation we take advantage of the fact that terms of gauge-invariant quantities.
h~ab hab and K ~ K in the Regge-Wheeler gauge. This Our final results are
1 and
Qa rc h~ca ra h~ ra h~ ra K;
~ (4.14)
r 2 l 1l 2 ]
ra Qa ra Qa Q Q[ 0: (4.22)
r 2r2
2 1
Q[ 䊐h~ ra rb h~ab ra rb h~ab ra ra h~
r r
C. Master equation
ll 1 ~ ~ 2 a ~
h K r ra K; (4.15) The Zerilli-Moncrief function is defined by
2r2 r
2r ~ lm 2 a b ~lm ~ lm
~ lm : K r r h rrar K ;
Q] h; (4.16) even
ll 1 ! ab a
104003-5
KARL MARTEL AND ERIC POISSON PHYSICAL REVIEW D 71, 104003 (2005)
a consequence of the field equations, the even-parity mas- perturbation, the monopole component of the metric per-
ter function satisfies the Zerilli equation turbation represents a shift in the mass parameter of the
Schwarzschild solution.
䊐 Veven even Seven ; (4.25)
The even-parity spherical harmonics for l 1 are Y 1m ,
with potential YA , and AB Y 1m ; YAB
1m 1m vanishes. The metric perturbation is
then
1 2 /2 6M 36M2 2M X
Veven / / pab h1m 1m
(4.31)
!2 r2 r3 r4 r ab Y ;
m
(4.26)
X
and source term paB j1m 1m
a YB ; (4.32)
m
4 1 2
Seven ra Qa Q] X
! r / 2! pAB r2 AB K 1m Y 1m ; (4.33)
24M m
2r2 ra ra Q r r Qab 2rfQ[
! a b and the fields G1m are not
P defined. Gauge transformations
P
r M M2 are generated by a m )1m a Y 1m
and A m )1m YA1m .
// 2 12/ 3 84 2 Q ; The perturbations change according to
! r r
(4.27) hab ! h0ab hab ra )b rb )a ; (4.34)
where / : l 1l 2 and Q : gab Qab . The validity 2
of Eqs. (4.25), (4.26), and (4.27) can be verified by brute- ja ! j0a ja )a ra ) ra ); (4.35)
r
force evaluation of both sides of Eqs. (4.25). The general
source term for the covariant Zerilli equation has never 2 2
been presented in explicit form in the literature. We display K ! K0 K ) ra ) a : (4.36)
r2 r
it here for the first time, but note that Eq. (A14) of Ref. [16]
gives an implicit expression for the source term, while their There is no analogue here of the gauge-invariant fields h~ab
Eqs. (A15)–(A17) give it explicitly in Schwarzschild and K ~ that were introduced in the general case. The rele-
coordinates. vant field equations for hab , ja , and K are the Qab , Qa , Q[
equations of Eqs. (4.13), (4.14), and (4.15), in which
D. Low multipoles we set l 1, h~ab hab ra jb rb ja , and K ~
1 a ]
K 2r r ja . The Q equation of Eq. (4.16) is not defined
To conclude our presentation of the even-parity sector for l 1. It is well known [3] that in the case of a vacuum
we now handle the special cases l 0 and l 1. perturbation, the dipole component of the even-parity met-
Additional details can be found in Appendix G of the paper ric perturbation is pure gauge: it can always be removed by
by Zerilli [3]. a gauge transformation.
When l 0 the only relevant spherical harmonic is Y 00 ,
which is a constant. It follows that YA YAB 0, and the
only nonvanishing metric perturbations are V. ODD-PARITY SECTOR
A. Perturbation fields and gauge transformations
pab hab Y 00 ; pAB r2 KAB Y 00 ; (4.28)
The odd-parity sector refers to those components of the
the fields ha and G are not defined. The freedom to perform metric perturbation that can be expanded in terms of the
a gauge transformation is contained in a )a Y 00 , A odd-parity spherical harmonics XAlm and XABlm
. Recalling the
0, and the perturbations transform as notation gab gab pab , gaB paB , and 4 gAB
4 4
hab ! h0ab hab ra )b rb )a ; (4.29) r2 AB pAB for the perturbed metric, the odd-parity sec-
tor of the metric perturbation is
2 pab 0; (5.1)
K ! K 0 K ra ) a : (4.30)
r
X
There is no analogue here of the gauge-invariant quantities paB hlm lm
a XB ; (5.2)
h~ab and K~ that were introduced in the general case. The lm
relevant field equations for hab and K are the Qab and Q[ X
equations of Eqs. (4.13) and (4.15), respectively, in which pAB hlm lm
2 XAB : (5.3)
lm
we set l 0, h~ab hab , and K ~ K. The Qa and Q]
equations of Eqs. (4.14) and (4.16), respectively, are not In most of this section the sums over l are taken to begin at
defined. It is well known [3] that in the case of a vacuum l 2; there is no odd-parity perturbation with l 0, and
104003-6
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
the case l 1 will be considered separately in Sec. V E. Our final results are
The fields hlm lm 2
a and h2 are defined on M and they depend 2
a
on the coordinates x only. They are closely related to the Pa 䊐h~a ra rb h~b rb ra h~b ra rb h~b
quantities first introduced by Regge and Wheeler [1], who r
worked exclusively in terms of the usual Schwarzschild 2 ll 1 ~
2 ra rb h~b ha ; (5.8)
coordinates (t, r). In these coordinates (discarding r r2
spherical-harmonic labels) we have ht h0 and hr h1 .
Except for a sign—refer back to the discussion following P ra h~a ; (5.9)
Eq. (3.11)—the function h2 is identical to the correspond-
where 䊐 : gab ra rb . The source terms are
ing Regge-Wheeler quantity.
Odd-parity gauge transformations are generated by a 16*r2 Z aB lm
dual vector field ! 0; A that is expanded as Pa T XB d; (5.10)
ll 1
X
A )lm XAlm ; (5.4) Z
16*r4
lm P T AB X lm
AB d: (5.11)
l 1ll 1l 2
in which )lm depends on the coordinates xa only. Under
such a transformation the perturbation quantities change In Eqs. (5.10) and (5.11) the stress-energy tensor is imag-
according to (see Appendix B) ined to be given in fully contravariant form; T aB and T AB
are its relevant components in the spacetime coordinates
2 (xa , A ). In the event where the stress-energy tensor would
ha ! h0a ha ra ) ra ); (5.5) be given in covariant or mixed form, its indices would
r
have to be raised with (gab , r2 AB )—the inverse
Schwarzschild metric —before evaluating the source
h2 ! h02 h2 2); (5.6) terms. In Eqs. (5.8) and (5.9) all lower-case Latin indices
are lowered and raised with gab and gab , respectively.
where we have discarded the spherical-harmonic labels for
The perturbation equations are not all independent. By
brevity (we shall continue with this practice until the end of
virtue of the Bianchi identities, or the conservation
the section). It is easy to show that the combinations
Eq. (2.10), they are related by
1 1 2 l 1l 2
h~ a ha ra h2 ra h2 (5.7) ra Pa ra Pa P 0: (5.12)
2 r r r2
are gauge invariant. Equation (5.6) reveals that one can
always choose a gauge in which h2 0; this is the Regge- C. Master equation
Wheeler gauge. Equation (5.7) implies that h~a ha in the
Regge-Wheeler gauge. The Cunningham-Price-Moncrief function is defined by
2r ~ lm 2 ~lm
lm : "ab
r h
a b r h ; (5.13)
B. Perturbation equations odd
l 1l 2 r a b
The Ricci tensor of the Schwarzschild spacetime van- where "ab is the Levi-Civita tensor on the submanifold
ishes, and as a consequence its perturbation is gauge- M2 . Apart from a different normalization factor, this is a
invariant. Its computation can therefore be carried out in covariant generalization (first formulated by Gerlach and
any convenient gauge, and as in the preceding section we Sengupta [10]) of the definition provided by Cunningham,
shall adopt the Regge-Wheeler gauge. We substitute Price, and Moncrief [24], who worked with the usual
Eqs. (5.1), (5.2), and (5.3), having set h2 0, into the Schwarzschild coordinates. Our definition (and normaliza-
Ricci tensor of Appendix B. We simplify the result and tion) agrees with the odd-parity function considered by
find that Rab vanishes, RaB is expanded in terms of XBlm , Jhingan and Tanaka [25], who also worked with the
and RAB in terms of XAB lm . From the Ricci tensor we
Schwarzschild coordinates, but in the frequency domain.
compute the Einstein tensor and set the result equal to (The plethora of odd-parity functions and normalization
8*T !" . Each spherical-harmonic component of the field conventions is conveniently catalogued in the review ar-
equations can then be extracted by involving the orthonor- ticle by Nagar and Rezzolla [17].) The normalization
mality relations (3.4) and (3.9) satisfied by the odd-parity adopted in Eq. (5.13) will be seen to be convenient when
harmonics. At the end of this calculation we take advan- we discuss gravitational radiation at future null infinity
tage of the fact that h~a ha in the Regge-Wheeler gauge. (Sec. VI) and at the horizon (Sec. VII); our definition of
This allows us to make the substitution ha ! h~a and there- the odd-parity master function is well adapted to the de-
fore to express the gauge-invariant Einstein tensor in terms scription of radiation fields. It is noteworthy that it can also
of gauge-invariant quantities. be expressed as
104003-7
KARL MARTEL AND ERIC POISSON PHYSICAL REVIEW D 71, 104003 (2005)
2r 2 䊐 Vodd RW SRW ; (5.19)
lm
odd
: "ab @a hlm
b ra h lm ;
b
l 1l 2 r with the potential of Eq. (5.15) and a new source term given
in terms of the original (gauge-dependent) perturbation by
quantities, and in terms of the partial differentiation opera-
1 2 3M
tor; the function is nonetheless gauge-invariant and a SRW ra ra P Pa 1 P : (5.20)
r r r
scalar.
The perturbation Eqs. (5.8) and (5.9) give rise to a wave This covariant form for the source term is also a new result.
equation for the function odd (we resume our practice of
discarding the spherical-harmonic labels). As in the pre- E. Low multipoles
ceding section we simply state the final result: As a con- To conclude our presentation of the odd-parity sector
sequence of the field equations, the odd-parity master we now handle the special cases l 0 and l 1.
function satisfies the Regge-Wheeler equation Additional details can be found in Appendix G of the paper
䊐 Vodd odd Sodd ; (5.14) by Zerilli [3].
When l 0 the only relevant spherical harmonic is Y 00 ,
with potential which is a constant. It follows that XA XAB 0, and
ll 1 6M we conclude that there is no odd-parity perturbation with
Vodd 3 (5.15) l 0.
r2 r
The only surviving odd-parity spherical harmonics for
and source term l 1 are XA1m , which are obtained from Y 1m using
1m
2r Eq. (3.2). The tensorial harmonics XAB vanish, and the
Sodd "ab ra Pb : (5.16) only surviving components of the metric perturbation are
l 1l 2 X
paB h1m 1m
a XB : (5.21)
The validity of Eqs. (5.14), (5.15), and (5.16) can be m
verified by brute-force evaluation of both sides of
Eqs. (5.14). The general source term for the covariant The perturbations h1m
acan be altered by
P a gauge trans-
Regge-Wheeler equation has never been presented in ex- formation generated by a 0, A m )1m XA1m ; they
plicit form in the literature; it is given only implicity by change according to
Eq. (17) of Ref. [10]. We display it here for the first time, 2
and note that in the usual Schwarzschild coordinates (and ha ! h0a ha ra ) ra ): (5.22)
r
in the frequency domain), Sodd agrees with the source term
presented in Eq. (18) of Jhingan and Tanaka [25]. There is no analogue here of the gauge-invariant fields h~a
that were introduced in the general case. The relevant field
D. Regge-Wheeler function equation for ha is the Pa equation of Eq. (5.8), in which we
set l 1 and h~a ha . The P equation of Eq. (5.9) is not
The Cunningham-Price-Moncrief function is a close defined. It is well known [3] that in the case of a vacuum
cousin to the more familiar Regge-Wheeler function [1], perturbation, the dipole component of the odd-parity met-
whose covariant and gauge-invariant definition is ric perturbation represents a shift (away from zero) in the
1 a ~lm angular-momentum parameter of the black hole.
lm :
RW r ha : (5.17)
r
VI. RADIATION AT FUTURE NULL INFINITY
As we shall see in Secs. VI and VII, the Regge-Wheeler
function is not well suited to describe the gravitational To examine the gravitational perturbations near future
radiation field, and in this paper we adopt the function null infinity we adopt the retarded coordinates (u, r, , )
lmodd of Eq. (5.13) as the fundamental odd-parity master
and express the two-dimensional Schwarzschild metric in
function. It is straightforward to use the perturbation equa- the form of Eq. (2.3). In these coordinates, future null
tions to show that these functions are related by infinity corresponds to taking the limit r ! 1 keeping u
fixed, and our strategy will be to expand the metric pertur-
1 r bations in powers of r1 . In asymptotically Cartesian
RW ta ra odd r Pa ; (5.18)
2 l 1l 2 a coordinates the radiative part of the metric would scale
where ta "ab rb is the Killing vector of Eq. (2.7). as r1 ; transforming to spherical coordinates produces the
Outside of sources, and apart from a factor of one-half, scalings
the Regge-Wheeler function is the time derivative of the prad 1
prad 0
prad
ab Or ; aB Or ; AB Or
Cunningham-Price-Moncrief function.
(6.1)
The function of Eq. (5.17) also satisfies the Regge-
Wheeler equation, for the radiative part of the metric perturbations. Our goal
104003-8
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
X
is to isolate this, and to calculate how much energy and prad;even r lm lm
even u; r 1YAB : (6.8)
AB
angular momentum is carried away by the radiation. We lm
find it advantageous to work in gauge in which
It is obtained by integrating the Zerilli Eq. (4.25) and
ta pab 0 ta paB ; (6.2) evaluating the Zerilli-Moncrief function at r 1.
We turn next to the odd-parity sector of Sec. V. The
where ta is the Killing vector of Eq. (2.7). In spite of the gauge conditions imply hu 0, and the scalings of
fact that ta is not a null vector (except on the even horizon, Eq. (6.1) imply that we are looking for the r0 part of hr
which is well outside our domain of consideration), this and the r1 part of h2 . These are determined by following
gauge happens to be a perfectly respectable ‘‘radiation the same procedure as in the even-parity case, and we
gauge.’’ obtain
In this section we take l 2; as is well known, the low
multipoles l 0 and l 1 do not contain radiative de- l 1l 2 Z u
hr bu0 du0 Or2 ; (6.9)
grees of freedom. We assume that the matter distribution 2r
responsible for the radiation is confined to a bounded
volume, and that our domain of consideration is outside h2 bur Or0 ; (6.10)
this volume; we shall therefore be solving the vacuum field
equations. where bu is not determined by the vacuum field equa-
We begin with the even-parity sector of Sec. IV. The tions. We see that the radiative part of the perturbation is
gauge conditions imply huu hur ju 0, and the scal- contained entirely in the function h2 u; r. With the metric
ings of Eq. (6.1) imply that we are looking for the r1 part perturbation of Eqs. (6.9), (6.10) we may evaluate the
of hrr , K, and G, as well as the r0 part of jr . These can be Cunningham-Price-Moncrief function of Eq. (5.13). The
determined by expanding the components of the metric result is
perturbation in powers of r1 and substituting them into odd bu Or1 ; (6.11)
the field equations of Eqs. (4.13), (4.14), (4.15), and (4.16).
The expansions for hrr , K, and G begin at order r1 and and we conclude that the radiative part of the odd-parity
each coefficient is a to-be-determined function of u; the sector is given by
expansion for jr begins instead at order r0 . The field X
prad;odd r lm lm
odd u; r 1XAB : (6.12)
equations return Qab , Qa , Q[ , and Q] expanded in powers AB
lm
of r1 , and setting each coefficient to zero determines the
metric perturbation. We obtain It is obtained by integrating the Regge-Wheeler Eq. (5.14)
and evaluating the Cunningham-Price-Moncrief function
ll 1 Z u at r 1. Notice that by virtue of Eq. (5.18), the radiative
hrr au0 du0 Or4 ; (6.3)
r3 field could instead be expressed in terms of the u-integral
of the Regge-Wheeler function. The need to perform this
au integration is inconvenient, and it is the simple relationship
jr Or2 ; (6.4) of Eq. (6.12) that has motivated the adoption of the
r Cunningham-Price-Moncrief function as the fundamental
odd-parity master function.
ll 1 Z u The full radiative field is obtained from Eqs. (6.8) and
K au0 du0 Or4 ; (6.5)
2r3 (6.12). We have prad rad
ab 0 paB and
X
prad
AB r lm lm lm lm
even YAB odd XAB ; (6.13)
2 _
au
G Or2 ; (6.6) lm
l 1l 2 r
where lm lm lm lm
even
even u; r 1 and odd
odd u; r
where au is a function that is not determined by the 1. As expected, the radiative field is transverse, and
vacuum field equations, and a_ : da=du. We see that the tracefree by virtue of Eqs. (3.11). The two fundamental
radiative part of the perturbation is contained entirely in polarizations of the gravitational wave can be defined by
the function Gu; r. With the metric perturbation of h : p =r2 and h : p =r2 sin. Using the compo-
Eqs. (6.3), (6.4), (6.5), and (6.6) we may evaluate the nents of the tensorial spherical harmonics listed in
Zerilli-Moncrief function of Eq. (4.23). The result is Appendix A, we obtain
2 1 X lm @2 1
even _
au Or1 ; (6.7) h ll 1 Y lm
l 1l 2 r lm even @2 2
and we conclude that the radiative part of the even-parity im @ cos lm
lmodd Y (6.14)
sector is given by sin @ sin
104003-9
KARL MARTEL AND ERIC POISSON PHYSICAL REVIEW D 71, 104003 (2005)
and port of energy and angular momentum across the horizon is
pAB . [Refer to Poisson’s Eq. (6.10), which establishes the
1 X lm im @ cos lm
h even Y equality between pAB and 61AB , the perturbation of the
r lm sin @ sin horizon’s intrinsic metric.]
2 In this section we take l 2, as the low multipoles l 0
lm @ 1
odd ll 1 Y lm : (6.15) and l 1 do not contain radiative degrees of freedom. We
@2 2
assume that the matter distribution responsible for the
The energy and angular momentum carried away by the radiation does not come near the event horizon; we shall
gravitational radiation can be calculated using the tech- therefore be solving the vacuum field equations in an
niques developed by Thorne [29]. We compare our empty neighborhood of the event horizon.
Eq. (6.13) to his Eq. (4.3), taking into account the relation- We begin with the even-parity sector of Sec. IV. The
ship between our tensorial harmonics and his pure-spin gauge conditions of Eq. (7.1) imply that hvv hvr jv
harmonics (this is spelled out in Appendix A). Thorne’s 0, so that hrr , jr , K, and G are the only nonvanishing
mass multipole moments are thus seen to be proportional to components of the metric perturbation. As stated above,
lm
even u; r 1, and his current moments are proportional the radiation crossing the event horizon is described en-
to lm odd u; r 1. Substituting these into Thorne’s tirely by K and G evaluated at r 2M. The field equa-
Eq. (4.16) we obtain tions, however, imply that K vanishes on the horizon. This
is verified by expanding hrr , jr , K, and G in powers of f
dE 1 X
l 1ll 1l 2 1 2M=r and substituting them into the field equations of
du 64* lm Eqs. (4.13), (4.14), (4.15), and (4.16). This calculation
hj_ lm 2 _ lm 2
even j jodd j i (6.16) reveals also that Gv; r 2M av, a function that is
not determined by the vacuum field equations. On the other
for the averaged rate at which the energy escapes to future hand, the expansions allow us to evaluate the Zerilli-
null infinity. Substituting instead into Thorne’s Eq. (4.23) Moncrief function of Eq. (4.23), and the result is even r
returns 2M 2Mav. We conclude that on the horizon, the
radiative part of the even-parity sector is given by
dJ 1 X
l 1ll 1l 2im X
du 64* lm prad;even 2M lm lm
even v; r 2MYAB : (7.2)
AB
lm _ lm lm _ lm lm
h even even odd odd i (6.17)
It is obtained by integrating the Zerilli Eq. (4.25) and
for the averaged rate at which the angular momentum
evaluating the Zerilli-Moncrief function at r 2M.
escapes to infinity. This is the component of the angular-
We turn next to the odd-parity sector of Sec. V. The
momentum vector in the arbitrary z direction which defines
gauge conditions imply hu 0, so that hr and h2 are the
the orientation of the angles and . The overbar indicates
only nonvanishing components of the metric perturbation.
complex conjugation, and it is not difficult to show that
The radiation crossing the event horizon is described en-
hdJ=dui is real. The averaging carried out in Eqs. (6.16)
tirely by h2 evaluated at r 2M, and the field equations
and (6.17) is over a characteristic time scale associated
imply that bv : h2 v; r 2M remains as an undeter-
with the gravitational wave.
mined function. They also imply that hr v; r 2M
cv, with
VII. RADIATION AT THE EVENT HORIZON
l 1l 2 Z v
To examine the gravitational perturbations near the cv : bv0 dv0 :
8M2
event horizon we adopt the advanced coordinates (v, r,
, ) and express the two-dimensional Schwarzschild These statements are verified by expanding hr and h2 in
metric in the form of Eq. (2.4). We want to calculate how powers of f 1 2M=r and substituting them into
much energy and angular momentum is transferred to the Eqs. (5.8) and (5.9). The expansions allow us also to
black hole by the perturbation, and we shall do so by evaluate the Cunningham-Price-Moncrief function of
following the methods devised by Poisson [31]. As in the Eq. (5.13), and the result is
preceding section we impose the gauge conditions
4M dc 1
odd r 2M bv:
ta pab 0 ta paB (7.1) l 1l 2 dv 2M
on the metric perturbation; ta is still the Killing vector of We conclude that on the horizon, the radiative part of the
Eq. (2.7). Recall that this vector is null on the event odd-parity sector is given by
horizon, and Eq. (7.1), evaluated at r 2M, is equivalent X
to Poisson’s Eq. (6.5). Poisson then shows that the part of prad;odd
AB 2M lm lm
odd v; r 2MXAB : (7.3)
the metric perturbation which is associated with the trans- lm
104003-10
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
It is obtained by integrating the Regge-Wheeler Eq. (5.14) 1 @ lm @ lm
and evaluating the Cunningham-Price-Moncrief function Xlm Y ; lm sin
X Y :
sin @ @
at r 2M. Notice that by virtue of Eq. (5.18), the radiative
field could instead be expressed in terms of the v-integral According to Eq. (3.6) we have
of the Regge-Wheeler function; this would give rise to 2
Poisson’s Eq. (7.3), a less convenient expression. lm @ 1
Y ll 1 Y lm ;
The full radiative field is obtained by adding Eqs. (7.2) @2 2
and (7.3). We have 2
lm @ cos @
X Y Y lm ;
prad lm lm lm lm @@ sin @
AB 2M even YAB odd XAB ; (7.4)
lm
2
lm @ @ 1 2
Y sin cos ll 1sin Y lm :
where lm
even
lm 2M and lm
even v; r
lm
odd
odd v; r @2 @ 2
2M. This result should be compared with Eq. (6.13). The
rates at which the gravitational perturbation transfers en- And according to Eq. (3.7) we have
ergy and angular momentum to the black hole can now be 2
lm 1 @ cos @
calculated using the method described in Sec. VII of X Y lm ;
sin @@ sin @
Poisson [31]. Our Eq. (7.4) replaces his Eq. (7.5), and the
2
rest of the calculation is identical. The final results are lm 1 sin @ 1 @2 @ lm
X cos Y ;
dE 1 X 2 @2 sin @2 @
l 1ll 1l 2 2
dv 64* lm lm sin @
X cos
@
Y lm :
@@ @
hj_ lm 2 _ lm 2
even j jodd j i (7.5)
The tensorial harmonics YABlm and X lm can be related to
and AB
the spherical-harmonic functions of spin-weight s
2
dJ 1 X [28]. Let mA and m
A be a complex orthonormal basis on the
l 1ll 1l 2im
dv 64* lm unit two-sphere, with mA 21=2 1; i sin. The relation-
lm _ lm lm _ lm ship is then
h even even odd odd i: (7.6)
lm 1 p
These equations replace Poisson’s Eqs. (7.8) and (7.9). YAB l 1ll 1l 2
Notice the similarity between Eqs. (7.5) and (6.16), and 2
between Eqs. (7.6) and (6.17). In Eq. (7.6), J represents the 2 Y lm mA mB 2 Y lm m B
Am
component of the hole’s angular-momentum vector in
and
the arbitrary z direction which defines the orientation of
the angles and . The overbar indicates complex con- lm i p
jugation. The averaging carried out in Eqs. (7.5) and (7.6) is XAB l 1ll 1l 2
2
over a characteristic time scale associated with the gravi-
2 Y lm mA mB 2 Y lm m Am B ;
tational perturbation.
where s Y lm are the spin-weighted spherical harmonics.
ACKNOWLEDGMENTS These equations can be compared with Eqs. (2.38e) and
We are grateful to Alessandro Nagar for his criticism of (2.38f) of Ref. [29]. This reveals that our tensorial har-
an earlier version of this paper. This work was supported monics are intimately related to the ‘‘pure-spin’’ harmon-
by the National Science and Engineering Research Council ics used by Thorne. The relationship is
of Canada. s
lm 1 E2;lm
YAB l 1ll 1l 2TAB
APPENDIX A: COMPONENTS OF THE VECTOR 2
AND TENSOR SPHERICAL HARMONICS, AND
and
RELATIONSHIP WITH OTHER SPHERICAL
HARMONICS s
lm 1 B2;lm
We first list the components of the tensorial harmonics XAB l 1ll 1l 2TAB :
2
introduced in Sec. III. According to Eq. (3.1) we have
E2;lm B2;lm
@ lm @ lm Notice that the pure-spin harmonics TAB and TAB are
Ylm Y ; Ylm Y : normalized on the unit two-sphere. Our convention here
@ @
differs from Thorne’s, who inserts a factor of r1 in mA and
According to Eq. (3.2) we have m A in order to normalize them on a two-sphere of radius r.
104003-11
KARL MARTEL AND ERIC POISSON PHYSICAL REVIEW D 71, 104003 (2005)
APPENDIX B: PERTURBATION OF THE RICCI riantly constant relative to the connection a bc ; for ex-
TENSOR FOR A GENERAL SPHERICALLY- ample ra AB
0.
SYMMETRIC BACKGROUND SPACETIME A straightforward calculation returns the components of
the perturbed connection, which we denote 4 ! "6
The perturbed spacetime metric is written as
! "6 . We obtain
4g gab pab ; 4g paB ;
ab aB
4 a bc Ca bc ;
gAB r2 AB pAB ;
1 1
where (gab , r2 AB ) are the components of the background abC DC pa b rb pa C ra pbC rb pa C ;
2 r
metric and (pab , paB , pAB ) are the components of the 1
perturbation. In this Appendix we allow the background a BC DB pa C DC pa B ra pBC rrm BC pam ;
2
metric to be completely general, so long as it is spherically
1
symmetric; we do not restrict it to be the Schwarzschild A bc 2 rb pc A rc pb A DA pbc ;
metric. We raise lower-case Latin indices with gab , the 2r
inverse to gab , and we raise upper-case Latin indices with 1 1
A bC 2 DC pb A DA pbC rb pA C 3 rb pA C ;
AB , the inverse to AB . The inverse perturbed metric is 2r r
thus 1 1
A BC 2 CA BC rm BC pmA ;
1 aB r r
4 ab
g gab pab ; 4 aB
g p ;
r2 where ra : ra r,
4 AB 1 1
g 2 AB 4 pAB ; 1
r r Ca bc : rc pa b rb pa c ra pbc ;
2
up to terms quadratic in the perturbations.
Covariant differentiation with respect to the coordinates and
xa on the submanifold M2 is indicated with ra :ra gbc 1
0. Covariant differentiation with respect to the coordinates CA BC : DC pA B DB pA C DA pBC :
2
A on the unit two-sphere is indicated with DA :DA BC
0. Quantities which depend only on xa are covariantly The Ricci tensor for the perturbed spacetime is equal to
constant relative to the connection A BC ; for example the background Ricci tensor R!" plus its perturbation
DA r
0. Quantities which depend only on A are cova- R!" . We obtain
2 1 1 1 1
Rab rm Cm ab rm Cm ab ra rb pm m 2 DM DM pab 2 DM ra pb M rb pa M 2 ra rb pM M
r 2 2r 2r 2r
1 1
3 ra rb pM M rb ra pM M 4 ra rb rra rb rpM M ;
2r r
1 1 1 1
RaB DB rm pm a ra pm m ra pm m 䊐paB rm ra pm B ra rm pm B rm ra pm B
2 r 2 r
1 1 1
2 ra rm rra rm rpm B 2 DM DB paM DM paB 2 ra DM pM B DB pM M
r 2r 2r
1
3 ra DM pM B DB pM M ;
r
1 1 1
RAB AB rrm rn pmn gmn pk k rm rn rrm rn rpmn DA DB pm m rm DA pm B DB pm A
2 2 2
1 1 1 1 1 1
rm AB DM pmM 䊐pAB 2 DM CM AB 2 DA DB pM M rm rm pAB AB pM M
r 2 r 2r r 2
2 m 1
2 r rm pAB AB pM M ;
r 2
where 䊐 : gab ra rb . These expressions can be simpli- Under a gauge transformation generated by the dual
fied by involving Eqs. (2.6) and (2.8) when the background vector field ! a ; A , the components of the metric
spacetime is the Schwarzschild spacetime. perturbation change according to
104003-12
GRAVITATIONAL PERTURBATIONS OF THE . . . PHYSICAL REVIEW D 71, 104003 (2005)
pab ! p0ab : pab ra b rb a ; It can be shown that when the background Ricci tensor
vanishes, Rab , RaB , and RAB are all invariant under this
2
paB ! p0aB : paB ra B DB a ra B ; transformation. We use this property in Secs. IV and V.
r
0
pAB ! pAB : pAB DA B DB A 2rrm m AB :
[1] T. Regge and J. A. Wheeler, Phys. Rev. 108, 1063 (1957). [16] A. Nagar, G. Diaz, J. A. Pons, and J. A. Font, Phys. Rev. D
[2] C. V. Vishveshwara, Phys. Rev. D 1, 2870 (1970). 69, 124028 (2004).
[3] F. J. Zerilli, Phys. Rev. D 2, 2141 (1970). [17] A. Nagar and L. Rezzolla, gr-qc/0502064.
[4] S. Chandrasekhar, The Mathematical Theory of Black [18] H. Kodama, A. Ishibashi, and O. Seto, Phys. Rev. D 62,
Holes (Clarendon, Oxford, 1983). 064022 (2000).
[5] V. P. Frolov and I. D. Novikov, Black Hole Physics: Basic [19] H. Kodama and A. Ishibashi, Prog. Theor. Phys. 110, 701
Concepts and New Developments (Kluwer, Dordrecht, (2003).
1998). [20] A. Ishibashi and H. Kodama, Prog. Theor. Phys. 110, 901
[6] K. Martel, Phys. Rev. D 69, 044025 (2004). (2003).
[7] M. Sasaki and H. Tagoshi, Living Rev. Relativity 6, [21] H. Kodama and A. Ishibashi, Prog. Theor. Phys. 111, 29
6 (2004), [Online article]: cited on http://www. (2004).
livingreviews.org/lrr-2003-6. [22] S. Kopeikin, J. Ramirez, B. Mashhoon, and M. Sazhin,
[8] R. Gleiser, C. Nicasio, R. Price, and J. Pullin, Phys. Rep. Phys. Lett. A 292, 173 (2001).
325, 41 (2000). [23] V. Mukhanov, H. Feldman, and R. Brandenberger, Phys.
[9] V. Moncrief, Ann. Phys. (N.Y.) 88, 323 (1974). Rep. 215, 203 (1992).
[10] U. H. Gerlach and U. K. Sengupta, Phys. Rev. D 19, 2268 [24] C. T. Cunningham, R. H. Price, and V. Moncrief,
(1979). Astrophys. J. 224, 643 (1978).
[11] U. H. Gerlach and U. K. Sengupta, Phys. Rev. D 22, 1300 [25] S. Jhingan and T. Tanaka, Phys. Rev. D 67, 104018 (2003).
(1980). [26] K. Martel, Ph.D. thesis, University of Guelph, 2003.
[12] C. Gundlach and J. M. Martin-Garcia, Phys. Rev. D 61, [27] C. W. Misner, K. S. Thorne, and J. A. Wheeler, Gravitation
084024 (2000). (Freeman, San Francisco, 1973).
[13] J. M. Martin-Garcia and C. Gundlach, Phys. Rev. D 64, [28] J. Goldberg, A. MacFarlane, E. Newman, F. Rohrlich, and
024012 (2001). E. Sudarshan, J. Math. Phys. (N.Y.) 8, 2155 (1967).
[14] O. Sarbach and M. Tiglio, Phys. Rev. D 64, 084016 [29] K. S. Thorne, Rev. Mod. Phys. 52, 299 (1980).
(2001). [30] C. O. Lousto and R. H. Price, Phys. Rev. D 55, 2124
[15] C. Clarkson and R. Barrett, Classical Quantum Gravity 20, (1997).
3855 (2003). [31] E. Poisson, Classical Quantum Gravity 21, R153 (2004).
104003-13