Multipole Radiation

interchangably as the unit radial vector.

Notice that throughout, we use r and n
Consider a localized, oscillating source, located in otherwise empty space. We know that the solution for
the vector potential (e.g. using the Green function for the outer boundary at infinity) is



J (x0 , t0 )
1
0
0
0
d3 x0 dt0

t
+
|x

x
|
A (x, t) =
4
|x x0 |
c
Let the source fields be confined in a region d  where is the wavelength of the radiation, and let the
time dependence be harmonic, with frequency ,
A (x, t) = A (x) eit
J (x, t) = J (x) eit
= (x) eit

(x, t)
Then
A (x) e

it

=
=

so that with k =

c,



0
) eit
1
0
0
d x
dt
t t + |x x |
|x x0 |
c
0
1

J (x0 ) ei(t c |xx |)

0
d3 x0
4
|x x0 |
0
4

0 J (x

3 0

we have

ik|xx0 |
0
0
3 0 J (x ) e
A (x) =
d x
4
|x x0 |
The electric and magnetic fields follow immediately. We know that B = A, so
H=

while the Maxwell equation, H

D
t

1
A
0

= 0, shows that
i (0 E) = H

Dividing by
i0

ikc0
ik0

0 0
r
0
ik
0

=
=
=

and defining the impedence of free space, Z =

0
0 ,

gives the electric field in the form

iZ
H
k
Now we consider the radiation. The problem divides into three approximate regions, depending on the
length scales d and . We assume d  . If r is the distance of the observation from the source, we will
consider
E=

d  r  (static zone)
d  r (induction zone)
d  r

(radiation zone)

or simply the near, intermediate, and far zones.

1

Near zone
For the near zone,
kr =

2r
1

implies
eikr 1
and the potential becomes
A (x)

=
=

0
J (x0 )
lim
d3 x0
4 kr0
|x x0 |

X
0
4 Ylm (, )

d3 x0 J (x0 ) r0l Ylm

lim
(0 , 0 )
4 kr0
2l + 1 rl+1
l,m

and the only time dependence is the sinusoidal oscillation, eit . The spatial integration depends on the
details of the source.

Far zone
The exponential becomes important in the radiation zone, kr  1. Setting x = r
n = rr, we have
p
r2 + x02 2r
n x0
|x x0 | =
r
x02
2
x0
= r 1+ 2 n
r
r
and since x02  r2 , we may drop the quadratic term and approximate the square root,
r
2
0
x0
|x x | r 1 n
r


1
0
x
r 1 n
r
x0
= rn
The potential is then
0
A (x) =
4
For the lowest order approximation, we neglect

0 eikr
A (x) =
4 r

0 eikr
=
4 r

0 eikr
=
4 r

) eik(rnx )
x0
rn

3 0 J (x

d x

the x0 term in the denominator,

d3 x0

J (x0 ) eiknx
x0
1 1r n

3 0

d x J (x ) e

ik
nx0



1
0
x
1+ n
r

d3 x0 J (x0 ) eiknx

x0  k n
x0 .
where the last step follows because d  implies 1r n
0
For higher orders in the kx / kd  1, note that powers of kx0 decrease rapidly in magnitude. We can
carry a power series in kx0 to higher order N in kx0 kd as long as we can still neglect dr ,
d
N
 (kd)
r
2

The expansion is then

A (x)

=
=

0
0 eikr
d3 x0 J (x0 ) eiknx
4 r
n
0 eikr X (ik)
n
d3 x0 J (x0 ) (
n x0 )
4 r n=0 n!

and because each term is smaller than the last by a factor on the order of kd, it is only the lowest nonvanishing
moment of the current distribution

n
d3 x0 J (x0 ) (
n x0 )
that dominates the radiation field.
The radiation zone solution is characteristic of radiation. Returning to lowest order

0
0 eikr
d3 x0 J (x0 ) eiknx
A (x) =
4 r
we note that the integral contributes only angular dependence of the field,

0
d3 x0 J (x0 ) eiknx = f (, )
so the waveform is

0 eikr
f (, )
4 r
The potential is therefore a harmonic, radially-expanding waveform, with amplitude decreasing as 1r ,
A (x) =

A (x, t)

ei(krt)
r

The magnetic field is given by

B (x)

=
=
=

A (x)



0
0 eikr

d3 x0 J (x0 ) eiknx
4 r




eikr
0
eikr
3 0
0
ik
nx0
3 0
0
ik
nx0

d x J (x ) e
d x J (x ) e
4
r
r

The first term is in the direction

d3 x0 J (x0 ) eiknx

and therefore transverse. Since the gradient includes a term

eikr
eikr
ikn
r
r

the magnetic field will fall off as 1r .

For the second term

0
0
d3 x0 J (x0 ) eiknx = d3 x0 J (x0 ) eik(xx )/r
the gradient gives
e

ik(xx0 )/r

=e

ik(xx0 )/r


x0
ikx0
ik n

+
n
r
r

The second term is also transverse, so we need only consider

0
d3 x0 (x0 J) eiknx
If we think of the exponential as giving a modified source,
= Jeiknx0
J
then this integral is just twice the magnetic dipole moment of that source. We know that the resulting
magnetic field may be written as


0 3n (n m) m
B=
4
r3
Although this does have a radial component, the magnitude falls off as the cube of the distance, and is
therefore negligible. The magnetic field is therefore transverse to the radial direction of propagation.
The electric field is also dominated by the
eikr ikneikr
term, falling off as 1r , and therefore being transverse to the radial direction.

Intermediate zone
In the intermediate zone, r , an exact expansion of the Green function is required. This is found by
expanding
0
eik|xx |
0
G (x, x ) =
4 |x x0 |
in spherical harmonics,
G (x, x0 ) =

glm (r, r0 ) Ylm (, ) Ylm

(0 , 0 )

l,m

and solving the rest of the Helmholtz equation for the radial function. The result is spherical Bessel functions,
(1)
(2)
jl (x) , nl (x), and the related spherical Henkel functions, hl , hl , which are essentially Bessel function times
1 . They are discussed in Jackson, Section 9.6. The vector potential then takes the form
r

A (x)

= ik0

(1)

hl (kr) Ylm (, )

d3 x0 J (x0 ) jl (kr0 ) Ylm

(0 , 0 )

l,m

The spherical Bessel function may be expanded in powers of kr to recover the previous approximations.

Explicit multipoles: n = 0 and n = 1

For higher multipoles (n > 0), we require the vector potential in order to get both magnetic and electric
fields. We can then find both fields using
H

E =

1
A
0
iZ
H
k

Since there is no magnetic monopole field, we may use the scalar potential to demonstrate the absence of
monopole radiation.

Monopole field
The lowest order far field is the electric monopole field. For this it is easiest to use the solution for the scalar
potential, in terms of charge density,



(x0 , t0 )
1
1
0
0
d3 x0 dt0
(x, t) =

t
|x

t
+
|
40
|x x0 |
c



1
1
d3 x0 dt0 (x0 , t0 ) t0 t + |x x0 |
=
40 r
c
However, all charge is confined to a central region, and total charge is conserved. This means that the spatial
integral is independent of time,

d3 x0 (x0 , t0 )

qtot =
so
(x, t)

qtot
40 r

The electric field is therefore a static Coulomb field; there is no radiation.

Dipole field
If the first nonvanishing term in the multipole expansion,

0
0 eikr
A (x) =
d3 x0 J (x0 ) eiknx
4 r
n
0 eikr X (ik)
n
d3 x0 J (x0 ) (
n x0 )
=
4 r n=0 n!
is the lowest (n = 0), then we have
A (x)

0 eikr
4 r

d3 x0 J (x0 )

From the continuity equation,

+J
t
i (x) + J

and a cute trick. Since the current vanishes at infinity, we may write a vanishing total divergence of xj J (x0 ):

x0j J (x0 )
d3 x0 x0j J (x0 ) =
d2 x0 n
=
Then

0
d3 x0 x0j J (x0 )
X


0
=
d3 x0 i x0j Ji (x0 )

 0

0
d3 x0 i x0j Ji (x0 ) + x0j i Ji (x0 )

X
i

=
and we have



0
d3 x0 ij Ji (x0 ) + x0j i Ji (x0 )



0
d3 x0 Jj (x0 ) + x0j J

d3 x0 Jj (x0 )

d3 x0 x0j J

= i d3 x0 x0j (x0 )
=

This integral is the electric dipole moment,

d3 x0 x0 (x0 )

p=
and the vector potential is
A (x)

i0 eikr
p
4 r

The magnetic field is the curl of this,

H (x)

=
=
=
=
=
=
=

1
A
0
 ikr 
i
e

p
4
r


i
eikr

4
r
 ikr 
i e
r p

4 r
r
 ikr

eikr
i ike
2 r p

4
r
r

 ikr
1
e
k
r p
1
4
ikr
r

 ikr
k2 c
1
e
r p
1
4
ikr
r

and is therefore transverse to the radial direction. For the electric field,
E =
=

iZ
H
k

 ikr

iZkc
1
e
r p

1
4
ikr
r

Using
(a b) = a ( b) b ( a) + (b ) a (a ) b
this becomes
E =



 ikr 


 ikr 
1
e
1
e
iZkc
r p 1
r
(p ) 1
4
ikr
r
ikr
r
6

We easily compute the first term in the brackets using the identity
(a ) (nf (r)) =
Thus, the first term is

 ikr 
1
e
r
(p ) 1
=
ikr
r
=
=

f (r)
f
[a (a n) n] + (a n) n
r
r

f (r)
f
[p (p r) r] + (p r) r
r
r

 ikr


 ikr 


1
1
1
e
1 eikr
1
e
ikeikr
+
1

1
[p (p r) r] + (p r) r

r
ikr
r
ikr2 r
ikr
r2
ikr
r




ikr
2
e
1
1
2
[p (p r) r] + (p r) r ik +

r
r
ikr2
r
ikr2

For the second term, we use

(rf (r)) =

2
f
f+
r
r

so that
 ikr 

e
1
r
=
1
ikr
r
=
=
=

 ikr 
1
e
1
ikr
r


 ikr
 ikr 

ikr
2
e
1 e
1
e
1
ikeikr
1
+

+
1

1
r
ikr
r
ikr2 r
ikr
r2
ikr
r


ikr
2
2
1
1
1
1
e

+
+
+ ik
2
2
2
r
r
ikr
ikr
r
ikr
r
2
r

1
1
ikr

=
=
=
=

eikr

+
r
r



ikeikr
r

Combining these results, and using Zc =

E =

0
0

1
0 0

1
0 ,






 ikr 
iZkc eikr
1
2
1
2
ike

+
[p

(p

r
)
r
]
+
(p

r
)
r
ik

p
4
r
r
ikr2
r
ikr2
r





ikr
iZkc e
1
2
1
2

[p (p r) r]
(p r) r ik (p (p r) r)
4 r
r
ikr2
r
ikr2






iZkc eikr
1
1
2
1
ik +
1
(p (p r) r)
1
(p r) r
4 r
r
ikr
r
ikr






1
1
2
1
iZkc eikr
ik (p (p r) r) +
1
(p (p r) r)
1
(p r) r
4 r
r
ikr
r
ikr




1 eikr 2
ik
1
k (p (p r) r) +
1
(p 3 (p r) r)
40 r
r
ikr

Finally, noting that

r (p r) = p (p r) r
we see that this is equivalent to the expression in Jackson. Notice that the first term is transverse, but not
the second.
The electric and magnetic fields for an oscillating dipole field are therefore,

 ikrit
k2 c
1
e
r p
H (x, t) =
1
4
ikr
r




1 eikrit 2
ik
1
E (x, t) =
k r (p r) +
1
(p 3 (p r) r)
40
r
r
ikr
7

In the radiation zone, kr  1, these simplify to

k 2 c eikrit
r p
4
r
ikrit
2
k e
r (p r)
E (x, t) =
40
r
= Z0 H r

H (x, t)

while in the near zone, kr  1,

H (x, t)

=
=

E (x, t)

ikc 1 ikrit
r p
e
4 r2
ikrc 1 ikrit
r p
e
4 r2
1
eikrit (3 (p r) r p)
40 r3

Notice that in the near zone, the electric field is just eit times a static dipole field, while
H

kr 1 p
Z 40 r3
1 p
|
|3 (
p r) r p
40 r3

so that H  E.

An aside: the trick, in general

Does the trick for expressing the moments of the current work for higher multipoles? Not quite! We show
here that two distinct distributions are required: the charge density and the magnetic moment density.
First, we know how to find the zeroth moment of the current in terms of the first moment of the charge
density,

d3 x0 Jj (x0 ) = i d3 x0 x0j (x0 )

Now, suppose we know the first n 1 moments of a current distribution in terms of moments of the charge
density, and want to find the nth ,

jk1 ...kn d3 x xk1 . . . xkn J (x)

Then consider the (vanishing) integral of a total divergence,

0 =
d3 x xk1 xk2 . . . xkn+1 J (x)
X


=
d3 xi xk1 xk2 . . . xkn+1 Ji
i

d3 x


ik1 xk2 . . . xkn+1 + xk1 ik2 . . . xkn+1 + . . . + xk1 xk2 . . . ikn+1 Ji + xk1 xk2 . . . xkn+1 i Ji

d3 x


Jk1 xk2 . . . xkn+1 + xk1 Jk2 . . . xkn+1 + . . . + xk1 xk2 . . . Jkn+1 + ixk1 xk2 . . . xkn+1

d3 x


Jk1 xk2 . . . xkn+1 + xk1 Jk2 . . . xkn+1 + . . . + xk1 xk2 . . . Jkn+1 + ixk1 xk2 . . . xkn+1

Each of the first n + 1 integrals,

d3 x Jk1 xk2 . . . xkn+1

is a component of jk1 ...kn , since there are n factors of the coordinates, xk2 . . . xkn+1 . The problem is that
st
we have expressed the (n + 1) moment of the charge density in terms of the symmetrized moments of the
current. We cannot solve for jk1 ...kn unless we also know the antisymmetric parts. Since the products of the
coordinates are necessarily symmetric (i.e., xk2 . . . xkn+1 is the same regardless of the order of the ki indices),
the only antisymmetric piece is

d3 x (Ji xk Jk xi ) xk2 . . . xkn

This does not vanish, but it can be expressed in terms of the magnetic moment density. Recall
M

Mi

1
xJ
2
1X
ijk xj Jk
2
j,k

Mi imn

1X
imn ijk xj Jk
2

1X
(mj nk mk nj ) xj Jk
2

i,j,k

j,k

1
=
(xm Jn xn Jm )
2
X
= xn Jm + 2
Mi imn

xm Jn

Consider the second term in our expression for the moments,

d3 x xk1 Jk2 . . . xkn+1
We can now turn it into the same form as the first term,

d3 x xk1 Jk2 . . . xkn+1

!
d3 x

xk2 Jk1 + 2

Mi ik1 k2

d3 xJk1 xk2 . . . xkn+1 + 2

!
. . . xkn+1

d3 xMi ik1 k2 xk3 . . . xkn+1

The first term on the right is now the same as the first term in our expression. Repeating this for each of
the n + 1 terms involving Jk gives

X
X
3
3
0 =
d x (n + 1) Jk1 xk2 . . . xkn+1 + 2
d xMi ik1 k2 xk3 . . . xkn+1 + . . . + 2
d3 xMi ik1 kn+1 xk2 xk3 . . . xkn+1 + i
i

and therefore,
jk2 ...kn+1

jk2 ...kn+1



i
2 X
d3 x Mi ik1 k2 xk3 . . . xkn+1 + . . . + Mi ik1 kn+1 xk2 xk3 . . . xkn+1
pk ...k

n+1 i
n + 1 1 n+1
2n! X
i

ik1 (k2 mik3 ...kn+1 )

pk ...k
n+1 i
n + 1 1 n+1
9

where

d3 xMxk1 . . . xkn

mk1 ...kn =

are the higher magnetic moments. The parentheses notation means symmetrization. For example,
T(ijk)

1
(Tijk + Tjki + Tkij + Tjik + Tkji + Tikj )
3!

This shows that there are two types of moments that we will encounter: electric multipole moments built
from the charge density, and magnetic multipole moments built from the magnetic moment density. We see
this explicitly in the calculation of the electric quadrupole term of our general expansion.

Electric quadrupole and magnetic dipole radiation

Return to the exact expression for the vector potential, and expand to the next order:
A (x)

0 eikr
4 r

0 eikr
4 r

d x

) eiknx
x0
1 1r n

3 0 J (x



1
x0 ) 1 + n
x0
d3 x0 J (x0 ) (1 ik n
r


ikr
ikr
0 e
0 e
1
x0 + n
x0
d3 x0 J (x0 ) +
d3 x0 J (x0 ) ik n
4 r
4 r
r

We know that the first term on the right leads to electric dipole radiation. Suppose this term vanishes, so
that the radiation is described by the next order terms:


0 eikr 1
A (x) =
ik
d3 x0 J (x0 ) (
n x0 )
4 r
r
Now we need the next higher moment of the current,

d3 x0 J (x0 ) (
n x0 )
We have seen that we can express the nth moment of the charge distribution in terms of symmetrized
st
(n 1) moments of the current. In order to get the symmetrized moments of J we may use a vector
identity. Starting with the magnetic moment density
M=

1
(x J)
2

we take a second curl

r M

=
=

1
r (x0 J)
2
1 0
(x (r J) J (r x0 ))
2

so that
J (r x0 ) = x0 (r J) 2r M
In components, this is
X
k

rk (Ji x0k ) =

rk x0i Jk 2

X
j,k

10

ijk rj Mk

Therefore, the ith component of the current moment is




d x J (x ) (
nx)
3 0

!
3 0

d x

=
=

rk Ji x0k

!
1X
1X
0
0
d x
rk Ji xk +
rk Ji xk
2
2
k
k

X
X
X
1
1
rk x0i Jk 2
ijk rj Mk
d3 x0
rk Ji x0k +
2
2
k
j,k
k

X 1
X
d3 x0
rk (Ji x0k + x0i Jk )
ijk rj Mk
2
k
j,k

X
i X

rk d3 x0 (x0i x0k ) + d3 x0
ijk rk Mj
2
k
j,k

X
X
i
d3 x0 (x0i x0k )
rk d3 x0
ijk Mj
2
j
3 0

Notice that

fi

X
k

d3 x0 r02 ik

rk

d3 x0 r02

ri

( r)

has vanishing curl,

d3 x0 r02

This means that this term may be added to the vector potential without affecting the fields (Jackson never
mentions this). This allows us to define the quadrupole moment of the charge distribution as the traceless
matrix

Qik =
d3 x0 3x0i x0k ik r02 (x0 )
X
Qkk = 0
k

without changing the fields. Then, with the magnetic dipole moment equal to

m = d3 xM
and setting
[Q (r)]i

rk Qik

the vector potential becomes

A (x)

0 eikr
4 r

1
ik
r

11

d3 x0 J (x0 ) (r x0 )

0 ikeikr
4 r

0 ikeikr
4 r



i
r m + Q (r)
6



1
i
1
Q (r) + r m
ikr
6
1

1
ikr

For comparison, here is the divergence trick applied to the present case. Consider

0 =
d3 x (xj xk J (x))
X
d3 xi (xj xk Ji (x))
=
i

d3 x [(ij xk + ik xj ) Ji + xj xk i Ji ]

i
d3 x [Jj xk + Jk xj + ixj xk ]

d3 x [2Jj xk + (Jk xj Jj xk ) + ixj xk ]

3
3
2 d x Jj xk + d x (Jk xj Jj xk ) + i d3 x xj xk

=
=
so that finally,

i
1
3
d x (Jk xj Jj xk )
d3 x xj xk
d x J j xk =
2
2
Now using the definition of the magnetic moment density,

M=

1
(x J)
2

and noting that

[r M]i

1X
ijk rj (x0 J)k
2

1 X
ijk kmn rj x0m Jn
2

1X
(im jn in jm ) rj x0m Jn
2 jmn

1X
rj x0i Jj x0j Ji
2 j

jk

jkmn

Returning to the expression for the vector potential



0 eikr 1
A (x) =
ik
d3 x0 J (x0 ) (r x0 )
4 r
r

X
0 eikr 1
Ai =
ik
rk d3 x0 Ji x0k
4 r
r
k

X 


ikr
1
0 e
1
i
ik
rk
d3 x0 (Jk x0i Ji x0k )
d3 x0 x0i x0k
=
4 r
r
2
2
k
!



0 eikr 1
i X
3 0
3 0
0 0
=
ik
d x ([r M]i )
rk d x xi xk
4 r
r
2
k

12

A (x)

0 ikeikr
4 r


1

1
ikr




i
Q (r) + r m
6

so the result is just as above.

The fields are now found in the usual way. The magnetic dipole potential is
A (x)

0 ikeikr
r m
4 r

Notice that the magnetic field for the electric dipole had exactly this form,
i0
0 ikeikrit
r p
He (x, t) =
kc
4
r
once we replace p m. Since the electric field in that case was given by Ee (x, t) =
write the curl of A immediately. We have
Hm (x, t)

1
Am (x)
0

=
=
=

iZ0
k

He , we can

!


i0

He (x, t)
kc
pm

i k
Ee (x, t)|pm
kc iZ0




ik
1
1 1 eikrit 2
k r (m r) +
1
(m 3 (m r) r)
cZ 40
r
r
ikr




ikrit
1 e
ik
1
k 2 r (m r) +
1
(m 3 (m r) r)
4
r
r
ikr

We can resort to this sort of magic again, because we know that this form of Ee (x, t) was achieved by taking
the curl of He , and the Maxwell equations for harmonic sources tell us that

iBm =
!


i0

=
He (x, t)
kc
pm
0 He (x, t)|pm

Em
Em
Em

Notice that dropping the curl on both sides is not quite allowed, since the right and left sides could differ by
a gradient, but the answer here is correct. The electric field for magnetic dipole radiation is correctly given
by
Em

=
=
=

0 He (x, t)|pm
 2 
 ikrit


k c
1
e
r p 
0
1
4
ikr
r
pm

 ikrit
Z0
1
e
r m
k2 1
4
ikr
r

This gives the magnetic dipole fields as





1 eikrit 2
ik
1
H =
k r (m r) +
1
(m 3 (m r) r)
4
r
r
ikr

 ikrit
Z0
1
e
r m
E = k2 1
4
ikr
r
13

in complete analogy to the electric dipole field, but with magnetic and electric parts interchanged. In the
radiation zone, these become
1 2 eikrit
k
[r (m r)]
4
r
Z0 eikrit
r m
E = k2
4
r

For the quadrupole fields, we begin with the quadrupole piece of the vector potential


i0 ikeikr
1
A (x) =
1
Q (r)
24
r
ikr
where writing
[Q (r)]i

rk Qik

allows us to write the potential in vector form. The magnetic field is then
H (x, t) =

1
A (x)
0

Keeping only terms of order 1r , this gives

H (x, t)

=
=
=



1
i0 ikeikr
Q (r)

0
24
r


1 ik 2 0 eikr
r Q (r)
0
24
r
ick 3 eikr
r Q (r)
24 r

since the only derivative term that does not increase the power of
field is then

1
r

is

 i0 ik
24 r


eikr Q (r) . The electric

E = Z0 H r


ick 3 eikr
r Q (r) r
= Z0
24 r
=

ik 3 eikr
[r Q (r)] r
240 r

Radiated power
The energy per unit area carried by an electromagnetic wave is given by the Poynting vector,
S=EH
For a plane wave, we have
E = E0 cos (k x t)
1 1
H =
k E0 cos (k x t)

14

So
S

= EH


1 1
= E0 cos (k x t)
k E0 cos (k x t)

k
r
1
=
E0 (k E0 ) cos2 (k x t)
k
r
1 2
=
E k cos2 (k x t)
k 0

with real part

r
S =

1 2
E k cos2 (k x t)
k 0

with time average


 r
1 2
E0 k
2

S=
For a complex representation of the wave,

E = E0 ei(kxt)
1 1
H =
k E0 ei(kxt)

k
we may write the same quantity as
1
Re (E H )
2

=
=



1
1
i(kxt)
i(kxt)
k E0 e
Re E0 e

r
1
2
|E0 | k
2

so the time-averaged energy flow per unit area per unit time is
1
Re (E H )
2

S=

Now consider the average power carried off by electric dipole, electric quadrupole, and magnetic dipole
radiation.
Electric dipole
The radiation zone fields were found above to be
H (x, t)

E (x, t)

k 2 c eikrit
r p
4
r
k 2 eikrit
r (p r)
40
r

so that
dP
dA

= r S
=

1
Re (r (E H ))
2
15

 2


1
k 1
k2 c 1
r [r (p r)]
Re
[r p]
2
40 r
4 r
4
1
ck
2
|p| sin2
=
32 2 0 r2

c2 k 4 0 0 1
2
=
|p| sin2
32 2 0 r2
c2 Z 0 1 4 2 2
k |p| sin
=
32 2 r2
This is the power per unit area. Since the area element at large distances is dA = r2 d, where is the solid
angle, we may write the differential power radiated per unit solid angle using
=

dP
1 dP
= 2
dA
r d
so that
dP
d

c2 Z 0 4 2 2
k |p| sin
32 2

Magnetic dipole
The radiation zone fields for magnetic dipole radiation are
k 2 eikrit
r (m r)
4
r
Z0 k 2 eikrit
r m
E (x, t) =
4
r
so the result is the same as for the electric dipole with the substitution p m/c,
H (x, t)

dP
d

Z0 4
2
k |m| sin2
32 2

Electric quadrupole moment

For electric quadrupole radiation the fields are given by
ick 3 eikr
r Q (r)
24 r
ik 3 eikr
E (x, t) =
[r Q (r)] r
240 r
giving an average power per unit solid angle of
H (x, t)

r2
|Re (E H )|
2 

  3 ikr


r2 
ik 3 eikr
ick e
r Q (r) 
=
Re
[r Q (r)] r

2
240 r
24 r
3
3
1 k
ck
=
|([r Q (r)] r) (r Q (r))|
2 240 24
ck 6
|([r Q (r)] r) (r Q (r))|
=
1152 2 0
Z 0 c2 6
2
=
k |[r Q (r)] r|
1152 2
Notice that the power radiated by the quadrupole moment depends on k 6 , whereas the power radiated
by the dipole moments both go as k 4 . This pattern continues for higher moments.
dP
d

16