7

Geomagnetism

The existence of a magnetic force was known for centuries before William Gilbert
pointed out in 1600 that the Earth itself behaved like a huge magnet. Gradually
maps were made of the geomagnetic elements. Systematic investigation of
magnetic behavior was undertaken in the late eighteenth and early nineteenth
centuries. The French scientist Charles Augustin de Coulomb showed experi-
mentally that forces of attraction and repulsion exist between the ends of long thin
magnetized rods, and that they obey rules similar to those determining the
interaction of electrical charges. A freely suspended magnet was observed to
align approximately north–south; the north-seeking end became known as its
north pole, the opposite end as its south pole. The origin of magnetic force was
attributed to magnetic charges, which, through association, became known as
magnetic poles. Subsequently, it was shown that individual magnetic poles, or
monopoles, do not exist. All magnetic fields originate in electric currents. This is
true even at atomic dimensions; circulating (and spinning) electrical charges
impart magnetic properties to atoms. However, the concept of multiple pole
combinations (e.g., the dipole, quadrupole, and octupole) proved to be very useful
for describing the geometries of magnetic fields.

7.1 The dipole magnetic field and potential

The most important field geometry is that of a magnetic dipole. This was
originally imagined to consist of two equal and opposite magnetic poles that
lie infinitesimally close to each other (Appendix A2). At distances several times
greater than the size of the source the field of a very short bar magnet is very
nearly a dipole field, as is the magnetic field produced by an electric current in a
small plane loop. In an external magnetic field B a magnetic dipole experiences
a torque τ that aligns it with the field (Appendix A4). The torque is governed by
the relationship

198
 7.1 The dipole magnetic field and potential 199

τ=m B (7:1)

In this equation m is the magnetic moment of the dipole, a measure of its

strength. For a current-carrying loop m is equal to the product of the current I
in the loop and its area A, and its direction en is that of the normal to the plane of
the loop (Appendix A4):
m ¼ ðIAÞen (7:2)
The dimensions of magnetic moment are by this definition A m2; the dimen-
sions of torque are N m; thus the SI unit of the magnetic field B, the tesla, has the
dimensions N A−1 m−1.
The potential W of a dipole magnetic moment m at distance r from its center
and at an azimuthal angle θ between the dipole axis and the radial direction is
(Appendix A2)
μ m cos θ
W¼ 0 (7:3)
4π r2
The constant μ0 is the magnetic field constant. It is defined in SI units to be
exactly 4π × 10−7 N A−2 (alternatively designated henry m−1). The dipole
potential is the most important component of the geomagnetic field, represent-
ing more than 93% of its energy density.
The dipole magnetic field B is the gradient of the dipole potential: B = −∇W.
In spherical coordinates the field has a radial component Br and an azimuthal
component Bθ. These are
!
∂ μ0 m cos θ μ 2m cos θ
Br ¼ $ 2
¼ 0 (7:4)
∂r 4π r 4π r3
!
1 ∂ μ0 m cos θ μ0 m sin θ
Bθ ¼ $ ¼ (7:5)
r ∂θ 4π r2 4π r3
For a dipole at the center of the spherical Earth, the azimuthal component of
the field, Bθ, is horizontal. Moreover, if the dipole is aligned with the Earth’s
axis, the angle θ is the complement of the magnetic latitude β. The direction of
the field makes an angle I with the horizontal called the inclination of the field
(see Fig. 7.1(b) and Appendix A, Fig. A1). The inclination, magnetic co-
latitude, and magnetic latitude are related by
Br
tan I ¼ ¼ 2 cot θ ¼ 2 tan β (7:6)
Bθ
This equation forms the basis of paleomagnetic determination of ancient pale-
olatitudes from the inclinations of remanent magnetizations measured in ori-
ented rock samples.
200 Geomagnetism

7.2 Potential of the geomagnetic field

The empirical laws that govern electricity and magnetism are summarized in
Maxwell’s equations (Appendix B). Analysis of the present geomagnetic field
requires Gauss’s law and Ampère’s law.
Gauss’s law established that the net magnetic flux through any closed surface
is zero. This is equivalent to stating that there are no magnetic monopoles:
dipole sources such as current circuits, even at atomic scale, produce zero net
flux through a surrounding surface. The corresponding equation is
r·B ¼ 0 (7:7)
Ampère’s law showed that an electric current produces a magnetic field in the
surrounding space, and it relates the strength of the magnetic field B to the
electric field E that causes the current:
∂E
r " B ¼ μ0 σE þ μ0 ε0 (7:8)
∂t
The first term on the right is the electric current associated with the flow of free
charges in a conductor and relies on Ohm’s law; the second term is the electric
displacement current that results from time-dependent motions of charges
bound to a parent atom. The parameter μ0 is the magnetic field constant, or
permeability of free space, and ε0 is the electric field constant, or permittivity of
free space; σ is the electrical conductivity of the medium.
In a region that is free of sources of the magnetic field (such as the space just
above the Earth’s surface in which the field is measured), we can assume that
there are no electric or displacement currents, thus
r"B¼0 (7:9)
Consequently, the magnetic field B can be written as the gradient of a scalar
potential, W:
B ¼ $rW (7:10)
On substituting for B in (7.7) the potential W of the Earth’s magnetic field is seen
to satisfy Laplace’s equation:
r2 W ¼ 0 (7:11)

7.2.1 The fields of internal and external origin

The geomagnetic potential at Earth’s surface arises from two sources. The most
important part of the field originates in the Earth’s interior, and the rest
 7.2 Potential of the geomagnetic field 201

originates outside the Earth, e.g., from current systems in the ionosphere. Let Wi
be the potential of the field of internal origin and We be the potential of the field
of external origin. The total geomagnetic potential W at Earth’s surface is

W ¼ We þ Wi (7:12)
The geomagnetic potential has to be conformable with Earth’s approximately
spherical geometry, so the solution of (7.11) requires spherical polar coordi-
nates. The general solution of Laplace’s equation is therefore as described in
Section 1.16. The variation of potential on a spherical surface is described by
spherical harmonic functions of the co-latitude θ and longitude !. The variation
of potential with radial distance r consists of two parts. In a region where r can
be zero, the potential is proportional to rn. At Earth’s surface this condition
applies to the field due to sources outside the Earth, so We must vary as rn. In a
region where r can be very large or infinite, the potential is proportional to
1/rn+1. Outside the Earth and on its surface, this applies to the potential of the
field of internal origin, so Wi must vary as 1/rn+1. These considerations lead to
the following definition for the potential We of the field of external origin:
1 X
n ! "n #
X r
Gm m
$ m
We ¼ R n cosð m! Þ þ H n sin ð m!Þ Pn ðcos θÞ; r 5 R
n¼1 m¼0
R
(7:13)
Similarly, the potential Wi of the field of internal origin is
1 X
n % &nþ1 #
X R
gm m
$ m
Wi ¼ R n cos ð m! Þ þ h n sin ðm! Þ Pn ðcos θÞ; r 4 R
n¼1 m¼0
r
(7:14)
Terms with n = 0 are absent from these expressions because magnetic monop-
oles do not exist. At the Earth’s surface the expressions simplify to
1 X
X n #
Gm m
$ m
We ¼ R n cos ð m! Þ þ H n sinð m!Þ Pn ðcos θÞ (7:15)
n¼1 m¼0

1 X
X n #
gm m
$ m
Wi ¼ R n cosðm!Þ þ hn sinðm!Þ Pn ðcos θÞ (7:16)
n¼1 m¼0

In a convention adopted in 1939 by the scientific body that preceded the

modern International Association of Geomagnetism and Aeronomy (IAGA), it
was agreed to base the spherical harmonic functions in the magnetic potential on
the partially normalized Schmidt polynomials (Section 1.15.2). The coefficients
202 Geomagnetism

(gm m m m
n , hn ) and (Gn , Hn ) are called the Gauss (or Gauss–Schmidt) coefficients of
the fields of internal and external origin, respectively. They have the dimensions
of magnetic field and their magnitudes diagnose the relative importance of the
external and internal sources of the field.

7.2.2 Determination of the Gauss coefficients

It is not possible to measure the geomagnetic potential directly, so the Gauss
coefficients are calculated from measurements of the northward (X ), eastward
(Y ), and vertically downward (Z ) components of the magnetic field at or above
the Earth’s surface (Fig. 7.1(a)). These components are related to other geo-
magnetic elements, such as the horizontal field (H ), total field (T ), angle of
inclination (I ), and angle of declination (D), as illustrated in Fig. 7.1(b). The
field components in spherical polar coordinates are

1 ∂W
X ¼ !Bθ ¼ (7:17)
r ∂θ r¼R

!1 ∂W
Y ¼ B! ¼ (7:18)
r sin θ ∂! r¼R

∂W
Z ¼ !Br ¼ (7:19)
∂r r¼R

The differentiations, after evaluating on the Earth’s surface at r = R, result in

the following set of equations involving the unknown Gauss coefficients:

geographic magnetic
North North
N H
(a) (b)
Br D
N
Bφ I E
E T
θ V Bθ
φ
r V

Fig. 7.1. (a) Relationship between the north (X), east (Y), and vertical (Z)
components of the geomagnetic field and the spherical polar components Br, Bθ,
and B . (b) The field may be described by the X, Y, and Z components, or by its
intensity (T), declination (D), and inclination (I). A magnetic compass aligns with
the horizontal component H, which is directed towards magnetic north.
 7.2 Potential of the geomagnetic field 203

1 X
X n #! $ ∂ m
gm m
" ! m m
"
X¼ n þ G n cos ð m Þ þ hn þ H n sin ð m Þ P ðcos θÞ
n¼1 m¼0
∂θ n
(7:20)
1 X
X n #! $ m m
gm m
" ! m m
"
Y¼ n þ Gn sin ð m Þ % h n þ Hn cos ð m Þ P ðcos θÞ
n¼1 m¼0
sin θ n
(7:21)
1 P
n #!
ðn þ 1Þgm m
P " $
Z¼% n % nGn cosðm Þ
n¼1 m¼0 (7:22)
þ ðn þ 1Þhm m m
! "
n % nHn sinðm ÞPn ðcos θÞ

Note that the Gauss coefficients have the same dimensions as the magnetic field
B, namely tesla. The tesla is a large magnetic field, so the geomagnetic field
intensity and the Gauss coefficients are usually expressed in nanotesla (1 nT =
10−9 T). In the north and east components the Gauss coefficients occur as
# m
gn þ Gm and hm m
$ # $
n n þ Hn , and therefore the horizontal components alone
do not allow separation of the external and internal parts. However, the Gauss
coefficients occur in a different combination in the vertical field, and by virtue of
this the external and internal fields can be separated.
In theory, the summations are over an infinite number of terms, but in practice
they are truncated after a certain degree N. The coefficients h0n and H0n do not
exist, because sin(m ) = 0 for m = 0, and these terms make no contribution to the
potential. For n = 1 there are three coefficients for the internal field (g01 ; g11 ; h11 )
and three for the external field (G01 ; G11 ; H11 ). Similarly, there are five of each for
n = 2, and in general 2(2n + 1) for degree n. The total number of coefficients Sn
up to and including order N for each part of the field is

SN ¼ ½2ð1Þ þ 1' þ ½2ð2Þ þ 1' þ ½2ð3Þ þ 1' þ ( ( ( þ ½2ðNÞ þ 1'

¼ 2ð 1 þ 2 þ 3 þ ( ( ( þ N Þ þ N (7:23)

The sum of the first N natural numbers is N(N + 1)/2, so the number of
coefficients up to degree and order N of the internal field is N(N + 2). The same
number is obtained for the external field. Thus separation requires knowing the
field values at a minimum of 2N(N + 2) stations.
From 1835 to 1841 Carl Friedrich Gauss and Wilhelm Weber organized the
semi-continuous (every 5 minutes, 24 hr/day) acquisition of data from up to 50
magnetic observatories distributed worldwide, albeit unevenly. Gauss in 1839
carried out the first analysis of the geomagnetic field up to degree and order 4, and
established that it is dominantly of internal origin; the coefficients of the external
204 Geomagnetism

field are small compared with those of the internal field, and may to a first
approximation be neglected. The potential of the internal field is given by (7.14).
Magnetic field components have historically been measured and recorded at
geomagnetic observatories. A drawback of the data from observatories is their
uneven geographic distribution. A superior global coverage has been obtained
during the last decades with the addition of data from satellites. The coefficients
of the modern geomagnetic field have now been evaluated reliably up to degree
and order 13. The data are updated and published regularly as the coefficients of
the International Geomagnetic Reference Field (IGRF). The coefficients up to
degree and order 3, corresponding to the dipole, quadrupole, and octupole
components of the field at the Earth’s surface are listed in Table 7.1 for some
selected field models. The terms with n = 1 describe a dipole field; the higher-
order terms with n ≥ 2 are referred to collectively as the non-dipole field.

Table 7.1. Dipole (n = 1), quadrupole (n = 2), and octupole (n = 3) Gauss–

Schmidt coefficients from some historical field analyses. The coefficients DGRF
are for Definitive Geomagnetic Reference Fields that will not be modified
further. Details of the construction of the International Geomagnetic Reference
Field IGRF 2010 are given in Finlay et al. (2010).

Epoch and source

1835, 1885, 1922,

Gauss, Schmidt, Dyson and 1965, 1985, 2010,
in 1839 in 1895 Furner (1923) DGRF DGRF IGRF

g10 −32,350 −31,730 −30,920 −30,334 −29,873 −29,496.5

g11 −3,110 −2,360 −2,260 −2,119 −1,905 −1,585.9
h11 6,250 5,990 5,920 5,776 5,500 4,945.1
g20 510 −520 −890 −1,662 −2,072 −2,396.6
g21 2,920 2,830 2,990 2,997 3,044 3,026.0
h21 120 −720 −1,240 −2,016 −2,197 −2,707.7
g22 −20 680 1,440 1,594 1,687 1,668.6
h22 1,570 1,500 840 114 −306 −575.4
g30 – 940 1,140 1,297 1,296 1,339.7
g31 – −1,230 −1,650 −2,038 −2,208 −2,326.3
h31 – −300 −460 −404 −310 −160.5
g32 – 1,430 1,200 1,292 1,247 1,231.7
h32 – 30 120 240 284 251.7
g33 – 400 880 856 829 634.2
h33 – 680 230 −165 −297 −536.8
 7.3 The Earth’s dipole magnetic field 205

7.3 The Earth’s dipole magnetic field

The dominant component of the Earth’s surface magnetic field is the dipole
component. The axis of the dipole is inclined to the rotation axis, thus it can be
separated into an axial dipole and two orthogonal equatorial dipoles. As we will
see, shifting these dipoles from the center of the Earth generates higher-order
components in the geomagnetic potential.

7.3.1 The geocentric axial dipole

Each term in the geomagnetic potential (7.14) represents the potential of a partic-
ular pole configuration. The potential described by the largest coefficient, g01 , is

R3 g01 0 R3 g01 cos θ

W 01 ¼ P 1 ð cos θ Þ ¼ (7:24)
r2 r2
Comparison with (7.3) shows that this is the potential at distance r from the mid-
point of a magnetic dipole and at angle θ from the dipole axis. In Earth coor-
dinates this is the potential at co-latitude θ of a geocentric dipole aligned with the
rotation axis and pointing to the north pole with magnetic moment m given by

4πR3 0
m¼ g (7:25)
μ0 1
The magnetic field of an axial dipole is horizontal at the equator (see (7.4) and
(7.5)). Its value at Earth’s surface is

1 ∂ R3 g01 cos θ ""

!"
Bθ ¼ # ¼ g01 sin θ (7:26)
r ∂θ r2 "
r¼R

At the equator this is equal to g01 .

7.3.2 The geocentric inclined dipole

The coefficients of degree n = 1 and order m = 1 also have an inverse-square
dependence on distance, so g11 and h11 too must represent dipoles. The combined
potential of the dipole terms is

R 2# 0 0
!
g1 P1 ðcos θÞ þ ðg11 cos ! þ h11 sin !ÞP11 ðcos θÞ
$
W1 ¼ R (7:27)
r

R 2# 0
!
g1 cos θ þ g11 cos ! sin θ þ h11 sin ! sin θ
$
W1 ¼ R (7:28)
r
206 Geomagnetism

geographic
pole
magnetic
pole

(θ 0 , φ 0 ) Θ

site
θ0
P(θ , φ )
θ

O
φ=0 φ0
Greenwich φ
meridian equator
Fig. 7.2. Angular relationships pertaining to the computation of the potential of an
inclined geocentric magnetic dipole.

Consider now the direction cosines of a line OP inclined at angle θ to the

reference axis and at angle to the reference axis = 0, as in Fig. 7.2. The
direction cosines (α, β, γ) of OP are

α ¼ sin θ cos
β ¼ sin θ sin (7:29)
γ ¼ cos θ

Suppose the axis of a magnetic dipole to be inclined at angle θ0 to the z-axis and
at angle 0 to the reference axis = 0. The direction cosines (α0, β0, γ0) of the
dipole axis are

α0 ¼ sin θ0 cos 0
β0 ¼ sin θ0 sin 0 (7:30)
γ0 ¼ cos θ0

If Θ is the angle between OP and the dipole axis, and r the distance of P from the
dipole center, the magnetic potential at P is

μ0 m μ0 m
W1 ¼ cos Θ ¼ ðαα0 þ ββ0 þ γγ0 Þ (7:31)
4πr2 4πr2
The components of the dipole moment m along the reference axes (Fig. 7.3) are
 7.3 The Earth’s dipole magnetic field 207

(α 0 , β 0 , γ 0 )

mz
m
θ0

θx θy
my
mx y
φ0
x
Fig. 7.3. Relationship between the Cartesian components and direction cosines of
a magnetic dipole m, which is inclined at angle θ0 to the rotation axis and has an
azimuth !0 in the equatorial meridian.

mx ¼ m cos θx ¼ mα0
my ¼ m cos θy ¼ mβ0 (7:32)
mz ¼ m cos θ0 ¼ mγ0

The potential of the inclined dipole becomes

μ
W1 ¼ 0 2 αmx þ βmy þ γmz
!
(7:33)
4πr
Using the relationships in (7.29), the potential of the inclined dipole is
μ
W1 ¼ 0 2 mz cos θ þ mx cos ! sin θ þ my sin ! sin θ
!
(7:34)
4πr
On equating individual terms with the expression for the potential using Gauss
coefficients (7.28) it is evident that the coefficients g11 and h11 represent orthog-
onal dipoles in the equatorial plane. The equatorial dipole components are
4πR3 1
mx ¼ g (7:35)
μ0 1

4πR3 1
my ¼ h (7:36)
μ0 1
The axial component of the dipole is

4πR3 0
mz ¼ g (7:37)
μ0 1
208 Geomagnetism

The points where the dipole axis intersects the Earth’s surface are called the
geomagnetic poles (Fig. 7.2). At these points the dipole magnetic field is normal
to the surface. The geomagnetic poles are antipodal to each other, because they
lie at the opposite ends of the inclined axis. The co-latitude θ0 of the pole is
equal to the tilt of the inclined axis. From (7.30) and (7.32)
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 4π qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
" 1 #2 " 1 #2ffi
2 2 3
m sin θ 0 ¼ mx þ my ¼ R g1 þ h1 (7:38)
μ0
Together with the axial component, this defines the tilt θ0 of the dipole axis,
which is also the co-latitude of its pole:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
" 1 #2 " 1 #2ffi
m sin θ0 g1 þ h1
tan θ0 ¼ ¼ (7:39)
m cos θ0 g01
The components of the dipole moment in the equatorial plane, mx and my, define
the longitude !0 of the pole. From (7.35) and (7.36)

β0 my h11
tan !0 ¼ ¼ ¼ (7:40)
α0 mx g11

The dipole magnetic moment m is obtained by squaring and summing mx, my,
and mz, giving
4π 3 " 0 #2 " 1 #2 " 1 #2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
m¼ R g1 þ g1 þ h1 (7:41)
μ0
Analysis of the geomagnetic field for epoch 2010 (Finlay et al., 2010) locates
the north geomagnetic pole at 80.08 °N, 287.78 °E and the south geomagnetic
pole at 80.08 °S, 107.78 °E. The places where the total magnetic field of the
Earth is normal to the surface are the magnetic dip poles. The total field is
expressed by all the terms in (7.14). Because of the non-dipole components the
magnetic dip poles are not antipodal; also, because of secular variation
(Section 7.4) the pole locations change slowly with time. For epoch 2010, the
north dip pole was at 85.01 °N, 227.34 °E; the south dip pole was at 64.43 °S,
137.32 °E, which is outside the Antarctic Circle.

7.3.3 Axial dipole with axial offset

The terms with n = 2 are referred to as the quadrupole component of the field.
However, one must keep in mind that the multipole expression of the magnetic
field is a mathematical convenience that simply allows us to subdivide it for
convenient reference. That is, just as there are no physical magnetic dipoles