Karl Frederick Gauss made extensive studies of

the Earth's magnetic field from about 1830 to 1842,

and most of bis conclusions are still valid. He con
cludcd from mathematical analysis that the magnetic
eld was entirely due to a source within the Earth,
rather than outside of it, and he noted a probable
connection to the Earth's rotation because the axis
of the dipole that accounts f or most of the field is
not far from the Earth's rotational axis.
The terrestrial magnetic ficld has beco studicd
almost continuously since Gilbert's time, but it was
not until 1843 that von Wrede first used variations in
the fteld to locatc dcposits of magnctic ore. Thc
publication, in 1879, of
by Thal~n
marked thc first use of the magnctic method.
Until thc late 1940s, magnetic field measurements
mostly were made with a magnetic balance, which
measured one component of thc carth's eld, usually
the vertical component. This limited measuremcnts
mainly to the land surface. Thc ftuxgatc magnetome
ter was developed during World War II for detecting
submarines from an aircraft. After the war, the ux
gate magnetometcr (and radar navigation, another
war development) made aeromagnetic measuremcnts
possible. Protonprecession magnetometers, devcl
oped in the midl 950s, are very reliable and their
operation is simple and rapid. They are the most
commonly used instruments today. Opticalpump al
kalivapor magnetometers, which began to be uscd in
1962, are so accurate that instrumentation no longer
limits tbe accuracy of magnetic measuremcnts. How
ever, protonprccession and opticalpump magne
tometers mcasure only tbe magnitudc, not the direc
tion, of the magnetic field. Airborne gradiometer
measuremcnts began in the late 1960s, although
ground mcasurcments werc made much earlier. Thc
gradiometer often consists of two magnctometers
vertically spaced 1 to 30 m apart. Thc differcnce in
readings not only gives the vertical gradient, but
also, to a large cxtcnt, rcmoves the eff ccts of tempo
3.1.2. History of Magnetic Methods
The study of the eartb's magnetism is the oldest
branch of geophysics. lt has been known for more
than thrcc ccnturies that the Earth bchaves as a large
and somewhat irregular magnct Sir William Gilbert
(15401603) made the first scicntific investigation of
terrestrial magnctism. He rccorded in de
that knowledge of the northsceking property of a
magnetite splintcr (a or leading stone) was
brought to Europe from China by Marco Polo.
Gilbert showcd that the Earth's magnetic fleld was
rougbly equivalcnt that of a permanent magnct
lying in a general northsouth direction near the
Earth's rotational axis.
3.1.1. General
Magnetic and gravity methods have much in com
mon, but magnctics is generally more complcx and
variations in the magnctic ficld are more erratic and
localizcd. 1bis is partly due to the diff erence bctwccn
the dipolar magnetic eld and thc mooopolar gravity
eld, partly duc to the variable direction of the
magnetic field, wbereas the gravity field is always in
the vertical direction, and partly due to the time
dependence of thc magnetic field, whereas the grav
ity field is timeinvariant (ignoring small tidal varia
tions). Whereas a gravity map usually is donnatcd
by regional c.trccts, a magnctic map gcncrally shows
a multitudc of local anomalics. Magnetic measure
ments are madc more easily and cheaply than most
geophysical measurements and correctioas are prac
tical)y unncccssary. Magnetic ficld variations are ot
ten diagnostic of mineral structures as well as re
gional structurcs, and the magnetic mcthod is thc
most vcrsatile of gcophysical prospecting technques,
Howcver, like ali potential methods, magnetic meth
ods lack uniquencss of interpretation.
3.1. INTRODUCTION
Chapter
3
where H has the SI dimensin amperes per meter
[ 4w x 10
3
oersted], and are in meters, 1 is
in amperes, and H, r1, and l ll/ have the directions
indicated in Figure 3.1.
A current flowing in a circular loop acts as a
magnetic dipolc located at the center of the loop and
oriented in the direction in which a righthanded
scrcw would advancc if tumed in the direction of the
current. lts dipole moment is measured in ampere
meter'' ( 1010 polecm), The orbital motions of
electrons around an atomic nucleus constitute circu
lar currents and cause atoms to have magnetic mo
AH (1 tl./) x r1/411r2 (3.4)
m is a vector in the direction of the unit vector r1
that extends from the ncgative pole toward the pos
tive pole.
A magnetic field is a consequence of the ftow of
an electrical current. As expressed by Ampre's law
(also called the BiotSavart law), a current 1 in a
conductor o length !l./ creares, at a point (Fig.
3.1), a magnetizing eld !J ,.ff given by
(3.3) m 2/pr1
unts); H' is measured in oersteds (equivalent to
dynes per unit pole),
A is envisioned as two potes of
strength + and separated by a distance 21. Tbc
is defined as
where 4H is in amperes per meter when 1 is in amperes
.<\H (1 d/) X r1/4w2
Figure 3. 1 . Ampere 's lew. A current 1 through a length of
conductor creates a magnetizing field 4H . ita pont P:
(we use a prime to indicate that H is in cgsem
where F is the force on p2, in dynes, the poles of
strength p1 and are r centimeters apart, . is the
[a property of the mdium: see
Eq. (3.7)), and r1 is a unit vector directed from p1
toward Pz - As in the electrical case (but unlike the
gravity case, in which the force is always attractive),
the magnetostatic force is attractive for poles of
opposite sign and repulsive Cor poles of like sign. The
sign eonvention is tbat a positioe p< > ie is attracted
toward the Earth's north '1\agnetic pote; the term
is also used.
The H (also callcd
is defined as the force on a unit pole:
1
T

(3.1)
3.2.1. Classical versus Electromagnetic
Concepts
Modem and classical magnetic theory ditrer in basic
concepts, Classical magnetic theory is similar to elec
trical and gravity theory; its basic concept is that
point magnetic poles are analogous to point electri
cal charges and point masses, with a similar inverse
square law Cor the forces between the potes, eharges,
or masses. Magnetic units in the centimetergram
second and electromagnetic units (cgs and emu) sys
tem are based on this concept, Systme Intemational
(SI) units are based on the fact that a magnetic field
is electrcal in origin. lts basic unit is the dipole,
which is created by a circular electrical current,
rather than the ctitious isolated monopole of the
egsemu system. Both emu and SI units are in
currcnt use.
The cgsemu system begins with the concept of
magnetic force F given by Coulomb's law:
3.2. PRINCIPLES ANO ELEMENT ARY
THEORY
ral field variatons, which are often the limiting fac
tor on accuracy.
Digital recording and processing of magnetic data
removed much of the tedium involved in reducing
measurements to magnetic maps. Interpretation al
gorithms now make it possible to produce computer
drawn proles showing possible distributions of
magnetization.
The history of magnetic surveying is discussed by
Reford (1980) and the state of the art is discussed by
Paterson and Reeves (1985).
Principies and elementary theorv
3.2.2. BH Relations: The Hysteresis loop
The relation bctween B and H can be complex in
ferromagnctic matcrials (3.3.5). This is illustrated
by bysteresis (Fig. 3.2) in a cycJ e of magnetization. lf
a demagnctiz.cd sample is subjected to an increasing
magnetizing ficld H, we obtain the ftrst portion of
the curve in which B increases with H until it ftattcns
off as wc approach the maximum valuc that can
have for the samplc When H is de
creased, the curve does not retrace thc samc path,
but it does show a positive value of B when H O;
wherc A is a vector arca (A.3.2). lbus IB I ~/IAI
when A and B are parallel, that is, B is the density o(
magnctic flux. Thc SI unit for magnetic ftux is the
weber ( Trrt) and the cm unit is the maxwell
( 1011 Wb).
(3.8)
Thcrc is often confusion as to wbether tbe quan
tity involved in magnetic exploration is B or H.
Altbough we measure B.,, we are interested in the
Earth's fteld H.,. However, because B and H are
linearly related [Eq. (3.7)) and usually I' l, we can
(and do) treat a map of B,. as if it were a map of H,.
We also speak of or
41> :
ly 109 T 1 nT
wben H and M (H' and M') are in tbc same drec
tion, as is usually the case. The SI unit for B is the
tesla 1 newton/amperemeter 1 weber/meter2
(Wb/nt). Tbe elcctromagnetic unit for B' is the
gauss [ 104 tesla (1)). The permeability of free
spacc has tbe value 4tr x 107 Wb/Am. In
vacuum 1 and in air I' l. Confusion some
times results between H' and B ' because the em units
gauss and oersted are numerically equaJ and dimen
sionally the same, althougb conceptuaJ ly ditrerent;
both H' and B' are sometimes caJ led thc "magnctic
eld st:rength." In magnetic prospecting, we measure
B to about 104 of tbc Earth's main fteld (whicb is
about S O 1). Thc unit of magnetic induction gener
ally used for geopbysical work is thc nanotesla (also
B p0(H + M) l'o(l + k)H l'l'oH (3.7a)
B' H' + 4trM' (1 + 4fl'k')H' l'H' (3.7b)
Susceptibility is tbe fundamental rock parameter in
magnetic prospecting. The magnetic response of
rocks and minerals is determined by the amounts
and susceptibilities of magnetic materials in them.
The susceptibilities o! various materials are listed in
Table 3.1, Section 3.3.7.
The B is tbe total ficld, nelud
ing the eff ect of magnetization. It can be written
(3.6)
Magnetic susceptibility in emu diff ers from that in SI
units by the factor 4w, that is, called the gamma, y):
Figure 3. 2. Hysteresis loop. s. s' saturation, r and r'
remanent magnetism, e and e' - coercive force.
O r O r' - Residual maptilin
Oc O c' - Coerci~ force
H
(3.5) MkH
ments. Molecules also bave spin, whlch gives them
magnetic momcnts.
A magnetizable body placed in an extemaJ mag
netic field becomes magnetized by induction; the
magnetization is due to the reorientation of atoms
and molecules so that their spins line up. The mag
netization is measured by the
M (a1so called or
The lineup of intemal dipoles
produces a field M, wbich, within the body, is added
lo the magnetizing field H. lf M is constant and has
tbe same direction througbout, a body is said to be
The SI unit for magnetization
is amperemeter! per meter3 [ ampere per meter
(A/m)).
For low magnetic elds, M is proportonal to H
and is in the direction of H. The degree to wbich a
body is magnetized is determined by its
k, which is defined by
64
{3.14c)
tan a F,/F, - (1/2)tan9
and the direction with respect to the dipole axis is
F- IF I ( m/r3}(1 + 3 cos2 9)112 (3.14b)
where unit vectors r1 and 8 are in tbe dircction of
increasing and fJ (counterclockwise in Fig. 3.3).
The resultant magnitude is
F QI ( m/r3)(2cosfJ r1 + sinl&) (3.14a)
where mis the dipole moment of magnitude
Equations (3.11) and (3.13) give [A.4 and Equation
(A.33)]
(3.13) mcos /r2
When r I, Equation (3.10) becomes
- p 3/2
( r2 + /2 + cos fJ )
r- lcosfJ }
(3.12a)
( r2 + 2rl cos 9)312
{
I sin 8
F, - p 3 / 2
( r + + 2rl cos
/sin fJ }
+ (3.12b)
( r2 + 1 2r/ cos IJ )
lar component is these are
Figure 3. 3. Calculating tne field ot d megnetic dipote.
- p +p
r.
65
lts radial component is F,. - and its angu
(3.11) F(r) ... VA(r)
We can derive the vector F by taking the gradient o
(Eq. (A.17)]:
1 }
1/2 (3.10)
( r2 + + 2/rcos 8)
{
1
1 / 2
(r2 + /2 2/rcosfJ )
p - !!...)
'1 '2
However, since a magnetic pole cannot exist, we
consider a magnetic dipole to get a realistic entity.
Referring to Figure 3.3, we calculatc atan externa)
point:
(3.9)
A ( r) J ' F( r) dr =p /
OQ
3.2.3. Magnetostatic Potential for a
Dipole Field
Conceptually the magnetic scalar potentiaJ at the
point is the work done on a unit positive pole in
bringing it from infinity by any path against a mag
nctic field F(r) [compare Eq. (2.4)]. (Henccforth in
this cbapter F, F indicate magnctic field rather than
force and we assumc J J - l.) When F(r) is dueto a
positive pole at a distance from P,
this is called magnetism. When
H is reversed, B finally becomes zero at sorne nega
tive value of H knowo as tbe The
other half of tbe hysteresis loop is obtained by
making H still more negative until reverse saturation
is reached and then retuming H to the original
postve saturation valu. The area inside the curve
represents the eoergy loss per cycle per unit volume
as a result of hysteresis (see Kip, 1962, pp. 235 7).
Residual eff ects in magnetic materials will be dis
cussed in more detail in Section 3.3.6. In sorne
magnctic materials, B may be quite large as a result
of previous magnetization having no relation to the
present value of H.
.....
Principies elementary theory
where the drectons of F, and F(r0) are not necessar
ily the same. If F( r0) is much smaller than F, or if the
body has no residual magnetsm, F and F, will be in
approximately the same direction. Whcre F(r0) is an
appreciable fraction (say, 25% or more) of F, and
F F, + F(r0)
The magnetic fteld in Equation (3.20) exists in the
presence of the Earths field F,. that is. the total fteld
F is given by
(3.20)
a ( dv )
lro rl
(Eq. (A.18)) and
)
M V M- - M (- + +
(3.19)
[f Mis a constant vector with direction a ti +
mj + nk. then the operation
F(r0) vf. M(r) v( l
1)
(3.18)
!ro r
The resultant magnetic eld can be obtained by
employing Equation (3.11) with Equation (3.17). This
gives
(3.17)
/.M(r) v(
1
)
!ro rl
the body (Fig, 3.4) is
Tbe potential for the wholc body at a point outsidc
M(r)cos9/r2 M(r) V(l/r) (3.16)
3.2.4. The General Magnetic Anomaly
A volume of magnetic material can be considered as
an assortment of magnetic dipoles that results from
the magnetic moments of individual atoms and
dipoles. Whether they initially are aligned so that a
body exhibts residual magnetism depends on its
previous magnctic history. They will, howcver, be
aligned by induction in the presence of a magnetiz
ing field. In any case, we may regard the body as a
continuous distribution of dipoles resulting in a vec
tor dipolc momcnt per unit volume, M, of magnitude
M. The scalar potential at P [see Fig. 3.3 and Eq.
(3.13)) some distance away from a dipole (r >/)
is
(3.15c)
o }
2m/r3
m/r3
If r >l, these simplify to
F, o
(3.15a)
m/( r2 + 12)
312
6 - '1/2
(3.15b)
F,- 0
Two special cases, 9 - O and .,, /2 in Equation
(3.12), are called the (endon) and
(sideon) positions. From Equations (3.12) they are
given by
Figure 3. 4. General magnelic ;moma/y.
z
.I'
66
3.3.1. Nature of the Geomagnetic Field
As far as exploration gcopbyics is concerned, thc
geomagnetic field of the Earth is composed of three
parts:
1. The main field, wbich vares relatively slowly and
is of intcmal origin.
2. A small field ( compared to the main eld), which
vares rather rapidly and originates outside the
Earth.
3. Spatial variations of the main eld, which are
usually smaller than tbe main eld, are nearly
constant in time and place, and are caused by
local magnetic anomalies in thc nearsurtace crust
of the Earth. These are thc targets in magnetic
prospecting.
3.3. MAGNETISM Of THE EARTH
These relations are used to makc pseudogravity maps
from magnctic data
M/yp)( 1 (3.28)
In particular, if M is vertical, the vertical component
of F is
(3.27b)
Fp - ( M/yp) U tl/J
where dV /da. Por a component of F in the
direction /11, this becomes
{3.27a)
F VA (M/yp)Vg,.
... (M/yp)V(VV ut)
::s ( V
lf we apply this result to an extended body, we
must sum contributions for each element of volume.
Provided that M and p do not change throughout
the body, the potentials and will be those for
the extended body. Therefore, Equations (3.24) to
(3.26) are vaJ id for an extended body with constant
density and uniform magnetization.
In terms of elds,
(3.26)
Thus,
fJ rY 1 a. ~ t'i t: ; e; l J O te. . n
u= tt!? L h~I
M. ::_ m C:...l~t'if? .b(... HO Y t1 . (. 67
nent of g in the direction m is
f 'to ~- '""1
- dU /da = - vU a1 ... ypV(l/r) u1
(3.25)
From Equations (2.3a). (2.5), and (A.18), the compo
A M V{l/r) - M V(l/r) Ui (3.24)
If we have an infinitesimal unit volume with mag
netic moment M and density p, then at a
distant point we have, from Equation (3.16),
3.2.S. Poisson's Relation
{3.23)
In a nonmagnetic medium, M = O and
(3.22) V2A 4trV. M(r)
is the net positive pote strength per unit voJ ume at
a point. We recall that a field F produces a partial
reorientation along thc field direction of the prev
ously randomly oriented elementary dipoles. This
causes, in efl'ect, a separation of positive and nega
tive poles. Por example, the component of F
separates pote strengths + and by a distance r
along tbe x axis and causes a net positive pote
strength ( M. dz to en ter the rear
face in Figure A.2a. Because the pole strength
leaving through the opposite face is { M. +
(a dz , the net positive pole strength
per unit volume ( p) created at a pont by the field F
is V M. Thus,
The magnetic interpretation problem is elearly more
complex than the gravity problem because of thc
dipolar field (compare 2.2.3).
The magnetic potential A, like the gravitational
potential V, satisfies Laplace's and Poisson's equa
tions. Following the method used to derive Equa
tions (2.12) and (2.13), we get
V . F v2A
a 2 ~
F .(r)=Mf - kF- f- -
D O iJ /2 viro r] , a/2 Viro - rl
(3.21b)
where 11 is a unit vector in the drection o( F,
(3.3.2a). Ir the magnetization is mainly induced by
F,, then
has a dfferent direction, the component of F(r0) in
the direction of F,, F becomes [Eq. (3.20)}
aA a2 do
Fo - f1 VA .... a =M aaaf r!
(3.2la)
(b) O rigin of the main field. Spherical harmonic
analysis of the observcd magnetic field shows that
ovcr 99% is duc to sourccs inside the Earth. The
present theory is that the main field is caused by
convection currents of conducting material circuJ at
ing in the liquid outer corc (which extends from
depths of 2,800 to 5,000 km). The Earth's core is
assumed to be a mixture of iron and nickeJ , both
good elcctrical conductors. The magnetic source is
thought to be a selfcxcited dynamo in which bighly
conductive ftuid moves in a complex manner causcd
by convection. Paleomagnetic data show that the
magnetic field has always bcen roughly along the
Earth's spin axis, implying that the convective mo
tion is coupled to the Earth's spin. Rccent explo
ration of the magnetic fte1ds of other planets and
their satellites provide fascinating comparisons with .
the Earth's field.
or positive pole; the end that dips downward in
southcrn latitudes is the southseeking or negative
pole.
Maps showing lines of equaJ declination, inclina
ton, horizontal intensity, and so on, are called
(Fig. 3.6). and
show, respectively, lines of equal
declination inclination and equal values of F,,
H,, or z,. Note that the inclination is large (that is,
Z, >H. ) for most of the Earth's land masses, and
hence corrections do not have to be made Cor lat
tude variations of ~or Z, ( 4 nT/km) exeept tor
surveys covering extensive arcas. The overall mag
netic eld does not reftect variations in surface geol
ogy, such as rnountain ranges, midocean ridges or
earthquakc belts, so the source of the main field les
deep witbin the Earth. The geomagnetic eld resem
bles that of a dipole whose north and south magnctic
potes are located approximately at 75N,l01 W and
69S, 145E. The dipole is displaced about 300 km
from the Earth's center toward Indonesia and is
inclined sorne 11.S to the Earth's axis. However, the
geomagnetic field is more complicatcd than thc field
of a simple dipole. The points where a dip needJ e is
vertical, the are at 75N, 101 W and
67S, 143E.
lbe magnitudes of at the north and south
magnetic poles are 60 and 70 rcspcctively. The
minimum valuc, 25 occurs in southcrn Brazil
South Atlantic. In a few locations, F, is larger
than 300 l'T because of nearsurf ace magnetic fea
tures. The line of zcro inclination (
where O) is never more than 15 from the
Earth's cquator. Thc largcst deviations are in Soutb
America and the eastem Pacific. In Africa and Asia
it is sligbtly north of the equator.
Magnetic methods
As stated earlier, the end of the needle that dips
downward in northern latitudes is the northseckiog
F, .f.li .f.( cos D cos li
+sin /J +sin /k)
Y, H,sinD
tan/ Z,/H,
X, - H,cos D
tan D - Y ,/X,
(3.29)
F.2 _ + z2 _ xi + y:2 + z2
' ' ' ' ' '
H, cos 1 Z, sin 1
3.3.2. The Main Field
(a) The Earth's magnetic field. lf an unmagnctizcd
steel needle could be hung at its eenter of gravity, so
lhat it is free to orient itself in any dircction, and if
otber magnetic fields are absent, it would assume the
dircction the Eartb's total magnctic eld, a drec
lion that is usually neither horizontal nor Inline with
the geographic meridian. The magnitude of this field,
F. , tbe or of the needle from the
horizontal, l, and the angle it makes with geographic
north ( the D, completely define the
main magnetic field.
The (Whitham, 1960) are illus
trated in Figure 3.5. The eld can also be describcd
in terms of the vertical component, Z,, reckoned
positive downward, and tbe horizontal component,
H,, which is always positive. X, and are the
components of H,, which are considered positive to
tbe ftorth and east, respectively, These elements are
re1ated as follows:
Figure 3. 5. emem of tbe Earth's magnetic
1
1
1
1
1
1 1
1 1
1 1
rr
1 1 ,'
1
I lt
;ir Geopapllic
North
68
.':~~~
. ,.
')
\
f \
..
1 1
1
1
I
l'
/
/
/
/
3.3.5. Magnetism of Rocks and Minerals
Magnetic anoma1ies are caused by magnetic mincrals
(mainly magnetite and pyrrhotite) contained in the
rocks. Magnetically important minerals are surpris
ingly few in number.
Substances can be divided on the basis of their
behavior when placed in an extemal field. A sub
stance is if its field is dominated by
atoms with orbital electrons orientcd to oppose the
extemal field, that is. if it exhibits negative suscepti
bility. Oiamagnetism will prevail only if thc net
magnetic moment of all atoms is zero when H is
zero, a situation characteristic of atoms with com
pletely filled electron shells. The most common da
magnetic earth materials are graphite, marble, quartz,
and salt. When the magnctic moment is not zero
when H is zero, the susceptibility is positive and the
substance is The eft'ects of diamag
netism and most paramagnctism are weak.
Certain paramagnetic clements, namcly iron,
cobalt, and nickel, have such strong magnctic inter
action that the moments align within rairJ y large
regions called This eff'ect is called
and it is 106 times thc eft'ects of
diamagnctism and paramagnetism. Ferromagnetism
dec:reases with increasing temperature and disap
pears entirely at the Curie temperature. Apparently
f erromagnetic minerals do not exist in nature.
The domains in some materials are subdivided
into subdomains that align in oppositc direc:tions so
that their moments nearly cancel; although they
would otherwise be considered ferromagnetic, the
susccptibility is comparatively low. Such a substance
is 1be only common example is
bematite.
In some matcrials, the magnetic subdomains align
in opposition but thcir net momcnt is not zero, either
bec:ause ooe set of subdomains has a stronger mag
netic alignment than the other or because there are
more subdomains of one type than of the other.
Thcse substanccs are Examples of the
fint type are magnetite and titanomagnetite, oxides
of iron and of iron and titanium. Pyrrhotite is a
magnetic mineral of the second type. Practically all
magnetic minerals are ferrimagnetic.
thc Canadian Shield, for example, shows a magnetic
contrast to the Western Plains). Many largc, erratic
variations often makc magnetic maps extremely
complex. The sources of local magnetic anomalies
cannot be very deep, because temperatures below
40 km should be above the the tem
perature ( 5S0q at which rocks tose their mag
netie properties. Thus, local anomalies must be asso
ciated with fcatures in the upper crust.
Masnetic
3.3.4. Local Magnetic Anomalies
Local changes in the main field result from varia
tions in the magnetic mineral content of nearsurface
rocks. These anomalies occasionalJ y are J argc enough
to double the main field. They usually do not persst
over great distances; thus magnetic maps gencrally
do not exhibir largescale regional features (although
3.3.3. The Extemal Magnetic Field
Most of the remaining small portian of the geomag
netc field appears to be associated with electric
currents in the ionzed laycrs of tbe upper atmo
sphere. Time variations of this portion are much
more rapid than for the main "permanent" field.
Some effects are:
1. A cycle of U years duration that correlatcs with
sunspot activity.
2. Solar diumal variations with a period of 24 b and
a range of 30 nT that vary with latitude and
season, and are probably controUed by action of
the solar wind on ionospheric currents.
3. Lunar variations with a h period and an am
plitude 2 nT that vary cyclically throughout
the month and seem to be associated with a
Moonionospberc interaction.
4. Magnetic storms that are transient disturbances
with amplitudes up to 1,000 nT at most latitudes
and even larger in polar regions, where they are
associated with aurora. Although erratic, they of
ten occur at 27 day intervals and correlate with
sunspot activity. At the beight of a magnetic
storm (which may last for several days), longrange
radio reception is affected and magnetic prospect
ing may be impractical.
These time and space variations of thc Earth's
main field do not significantly affect magnetic
prospecting except for the occasional magnetic storm.
Diumal variations can be corrected for by use o a
basestation magnetometer. Latitude variations ( 4
nT /km) require corrections only for higbresolution,
higblatitude, or largescale surveys.
(e) S ecular vsristions of tbe main field. Four hun
dred years of contnuous study of the Earths eld
has established that it changcs stowty. The inclina
tion has changed sorne 10 (75 to 65) and the
dcclination about 3S (lOE to 2SW and back to
lO"W) during this period. The source of this wander
ing is thought to be changcs in convection currents
in the core.
The Earth's magnetic field has also reversed drec
tion a number of times. The times of many of the
periodic eld reversals have been ascertained and
provide a
72
3.3.8. Magnetic Susceptibility
Measurements
(a) Measurement of k, Most measurements of
invoJ vc a comparison of the sample with a standard.
The simpJ cst laboratory method is to compare the
deection produced on a tangent rnagnetometer by a
3.3.7. Magnetic Susceptibilities of Rocks
and Minerals
Magnetic susccptibility is the significant variable in
magnetics, lt plays the same role as density docs in
gravity interpretation. Although instruments are
available for measuring susceptibility in the eld,
they can only be used on outcrops or on rock sam
ples, and such measurements do not necessarily givc
the bulle susceptibility of tbc formation.
From Figure 3.2, it is obvious that (hence I'
also) is not constant for a magnetic substance; as H ,
increases, k increases rapidly at rst, reaches a maxi
mum, and then decreases to z.ero. Furthermore, a).
though magnetization curves have the same general
shape, the value of H for saturation vares grcatly
with thc typc of magnetic mineral. Thus it is impor
tant in making susceptibility determinations to use a
value of H about the same as tbat of thc Earth's
field.
Since tbe f errimagnetic minerals, particularly
magnetite, are the main source of local magnctic
anomalies, there have bcen numerous attcmpts to
establish a quantitative relation between rock sus
ceptibility and Fe_,04 concentration. A rougb linear
dependence (k ranging from 103 to 1 SI unit as thc
volume percent of Fe304 increascs from 0.05$ to
35%) is shown in one report, but the scatter is large,
and results f rom other arcas diff er.
Table 3.1 lists magnetic susceptibilities for a vari
ety of rocks. Although there is grcat variation, even
for a particular rock, and wide overlap between
dff erent types, sedimentary rocks have the lowest
average susceptibility and basic igncous rocks have
the highest. In every case, the susceptibility depends
only on the amount of Ierrimagnetic minerals pre
sent, mainly magnetite, sometimes titanomagnetite
or pyrrhotite. The values of chalcopyrite and pyritc
are typical of many sulfide minerals that are basi
cally nonmagnctic. It is possible to locate minerals of
negative susceptibility, although the negative values
are very small, by means of detailed magnetic sur
veys. It is also worth noting that many iron minerals
are only slightly magnetic,
laboratory methods separate residual from induced
magnetization, something that cannot be done in the
ficld.
1. (TRM), which re
sults wbcn magnctic material is cooled below the
Curie point in the presence of an externa] field
(usually the Earth's eld), lts direction depends
on the direction of the field at the time aod place
where the rock cooled. Remanence acquired in
this fashion is particularly stable. This is the main
mechanism for the residual magnetization of ig
neous rocks.
2. (DRM), which occurs dur
ing the slow scttling of finegrained particles in
the presence of an external field. Varicd clays
exhbt this type of remanence.
3. (CRM), which
takes place when magnetic grains increase in size
or are changed from one Iorm to another as a
result of chemical action al modrate tempera
tures, that is, below the Curie point. This process
may be significant in sedimcntary and metamor
phic rocks.
4. (IRM), which
is the residual left following the removal of an
externa] field (see Fig. 3.2). Lightning stri.kes pro
duce IRM over very small arcas.
(VRM), wbich is
produccd by long exposure to an external field;
the buildup of remanence is a J ogarithmic Iunc
tion of time. VRM is probably more characteristic
of fnegrained than coarsegrained rocks. This
remanence is quite stable,
Studies of the magnetic history of the Earth
indicate that the Earth's field has
varied in magnitude and has reversed ts polarity a
number of times (Strangway, 1970). Furthermore, it
appears that the reversals took place rapidly in geo
logic time, because there is no evidence that the
Earth existed without a magnetic field for any signif
icant period. Model studies of a selfexcited dynamo
show such a rapid tumover. Many rocks have rema
nent magnetism that is oriented neither in the drec
tion of, nor opposite to, the present Earth field, Such
results support the plate tectonics theory, Paleomag
netism belps agedate rocks and determine past
movements, such as plate rotations. Paleomagnetic
3.3.6. Remanent Magnetism
In many cases, the magnetizaton of rocks dcpcnds
mainly on the present geomagnctic ficld and the
magnetic mineral content. Residual magnetism
(called NRM) oftcn
contributes to the total magnetization, both in ampl
tude and direction. The effect is complicated because
NRM depends on the magnetic history of the rock.
Natural remanent magnetization may be due to sev
eral causes. The principal ones are:
Magnetism of the E arth
Magnetic methods
Table 3.1. Magnetic susceptibiliies of varous rocks and minertJ ls.
Susceptibility 1<> 3 (SI)
Type Range Average
S edimenrary
Dolomite 00.9 0.1
limestones 0- 3 0.3
Sandstones 020 0.4
Shales 0.0115 0.6
Av. 48 sedimentary 018 0.9
MetJ morphic
Amphibolite 0.7
Schist 0.33 1.4
Phyllite 1.5
Gneiss 0.125
Quartzite 4
Serpentine 317
SI ate 035 f,
Av. 61 metamorphic 070 4.2
Granite 050 2.5
Rhyolite 0.235
Dolorite 135 17
Augitesyenite 3040
Olivi nedi abase 2 5
Di abase 1160 S S
Porphyry 0.3200 60
Cabbro 190
Basalts 0.2175 70
Diorite 0.6120 85
Pyroicenite 125
Peridotite 90200 150
Andesite 160
Av. acidic igneous 080 6
Av. basic igneous 0.597 2 5
Minera Is
Graphile 0.1
Quartz 0.01
Rock salt 0.01
Anhydrite, gypsum 0.01
Calcite 0.001 0.01
Coal 0.02
Clays 0.2
Chalcopyrite 0.4
Sphalerite 0.7
Cassiterite 0.9
Siderite 1 - 4
Pyrite 0.055 1.5
Limonite 2.5
Arsenopyrite 3
Hematite 0.535 &.5
Chromite 3110 7
Franklinite 430
Pyrrhotite 16000 1500
llmenite 3003500 1800
Magnetite 120019200 6COO
3.4.2. Fluxgate Magnetometer
This device was originally developed during World
War 11 as a submarine detector, Several designs bave
been used for recording diumal variations in the
Earth's eld, for airborne geomagnetics, and as
portable ground magnetometers,
The fluxgate detector consists esscntially of a core
of magnetic material. such as mumetal, pennal1oy,
or ferrite, that has a very high permeability at low
magnetic fields. In the most common design, two
cores are each wound wi th primary and secondary
coils, the two assemblies being as nearly as possible
identical and mounted parallel so that the windings
are in opposition. The two primary windings are
connected in series and energized by a low frequency
(50 to 1,000 Hz) current procluced by a constant
current source. The mximum current is sufficient to
magnetize the cores to saturation, in opposite polar
i ty, twice each cycle. The secondary colls, which
consist of many turns of fine wire, are connected to a
whose output is proportional to
the diff erence bctwecn two input signals.
The effeet of saturation in the fluxgate elements is
illustrated in Figure 3.7. In the absence of an exter
nal magnetic ficld, the saturation of the cores is
Typical sensitivity required in ground magnetic in
struments is between l and 10 nT in a total eld
rarely larger than 50,000 nT. Recent airbome appli
cations, however, have led to the development of
magnetometers with sensitivity of 0.001 nT. Sorne
magnetometers measure the absolute eld, although
this is nota particular advantage in magnetic survey
ing.
The earliest devices used for magnetic exploration
were modifications of the mariner's compass, such as
the Swedish mining compass, which measured dip I
and declination D. lnstruments (such as
which are essentially dip needles of high
sensitivity) were developed to measure and
but they are seldom used now. Only the modero
instruments, the ftuxgate, protonprecession, and op
ticalpump (usually rubdurnvapor) magnetometers,
will be discussed. The latter two measure the abso
tute total eld, and the uxgate instrument a1so
generally measures the total eld.
3.4.1. General
3.4. FIELO INSTRUMENTS F OR
MAGNETIC MEASUREMENTS
Overton, 1981). They achieve great sensitivity be
cause of the high magnetic moments and low noise
obtainable at superconducting ternperatures.
r

1
r
(b) Measurement of remenent magnetism. Mea
surement of remanent susceptibilty is considerably
more complicated than that of One method uses
an astatic magnetometer, which consists of two mag
nets of equal moment that are rigidly mounted paral
lel to each other in the same horizontal plane with
opposing poles. The magnetic system is suspended
by a torsin ftber. The specimen is placed in various
orientations below the astatc system and the angular
deflections are measured. This device, in eff ect, mea
sures the magnetic field gradent, so tbat extraneous
fields must eitber be eliminated or made unifonn
over the region of the sample. Usually the entire
assembly is mounted inside a threecomponent col
system that cancels tbe Eartb's fteld.
Anothcr instrument for tbc analysis of the resid
ual component is the The
rock sample is rotated at high speed near a small
pickup coil and its magnetic moment generates alter
nating current (ac) in the coil. The phase and inten
sity of the coil signal are compared with a reference
signal generated by the rotating system. The total
moment of the sample is obtained by rotating it
about diff erent axes.
Cryogenic instruments for determining two
axes remanent magnetism have been developed
(Zimmerman and Campbell, 1975; Weinstock and
l
'
,
~
d, and d, are the deections for the sample and
standard, respectively. The samples must be of the
same size.
A similar comparison method employs an induc
tance bridge (Hague, 1957) having several aircore
coils of diff erent cross sections to accommodate sam
ples of different sizes. The sample is inserted into
one of the coils and the bridge balance condition is
compared with the bridge balance obtained when a
standard sample is in the coil. The bridge may be
calibrated to give susceptibility directly, in which
case the sample need not have a particular geometry
(although the calibration may not be valid for sam
ples of highly irregular shape). This type of instru
ment with a large diameter coil is used in field
measurements on outcrop. The bridge is balanced
first with the coil remote from the outcrop and then
lying on it. A calibration curve obtained with a
standard relates and the change in inductance.
prepared sample ( either a drill core or powdered
rock in a tube) with that of a standard sample of
magnetic material ( often FeCl
3
powder in a test
tube) when the sample is in the GaussA position
[Eq. (3.lSa)]. The susceptibility of the sample is
found from the ratio of deflections:
...1
'
Field instruments far magnetic messurements
FiBure 3. 8. Portable fluxgate megnetometer.
symmetrical and of opposite sign near the peak of
eacb halfcycle so that the outputs from thc: two
secondary windings cancel. The presence of an exter
nal field component parallel to the cores causes
saturation to occur earlier for one halfcycle than the
other, producing an unbalance. The derence be
tween output voltages from the secondary windings
is a series of voltage pulses wbich are fed ioto the
amplifier, as shown in Figure 3.7d. The pulse hcigbt
is proportional to the amplitude of the biasing eld
of the Earth. Obviously any component can be mea
sured by suitable orientation of the cores.
The original problem with tbis type of magnc
tometer a lack of sensitivity n the core has been
solved by thc: development and use of materials
having sufficient initial permeability to saturate in
small fields. Clearly the hysteresis loop should be as
tbin as possible. Thcrc rcmains a relatively higb
noise level, caused by hysteresis cff ects in thc core.
The tluxgate e1emcnts should be long and thin to
reduce cddy currcnts. Improvcmcnts introduccd to
increase thc sgnaltonoise ratio include the follow
ing:
l. By deliberately unbalancing the two elements,
voltage spikcs are present with or without an
ambicnt ficld. The presence of the Earth's fteld
increases the voltage of one polarity more than
thc other and this diff erenee is amplificd.
2. Because the odd harmonics are canceled fairly
Fisure Principie of the fluxgate magnetometer. Note ttut He Fe, etc. (From
Whitham, 1 960. ) (a) Magnetz aton of tbe cores. (b) Flux in the two cores for F,
(e) Flux in the cores for F. O . (d) + fi tor F. " O . (e) O utput volt. J ge tor F, " O .
(a)
H, O
ltll
di
H,,. O
, , ; 1
. . . Tt
A_
A
y
V
81 "I'-
M .. neiir.alion
curves for
lluipte cores
B,, B,
H, ""'O
B,, B,
Magnetic methods
Prima ry a .c. Iie Id ..=.; ....
in the 2 coils
H,- O
7 6
where the factor 2'f1 /'Y p =23.487 0.002 nT/Hz.
Only the total ficld may be mcasured.
(3.30b)
The constant is the
the ratio of ts magnetic moment to its spin
angular momentum. The value of is known to an
accuracy of 0.001 %. Since precise frequency mea
surements are relatvely easy, the magnetic field can
be determined to the same accuracy. The proton,
which is a moving charge, induces, in a coil sur
rounding the sample, a voltage that vares at the
precession frequency Thus we can determine the
magnetic field from
(3.30a) =
havc a net rnagnetic moment that, coupled with their
spin, causes them to precess about an axial magnetc
eld.
The protonprecession magnetorneter depends on
the measurement of the freeprecession frequency of
protons (hydrogen nucJ ei) that have been polarized
in a direction approximately normal to the direction
of the Earth's field. When the polarizing field is
suddenly removed, the protons precess about the
Earth's field like a spinn.ing top; the Earth's field
supplies the precessing force corresponding to that
of gravity in the case of a top. The analogy is
llustrated in Figure 3.9. The protons precess at an
angular velocity known as the
which is proportional to the magnetic field
F, so that
3.4.3. ProtonPrecession Magnetometer
This instrument grew out of tbe discovery, around
1945, of nuclear magnetic resonance. Sorne nuclei
well in a reasonably matched set of cores, the
even harmonics (generally only the second is sig
nificant) are amplified to appear as positive or
negative signals, depending on the polarity o the
Earth's field.
3. Most of the ambient eld is canceled and varia
tions in the remainder are detected with an extra
secondary winding,
4. Negative feedback of the amplifier outputs is used
to reduce the effect of the Earth's field.
5. By tuning the output of the secondary windings
with a capacitance, the second harmonic is greatly
increased: a phasesensitive detector, rather than
the difference amplifier, may be used with this
arrangement.
There are several fundamental sources of error in
the ftuxgate instrument. These include inherent un
balance in the two cores, thermal and shock noise in
cores, drift in biasing circuits, and temperature sensi
tivity (1 nT /C or less). Thcse disadvantages are
minor, however, compared to the obvious advan
tages direct readout, no azimuth orientation, rather
coarse leveling requirements, light weight (2 to 3 kg),
small sze, and reasonable sensitivity. Another at
tractive feature is that any component of the mag
netic field may be measured. No elaborate tripod is
required and readings may be made very quickJ y,
generally in about 15 s. A portable fluxgate instru
ment is shown in Figure 3.8.
9.
Earth'
1ravity field
Gnvit.ional
torque on top
~

Gyration
. :
..
Magnetic torque
on proton
Earth's
manetic licld Ma1netc
F momcnt Spin
r Prccession momcntum
;th
Field instruments for megneticmessurements
3.4.4. Optically Pumped Magnetometer
A variety of scientific instruments and tecbniques
ha.s been developed using the energy in transferring
atomic electrons from one energy leve] to anotbcr.
For example, by irradiating a gas with light or
radiofrcqucncy waves of the proper frequency, elee
trons may be raised to a higher cnergy level. If they
can be accumulated in such a state and thcn sud
denly rctumed to a lower level, they releasc some of
their cnergy in the process. This cnergy may be used
for amplification (masers) or to gct an intense light
beam, such as that produeed by a laser.
Thc optically pumpcd magnctomctcr is another
application. The principie of operation may be un
derstood from an examination of Figure 3.lla, whi?t
shows three possible energy levcls, A1, A2, and
for a hypothetical atom. Under normal conditions of
pressure and temperature, thc atoms occupy ground
state levels A1 and
2
The energy dift"ercnce be
twcen A1 and A2 is vcry small (,. 10electron
volts (eV)J , represcnting a fine structure dueto atomic
electron spins that normally are not alt aligned in tbe
a fixed installation, it posses sorne problcms in small
portable equipment.
The protonprecession magnetometer's scnsitivity
( 1 nT) is higb, and it is essentially free from drift.
The fact that it requires no orientation or lcveling
malees it attractive for marine and airbomc opera
tions. lt has essentially no mechanical parts, al
though the electronc components are rclatively com
plex, The main disadvantage is that only the total
field can be measured. It a1so cannot record continu
ously because it requires a sccond or more between
readings. In an aircraf t traveling at 300 kmjhr, the
distance interval is about 100 m. Protonprecession
magnetometcrs are now thc donnant instrument for
both ground and airbome applications.
The essential components of this magnetometer
include a source of protons, a polarizing rnagnetic
fteld considerab)y stronger than that of the Earth
and drected roughly normal to it (tbe direction of
this fieJ d can be off by 45), a pickup coil coupled
tightly to the source, an amplificr to boost the minute
voltagc induccd in the pickup coil, and a frequeney
measurng device. The latter operates in the audio
range because, from Equation (3.30b), " 2130 Hz
for 50,000 nT. It must also be capable of indi
cating frequency diff erences of about 0.4 fu for an
instrument scnsitivity of 10 nT.
The protn source is usually a small bottle of
water (the nuclear moment of oxygen is zcro) or
sorne organic fluid rich in hydrogen, such as alcohol.
The polarizing field of 5 to 10 mT is obtained
by passing direct current through a solenoid
wound around the bottle, which is oriented roughly
eastwest Cor the measurement. When tbe solenoid
current is abruptly cut off, the proton precession
about the Earth's eld is detccted by a second coitas
a transient voltage building up and decaying over an
interval of 3 s, modulated by the precession fre
quency. In sorne models the same coil is used for
both polarization and detcction. The modulation sig
nal is amplified to a suitable level and the frequcncy
measurcd. A schematic diagram is shown in Figure
3.10.
The measurcment of frequency may be carried
out by actually counting preccssion cyclcs in an
exact time nterval, or by comparing thcm with a
very stable frequency gencrator. In ene ground
model, the precession signa) is mixed with a signal
from a local oscillator of high precision to produce
lowfrequency beats ( 100 Hz) that drive a vibrat
ing reed frequency meter. Regardless of the method
used, thc frequency must be measured to an accu
racy of 0.001 % to realize the capabilities of tbe
method. Althougb this is not particuJ arly difficult in
Figure 3. 1 0. Proton- precession magnetometer. (From S heriff, 1 984. )
G11t
Co.nter to ''" 11tt
1fltr 1h1t !100 crctu
~... ''"''
01c 1111110~ of
~"""" , , .~ .. llC
Magnetic
T .,., lt ctnlrtl
llcll
'r .,,,1, to .._.,..__,
11111u11co11,
tltrOI 1111,11
78
where y, is the
For Rb, the vale of y,/2.,, is approxirnately 4.67
Hz/nT whereas the corresponding frequency for ~
50,000 nT is 233 kHz. Because )', for the electron
is known to a precision of about 1 part in 107 and
because of the relatively high frequencies involved, it
( 3.31)
the energy levels 1 and
2
(actually the sublevels
are more complicated than this, but the simplifica
tion illustrates the pumping action adequately), and
there is a difference of one quantum of angular
momentum between the parallel and antiparallel
states. The irradiating beam is circularly polarized so
that the photons in the light bearn have a single spin
axis. Atoms in subleve} A1 then can be pumped to B.
gaining one quantum by absorption, whereas those
in A2 already have the same momentum as and
cannot make the transition.
Figure 3.12 is a schematic diagrarn of the rubid
iumvapor magnetometer. Light from the Rb lamp is
circularly polarized to illuminate the Rb vapor cell
after which it is refocused on a photocell, The axis of
this bearn is inclined approximately 45 to the Earth's
field, which causes the electrons to precess about the
axis of the eld at the Larmor requency. At one
point in the precession cycle the atoms will be most
nearly parallel to the lightbeam direction and one
half cycle later they will be more antiparallel. In the
rst position. more light is transmitted through the
cell than in the second. Thus the precession Ire
quency produces a variable light intensity that ick
ers at the Larmor frequency. If the photocell signa) is
amplified and f ed back to a coil wound on the cell,
the coilamplifier system becomes an oscilJ ator
whose frequency is given by
.

'
1
!
same direction. Even tbermal energies ( =10
2
eV)
are much larger than this, so that the atoms are as
J ikely to be in level A1 as in A2.
Leve! B represents a much higher energy and the
transitions from A1 or A2 to correspond to in
frared or visible spectral lines. If we irradiate a
sample with a bcam from which spectral line A
2
B
has been removed, atoms in leve) A1 can absorb
energy and rise to B. but atoms in A2 wiU not be
excited, Wben the excited atoms fall back to ground
state, tbey may return to either level, but if they Iall
to A1, they wiJ J be removed by photon excitation to
8 again. The result is an accumulation of atoms in
level A2
The technique of overpopulating one energy leve]
in this Iashon is known as As thc
atoms are moved from leve) A1 to A
2
by this selec
tive process, 1ess energy will be absorbed and the
sample becornes increasingly transparent to the irra
diating beam. When ali atoms are in the A2 state, a
photosenstive detector will register a maximurn cur
rent, as shown in Figure 3.llb. lf now we apply an
RF signal, having energy corresponding to the tran
sition between A1 and A
2,
the pumping eff ect is
nu11ified and the transparency drops to a mnimum
again. The proper frequency for this signal is given
by si E/h. where E is the energy difference be
tween A1 and A2 and is Planck's constant 16.62 X
10
34
jouleseconds].
To malee this device into a magnetometer, it is
necessary to select atoms that have magnetic energy
sublevels that are suitably spaced to give a measure
of the wealc magnetic field of the Earth. Elements
that have been used Ior this purpose include cesium,
rubidium, sodium. and heliurn. The first three each
have a single electrn in the outer shell whose spin
axis lies either parallel or antiparallel to an external
magnetic eld. These two orientations correspond to
J
..
Figure J 1 1 . O pticel pumping. (a) fnergy leve! trensttions. (b) ol pumping 01 1
lisht
(b)
7 9
far
- B It/J Min.
Random ""'""!++~ Microammeter cumut
distribution 5 ~ t.~1rj. ~
~~~=:;:;~~: .. . .. . ~; /~f.{~ ? 4::=~_::~::Y 'LJ
=:r L &mp FilW' A rpUoll Pbococell
2 remoYet spoctral cell Mu.
' ;~ : c;.=::~:_: ~_(_,_~_;;:1_~_1...J if~ilkJ to[i
Min.
1 1 "I
j1. _; . Ai=1 t- i- - - - =- - - 4- : ~
11
Pumpinanulli1ied_' 11""_'....:,~.:_=:_lf:=:i: :::1 ; .r - v- 'J l_ _ J
by RF sipal; ~ f
A i 1 1 1 1 A i
A1 A1 RF sisnaJ
(a)
Magnetic exploration is carried out on land, at sea,
and in thc air. For arcas of appreciable extent,
surveys usually are done with the airbome magne
torneter.
In oil exploraton, airbome magnetics (along with
surface gravity) is done as a preliminary to seismic
work to establish approximate depth, topography,
and character of the basement rocks. Since the sus
ceptibilities of sedimentary rocks are relatively small,
the main response is dueto igneous rocks below (and
sometimes within) the sediments.
Within thc last few years it has become possible
to extraer from aeromagnetic data wcak anomalies
originating in sedimentary rocks, such as result from
tbe faulting of sandstones. This results from (a) the
improved sensitivity of magnetometers, (b) more pre
cise determination of location with Doppler radar
(B.S), (e) corrections for diumal and other temporal
3.5.1. General
3.5. FIELD OPERATIONS
(3.32b) !lH 9. 0N!ll/a
whcre is in microamperes, in nanoteslas, and a
in mcters. Bccause varies directly with thc cur
rcnt, this can be written
(3.32a) Ha 9. 0Nl/a
method ernploys a large enough to
surround the instrument. This is a pair of identical
coils of N turns and radii coaxially spaced a
distance apart equal to the radius. The resulting
magnetic eld, for a current I flowing through the
coils connected in seriesaiding, is directed along the
axis and is uniform within about 6~overa cylinder
of diameter and lcngth concentrlc with the
coils. This tield is gven by
3.4.7. Calibration of Magnetometers
Magnetometers may be calibrated by placing them
in a suitably oriented variable magnetic field of
known valuc. The most dependable calibration
3.4.6. lnstrument Recording
Originally the magnetometer output in airbome in
stallations was displayed by pen recorder. To achieve
both bigb sensitivity and wide rangc, the graph would
be "paged back" (the relerence value changed) Ire
quently to prevent the pen from running off the
paper. Today recording is done digitally, but gener
ally an analog display is also made during a survey.
Some portable instruments for ground work also
digitally record magnetometer readings, station coor
dnates, diumal corrections, geological and terrain
data.
3.4.5. Gradiometers
The sensitivity of the optically pumped magnetome
ter is considerably greater than normaJ ly required in
prospecting. Since 1965, opticaJ ly pumped rubidium
and eesumvapor magnetometers bave been ncreas
ing)y empJ oyed in airbome gradiometers. Two detec
tors, vertically separated by about 35 m, measure
dF /dz, the totaleld vertical gradient. The senstv
ity is rcduced by pitcb and yaw of the two birds.
Major improvements by the Geological Survey of
Canada involve reducing the vertical separation to 1
to 2 m and using a more rigid conncction bctween
lhe sensors. Gradient measurements are also made in
ground surveys. The two sensors on a staff in the
Scintrex MP3 protonmagnetometer system, for ex
ample, measure the gradient to 0.1 nT/m. Gra
diometer surveys are discussed further in Section
is not difficult to measure magnetic field variations
as small as 0.01 nT with a magnetometer of this type.
Figure 3. 1 2. Rubidium- vspor magnetometer (schemetic).
1lecordcr Bias Frequcncy __ _.
s1eppcr mc1er
r ..,
1..
Magnetic
(e) Effect of variations in flight path. Altitude
differences between ftight lines may cause herring
bone pattems in the magnetic data. Bhattacharyya
(1970) studied errors arising from tligh t deviations
td) Flight pattern. Aerornagnetic surveys almost
always consist of parallel lines (Fig. 3.1 J c) spaced
anywhere from 100 m to severa) kilometers apart.
The heading generally is normal to the main geologic
trend in the area and altitude usuaJ ly is maintained
al fixed elevations, the height being continuously
recorded by radio or barometric altimeters, lt is
custornary to record changes in the Earth's eld with
time (dueto diumal or more sudden variations) with
a recording magnctometer on the ground. A further
check generally is obtained by flying severa) cross
lines, wbich verify readings at line intersections,
A which approximates constant
clearance over rough topography, is generally own
with a helicopter. lt is often assumed that drape
surveys minirnize magnetic terrain eff ects, but Grauch
and Campbell (1984) dispute this. Using a uniformly
magnetized model of a mountainvalley systern, four
profiles ( one leve), the others at diff erent ground
clearance) ali showed terrain effects. However,
Grauch and Campbell recommend drape surveys
ovcr levelflght surveys because of greater sensitivity
to smaJ I targets, particularly in valleys, The disad
vantages of draped surveys are higher cost, opera
tional problems, and less sophisticated interpretation
techniques.
(e) S tabiliz ation. Since protonprecession and opti
cally pumped magnetometers measure total field, the
problem o stable orientation of the sensing element
is minor. Although the polarizing field in the
protonprecession instrument must not be parallel to
the totaleld direction, practically any other orien
tation will do because the signal amplitude becomes
inadequate only within a cone of about 5.
Stabilization of the uxgate magnetometer is more
difficult, because the sensing element must be main
tained accurately in the F axis. This is accomplished
with two additionaJ ftuxgate detectors that are or
ented orthogonally with the first; that is, the three
elernents form a threedimensional orthogonal coor
dinate system. The set is mounted on a small plat
form that rotares freely in a11 directions. When the
sensing uxgate is accurately aligned along the
totaleld axis, there is zero signal in the other two.
Any tilt away from this axis produces a signal in the
control elements that drive servomotors to restore
the system to the proper orientation.
rnounting J ocation. Figure 3.13b shows an installa
tion with the magnetometer head in the tail.
81
(b} lnstrument mounting. Aside from stabiliza
tion, there are certain problems in mounting the
sensitive magnetic detector in an airplane, because
the latter has a complicated magnetic eld of its
own. One obvious way to eliminate these eff ects is to
tow the sensing element some distance behind the
ai.craft. This was the original mounting arrangement
and is still used. The detector is housed in a stream
lined cylindrical container, known as a con
nected by a cable 30 to 150 m long. Thus the bird
may be 75 m nearer the ground than the aircraft. A
photograph of a bird mounting is shown in Figure
3.l 3a.
An altemative scheme is to mount the detector on
a wing tip or slightly behind the tail, The stray
magnetic eff ects of tbe plane are minimized by per
manent magnets and soft iron or permalloy shielding
strips, by currents in compensating coils, and by
metallic sheets for electric shielding of the eddy
currents. The shielding is a cutandtry process, since
the magnetic effects vary with the aircraf t and
+
J .5.2. Alrborne Magnetic Surveys
(a) General. In Canada and sorne other countries,
govemment agencies have surveyed much of the
country and aeromagnetic maps on a scale of 1 rnile
to the inch are available at a nominal sum. Large
areas in ali parts of the world have also becn sur
veyed in the course of oil and mineral exploration.
The sensitivity of airborne magnetometers is gen
erally greater than those used in ground explora
tion about 0.01 nT comparcd with 10 to 20 nT.
Because of the initial large cost of the aircraft and
availability of space, it is practical to use more
sophisticated equipment than could be handled in
portable instruments; their greater scnsitivity is use
ful in making rneasurements severa! hundred meters
above the ground surface, whereas the same sensitv
ity is usuaJ ly unnecessary (and rnay even be undesir
able) in ground surveys.
field variations, and (d) computeranalysis tech
niques to remove noise eff ects,
Airborne reconnaissance for minerals frequently
combines magnetics with airborne EM. In most cases
of f ollowup, detailed ground magnetic surveys are
carried out. The method is usually indirect, that is,
the primary interest is in geological rnapping rather
than the mineral concentration per se. Frequently
the association of characteristic magnetic anomalies
with basemetal suldes, gold, asbestos, and so on,
has bcen uscd as a marker in mineral exploration,
There is also, of course, an application f or magnctics
in the direct search for certain iron and titanium
ores.
Field operstions
j
'
1
1
1
1
,
1
I
'
,
I
,
(b) Corrections. In precise work, either repeat
readings should be made every few hours at a prev
ously occupied station or a basestation recording
magnetometer should be employed. This providcs
corrections for diumal and erratic variations of the
magnetic eld, However, such precautions are un
necessary in most mineral prospecting because
anomalies are large ( > 500 nT).
Since most ground magnetometers have a sensi
tivi ty of about 1 nT, stations should not be locatcd
near any sizeable objects containing iron, such as
railroad tracks, wire fences, drillhole casings, or
culverts. The instrurnent operator should also not
wear iron articles, such as belt buckles, compasses,
knives, iron rings, and even steel spectacle frames.
Apart from diurna) eff ects, the reductions re
quired for magnetic data are insignicant. The vert
cal gradient vares from approximately 0.03 nT/m at
the poles to 0.01 nT /m at the magnetic equator. The
3.5.4. Ground Magnetic Surveys
(a) General. Magnetic surveying on the ground
now almost exclusively uses the portable protonpre
cession magnetometer. The main application is in
detailed surveys for minerals, but ground magnetics
are also employed in the followup of geochemicaJ
reconnaissance in basemetal search, Station spacing
is usually 15 to 60 m; occasionally it is as small as 1
m. Most ground surveys now measure the total field,
but verticalcomponent ftuxgate instrumenta are also
used. Sometimcs gradiometer measurements (3.5.S)
are made.
3.5.3. Shipborne Magnetic Surveys
Both the fluxgate and protonprecession magnetome
ters have been used in marine operations. There are
no major problems in ship installation. The sensing
element is towed sorne distance (150 to 300 m) astern
(to reduce magnetic effects of the vessel) in a water
tight housing called a fish , which usually rides about
15 m below the surface. Stabilization is similar to
that employed in the airborne bird. Use of a ship
rather than an aircraft provides no advantage and
incurs considerable cost increase unless the survey is
carried out in conjunction with other surveys, such
as gravity or seismic, The rnain application has been
in largescale oceanographic surveying related to
earth physics and petroleum search. Much of the
evidence supporting plate tectonics has come from
marine magnetics.
ing small areas may be prohibitive. The attenuation
of nearsurface features, apt to be the survey objec
tive, becorne limitations in mineral search.
(h) Advantages and disadvantages of
Airbome surveying is extremely attrac
tive for reconnaissance because of low cost per kilo
meter (see Table 1.2) and bigh speed, The speed not
only reduces the cost, but also decreases the effects
or time variations of the magnetic eld. Erratic
nearsurace Ieatures, frcquently a nuisance in ground
work, are considerably reduced. The ftight elevation
may be chosen to favor structures of certain size and
depth, Operational problems associated with irregu
lar terrain, sometimes a source of difficulty in ground
magnetics, are minimized. The data are smoother,
which may malee interpretation easer, Finally, aero
magnctics can be uscd over water and in regions
inaccessible for ground work.
The disadvantages in airbome magnetics apply
mainly to mineral exploration. The cost for survey
(g) ro magnetic data. Magnetic data
are corrected for drift, elevation, and line location
differences at line intersections in a leastsquares
manner to force tics. lnstrument drift is generally
not a major problem, especially with proton and
optically pumped magnetometers whose measure
ments are absolute values.
The valuc of the main magnetic eld of the Earth
is often subtracted from measurement values. The
Earth's eld is usualJ y taken to be that of the
(IGRF) model.
A stationary base magnetometer is often used to
determine slowly varying diurna) eff ects. Horizontal
gradiometer arrangements help in eliminating rapid
temporal variations; the gradient measurernents do
not invoJ ve diumal effects. Usually no attempt is
made to correct for the large eff ects of magnetic
storms.
(f) Aircraft The simplest method of locat
ing the aircraft at ali times, with respect to ground
location, is for the pilot to control the flight path by
using aerial photographs, while a camera takes pho
tos on strip film to determine locations later. The
photos and magnetic data are simultaneously tagged
at intervals. Over featureless terrain, radio naviga
tion (see B.6) gives aircraft position with respect to
two or more ground stations, or Doppler radar (B.5)
determines the precise flight path. Doppler radar
increasing)y is employed where high accuracy is re
quired.
over an idealized dike (prism) target. Altitude and
heading changes produced eld measurement changes
that wou)d alter interpretations based on anomaly
shape measurements, such as those of slope, Such
deviations are especially significant with highresolu
tion data.
Field operetions
3.6.1. General
Because ground surveys (until about 1968) measured
the verticalfield componcnt, whereas airborne sur
veys measurcd the total eld, both verticalcompo
ncnt and totaleld responses will be developed.
Depth detenninations are most important and lateral
extent J ess so, wbereas dip estimates are the lcast
important and quite difficult. In this regard, aero
magnetic and electromagnetic interpretation are sim
ilar. In petroleum exploration the depth to bascment
is the prime concern, whereas in mineral exploration
more detail is desirable. The potentialities of high
rcsolution and verticalgradient acromagnetics are
only now being exploited to a 1imited extent.
\sin gravity and clec~ctics. an2m1llies &&..
often matched with models. The magnetic_problem is
ioiCdi'~![~~ciH~~J fi~91~~aiaiac~C
~~~!~~ ~~}d_~d the possi~~o~~manencc. Very
simple geometncal mooels are usual]y cmpJ oyed:
isolated pote, dipole, lines o poles and dipoles, ttn
plate, dike (prism), and vertical contact. Becausc the
shape of magnetic anomalies relates to the magnetic
fieJ d, directions in the following sections are with
respect to magnctic north (the direction), magnetic
east, and so forth, the z axis is positive downward,
and we assume that locations are in the northem
hemisphcre. Wc use 1 for thc field inclination, (
3.6. MAGNETIC EFFECTS OF SIMPLE
SHAPES
cent. For the vertical contact, balf the separation
between maximum and minimum vales equals the
depth. Gradiometer measurcments are valuable in
field continuation calculations (3. 7.5).
Ground gradiometer measurements (Hood and
McClure, 1965) have recently been carried out for
gold deposits in castem Canada in an arca where tbe
overburden is only a f ew meters thick, The host
quartz was located because of its slightly negative
susceptibility using a vertical separation of 2 m and
a station spacing of 1 m. Gradiometer survcys
bave also been used in the search f or archeologcal
stes and artifacts, mapping buried stone structures,
J orges, kilos, and so forth (Clark, 1986; Wynn, 1986).
Vertical gradient acromagnetic surveys (Hood,
1965) are often carried out at 150 to 300 m aluuide.
Detailed coverage with 100 to 200 m line spacing is
occasionally obtained at 30 m ground clearance.
Two magnetometers horizontally displaced from
cach other are also used, especially with marine
measurements where they may be separated by 100
to 200 m. This arrangement permits the elimination
of rapid temporal variations so that small spatial
anomalies can be interpreted with higher confidence.
Magnetic
where J j and f2 are readings al the higher and lowcr
elevations, and is the separaton distance.
Discrimination between neighboring anomalics is
enhanced in the gradient measurements. For exam
ple, the anomalies for two isolated poles at depth h
separated by a horizontal distance h yield separate
peaks on a a F I a profile but they have to be
separated by 1.4 h to yield separate anomalies on an
profile. The eff ect of diumal variations is also
minimized, which is especially beneficial in high
magnetic latitudes. For most o the simple sbapes
discussed in Scction 3.6 (especially for thc isolated
pote, finitelength dipole, and vertical contact o
great depth extent), better depth estimates can be
made from the first verticalderivative proles than
from either the Z or F proles, For features of the
first two types, the width of the profile at
(az;az)mu./2 equals the depth within a few per
(F2 F)/4z
3.5.5. Gradiometer Surveys
The gradient of F is usually calculated from the
magnetic contour map with tbe aid of templates.
Thcre is, however, considerable merit in measuring
the vertical gradient directly in the eld, It is merely
necessary to record two readings, one abovc the
other. With instrument scnsitivity of 1 nT, an eleva
tion diff erence of o 1 m suffices. Then the vertical
gradient is given by
Z( X, y. O ) - Z( X, y. h) h( az az )z _,, (3.33)
latitude variation is rarely >6 nT jkm. Thus eleva
tion and latitude corrections are generally unneces
sary.
The inftuencc of topography on ground magnet
les, on the other hand, can be very important. Th.is is
apparent when taking measurements in stream
gorges, for example, where the rock wa11s above the
station frequeotly produce abnormal magnetic lows.
Terrain anomalies as large as 700 nT occur at steep
(45) slopes of only 10 m extent in formations con
taining 2% magnetite (k 0.025 SI unit) (Gupta and
Fitzpatrick, 1971). In such cases, a tcrrain correction
is required, but it cannot be applied merely as a
unction of topography alone because there are situ
ations (Ior example, scdimentary formations of vcry
J ow susceptibility) in which no terrain distortion is
observed.
A terrain smoothing correction may be carried
out by reducng measurements from an irregular
surface ht, y) to a horizontal plane, say O,
above it. This can be done approximately by using a
Taylor series (A.5) with two terms:
(Note that the total field anomaly F. which is on1y a
component of F,., may be smaller than Z. and that
in general F * ( Z2 + H1 )112.J
Profiles are shown in Figure 3.14a for I =45:
Zmu is located directly over the pole. The prole
is perfectly asymmetric and its positive half inter
sects the Z profile near)y at Zmax/3. The horizontal
distance between positive and negative peaks of H is
approximately 3zp/2. This profile is independent of
the traverse direction only if the effect of the pole is
much larger than the horizontal component of the
Earth's field.
A set of totalfleld profiles for various values of
is shown in Figure 3.14c. Fmu occurs south of the
monopole and north of it. is zero north of
the po1e at tan l. The curves wou)d be re
ftected in the vertical axis in southem latitudes. A
totalfield profile on a magnetic meridian becomes
progressively more asymmetric as the inclination de
creases (that is, as we move toward the magnctic
equator). At the same time, the max.imum decreases
and the mnimum increases and botb are displaced
progressively southward. The statement also applies
=Fp f1 =( cos + sin 1)
( 3.34b)
anomaly is approxirnately the component of Fp in
the F., direction. Using Equation (3.29).
Usually the fie)d of the pole, Fp, is much smaller
than the eld of the Earth, F~. and the total field
( 3 .34a) Z = (
where r1 is a unit vector from Pt y, O) toward the
pole - p. The vertical anomaly is
3.6.2. The lsolated Pole (Monopole)
Although an isolated pole is a ction, in practice it
may be used to represent a steeply dipping dipole
whose lower pole is so far away that it has a negligi
ble eff ect. The induced magnetization in a long,
slender, nearvertical body tends to be along the axis
of the body except near tbe magnetic equator. If the
length of the body is large, we have, in effect, a single
negative pole p located at (O, O,
From Equation (3.2) or Equations (3.9) and (3.11),
we get for the field at P( 0),
the dip of bodies and /3 for the strike angle relative
to magnctic north ( x axis). Note that depths are
measured with respect to the measurement elevation
(the aircraft elevation for aerornagnetic rneasure
ments).
Figure 3. 1 4. M. agnetic ettects of an pole. (a) f. Z, snd H protes tor I = . J 5 .
(a)
Surface
o
1
0.4
0.8
0.4
"l::..
..
<,
Q ..
:::::.
:>
;:
:
o
Magnetic effects of simple snepes
Fi1 ure 3. 1 4. (Continued) (b) Contours of + H.1 for H ... He 0.38. (e} F
proliles for various inclinations. (Alter S mellie, 1 967. }
(e)
o
(b)
2 3
3
z/z,
ma. N
Magnetic methods 86
Curves for these limiting cases are shown in Figure
3.15c and d.
m(2x2 z!)/r5 and Z =
(3.38b)
Near the rnagnctic equator, 1 ,. O and
=Z ... 2::! x2) 5 and = 5
( 3 .38a)
where 2pl and is the depth ot the dipole
below the surface of measurement.
Profiles are shown in Figure 3.15b Cor l 45.
The dipoJ e curves are somewhat sharper than the
monopole. The width, of the Z curves in Figure
3.15b at Zmax/2 is The same relation
holds for the F curves in Figure 3.ISb. e in the range
30 s / s: 90. For O. however, the profile is
sharper and =0.7;:m at Fmu./2. The F and Z
profiles would be symmetrical on E W traverses,
with the ftanks asymptotic to zero. The width of a
gradiometer profile peak at (1/2)( az a: : )mu. gives
a good estmate of ::,,..
When the dipole is polarized approximately vert
cally, which would be the case where l ~ 70, we
have
H . . . cos f/J sin et>
)(2 cos cos sin sin cf> )
(nr/rs){(2x2 3x.:,,.sinl}
(3.37c)
F,cosfJ - F, sin m/r3)(3cos28 1)
( nr/r3){ 3cos2 ( I + !/>) l}
(3cos2 / l)x2 sin
+ ( 3 sin2 1) .::! } ( 3 .37a)
Z = ( sin cf> + F, cos et> )
= (m/r3}(2cosflsin(/> + sin8coscf))
(2:! )sin cos 1}
(3.37b)
response may also be obtained f rom these equations,
replacing the inclination with the dip of the di
pole E.
Figure 3.lSa shows the geomctry. For a traverse
in the clip dircction. we find F. H. and Z by
resolving and F, in Equation (3.14a) along the
dipole, the vertical. and the horizontal directions,
respectively, Thus,
3.6.3. The Dipole
A small threedimensional structure containing
anomalous concentrations of magnetic materials and
varying in section from rodlike to spherical often
may be represented by a dipole model. The dpole
field was developed in Section 3.2.3. Assuming tbat a
structurc is magneized mainly by induction in the
direction of the Earth's eld, the dipole dip will be
that of the lnclination, and the magnetic response
may be obtained from Equation (3.14). lf the body's
intrinsic field due to remanence is much larger than
the cxtemal field (an unlikely case), the magnetic
~
lt may be difficult to locate xmin and x0 in high
latitudes, however, since the curves are practically
symmetrical when l >60. An alternative estimare.
good within 10%. is given by =1.4x112, where
is the halfwidth of the profile.
A first verticalderivative profile provides a good
depth estmate. The width or the prole at
(l/2)(aZ/dz)mu is equal to z P to within a few
percent.
Xmu)/{ tan + (8 + 9tan2 /)112}
(3.36b)
(3.36a)
_ 2 1/2
2( 'aun Xmu) /( 8 + 9 tarr !)
There are several relations between the profile
characteristics and the pole depth. When Z ...
Zmu/2, x112 1 (Fig, 3.14a), and when Z
Zmu/3, .tl/3 where X2 and are the
hal(widths at Zmu/2 and Zmu/3. respectively, Pole
depth may also be estimated frorn Equations (3.35).
For example,
F max.min =(psin (9 + Rcot21)112)/
{
2} 3/2
1 + ( xmu.. mm/z,} (3.35b)
where the plus sign gives Fmu. The maximum and
minimum values of F (recalling that the pole is p,
hence is positive) are
Xmu..ao 3 (9 + 8cot21)112} /4cot l
( 3 .35a)
to the southern hemisphere if we interchange maxi
mum and minimum.
The rnax.imum and mnimum values of in
Equation (3.34b) occur at
Magnetic eff ects of simple sbspes
3.6.4. TwoDimensional Features
(a) General. Clcarly the strike of a rwodimen
sional feature with respect to the carth's eld will
The direction of dipole dip is toward the side of
the profile that has thc steeper slope and negativc
tail. Tb.is tail is not pronounced, however, unless the
dip <50. The profiles are even more diagnostic
of dip. In Figure 3.15d the Z prole is asyrnmetrical,
with peak and trough above the ends of a long
dipole for 2. When z,,,// >1, the peak and
trough occur beyond the ends and depth cannot be
estimated, although a steep slope at the zero crossover
would indicate a shal]ow source.
A finite dipole sometimcs may be represented by
a dipping sheet of finitc length and depth extcnt [Eq.
(3.58)) or by an innite vertical dipping dike or finite
strike length [Eq. (3.49)).
(3.39b)
At z,,./ .2,
At l.2z,,.,
F min= m/z! }
Hmu 0.20m/z!
H- F- 0
At
Similar relations hold for the horizontal dipole:
At
Z """ O 036m/z1
1n111 aua "'
At 2z.,., Z, are minima: (3.39a)
At O, Z, F are maxima:
Zmu Fmu 2m/z!
The following relations hold for the vertical
dipole:
Figure 3. 1 5. Magnetic ettects of en isoleted dipote. (a) Ceometry. (b) Profiles far
(b)
0.8
0.4

"'1
..
' E
2
31Mag.
'
"O
o
'ii
- 3
i.i:
0.4
+p
(a)
"
/ \ ..
t. 2
\\
\
0.8
Magnetic
1 I P Surface
1 o ' , 1 "" 2'
7
- . " . - -
\ \ : ... ,e
' o
. ' ,
. . , /
. ".
!,..,, ,
Oipolc - - . . ,x __ /
.. ,
(d)
'
r
.i. ~orizontal dipole
Surface
(e)
J ~7
$Vertical dipole
o 3
Figure 3. 1 5. (Continued} (e) Profiles for vertical dipole. (d) Profiles tor horiz ontal
dipole. (e) f profiles ior different incirutions ot field end dipofe. (After S metlie. 1 967. )
mag. N
3 Mag.
Surface
2
3
Mag. N
89
20
F
16
m/z!
12
!HI
O
s
o
J
O
08
1.0
1.0
2.0
1.0
;)
"'
<,
I!

<,
o
:l2
- 3 - 2
.s
Magnetic effects of simple sbepes
Totaleld principal proles (normal to strike) are
shown in Figure 3.16 for scveral inclinations and
striie fJ - 45. Obviously the Z profiles are the
same for any strike direction. The F curves in Figure
3.16 have the same character as for the single pote,
aJ though thcy are somewhat broader. For l < 30,
the halwidth of the prole at Fmu./2 is about equal
to the depth. When is smaller, the depth is roughly
equal to half tbe horizontal distance between
and Fin1a
(3.40d) H- (2p1/r2)x'sinfJ
The northsouth component of H is from Equation
(3.40a):
(3.40c) Z
wherc r2 x'2 + The vertical componenl is
F- F,1 f1
(2p1/r2)( x'sinPI + z11k)
( cos /i + sin /k)
(2p1/r2)( cos /sin fJ + 1 sin I) (3.40b)
Now the component along F~ is
F,1 VA 2p1V{ln r}
x'I' + z,,k)
(2p1/r2)( x'sin/H + z,1k) (3.40a)
temporarily replace P(x',0) with P'(x', z). Then,
,2 (x')2 + (z,, z)2, ar/iJ x - x'/r, and
fJ r/az - lim.r_0((z,1 z)/r} - z 1 1 /r. Then,
(b) Line of poles (thin dike of infinite
An infinite horizontal line of poles
is an approximation to a long shcar or fracture zonc
or thin dik:e, whicb has appreciable susceptibility
contrast and which extends to considerable depth,
The magnetic potential is given by the logarithmic
rclation A - - 2p1 ln(l/r), wbere - p1 is the pote
intensity per unit lcngth [this equation comes from
Eq. (3.9) following the procedure in problem l,
Chapter 2, and assuming infinitesimal cross section
in Eq. (2.8)). We pass a vertical planc through
P( x, O) perpendicular to the line of polcs and take
and z' axes in this plane such tbat is horizon
tal, and the line of poles intcrsects the axis
ata depth 1 is tbe angle between the and
axes. The field F,1 is in this plane and is directed
down from P toward the linc of poles along the
vector r ( x'I' + 1 k). To get derivativcs of at
P, we write r2 in temu of the components of and
control induced magnetization, so the strike direc
tion fl will be incorporated in the expressions for
various structurcs. Traverse direction is more critica)
in identifying a 2D target. A profile 'f)proximately
along strike witl be unproductive. For 2D models, it
is customary to exhibit only (nor
mal to strike of the bodies). As in gravity, a magnetic
body generally is considcrcd to be 2D when its
strike lcngth is at least 10 to 20 times larger than
other dimensions. This situation is cven lcss likely to
be true in magnetics than it is in gravity, and fonnu
las are often modified Cor finite strike length
(Rasmussen and Pederscn, 1979).
In addition to using dircea magnetic analysis, 2D
features may also be derived witb relative case from
the corresponding gravity shapes by using Poisson's
relation (3.2.5).
Figure j, 1 6. Total- field profiles normal to a line of potes striking NW- S E.
.
.. , ,
..J .:. - p
Surfacc
Magnetic methods
04
02
O
06
F
OB 1
I
I
I
~
, '
"' ..
;., ,
.~
~~,
o L - 5 w ~- - = - ~s ; : : : : : : : : : : : r : : : : : : . ~- - 1 - ~2 _ J - ~, _ _ : o ~~~~. , - . ~~- ~l : _ : _ ~, : _ ~_ = _ s ~- = = - - - N ~E ~o ~r ~N - - W
SE ' ....._..;::.::.
02 /' ..
, _o '.
. , ,
'
figure J . 1 7. Promes normal to a horiz ontal line ot dipoles. (a} Vertic. il- field profiles.
I - 45 . (b} F profiles for fJ 45
1 Dipole 1
~
(b)

..
04
02
02
H 2m,J z!,
I
F
t - : s : w ; - - ~' t . 1 1 : & : ~~- - . . . . . ~~ . . . . . . ~~~- - ' - : ; ~: ! : : ! : : : = ~: - ; N ; E ~o : r ~N ~W
S E
(a)
,
Plaa
"
o SurflCI
1
2 1
Z profilc
fJ - 90 z
1 1
05
Magnetic effects of simple shapes
=( V( cos sin + sin!)
(U.u cos sin + sin /)l'
F M/yp)V(VUf1)
where f1 =(cos sin /W + sin /k). Using Equation
(3.27a) this becomes
(M/yp)VU f1 (3.42}
and so on. Starting with Equation (3.26), we have
2 + ( x + cot i)2
D2 + + Dcott)2
+ ( + cot E 2
D2 + ( + D cot b)
2
4'1 tan1{d/(x + dcot()}
3.6.5. Dipping Dike (Prism)
(a) General case. Magnetic anomalies caused by
intrusions, flows, or ironrich sedimentary horizons
are common Ieatures in regions favorable for min
eral exploration, and there is frequently a contrast in
the magnetic mineral content of such f eatures with
respect to the host rock. Such reatures may often be
simulated by a twodimensionaJ dipping dike (prism).
A vertical dike is also commonly used in making
basement depth determinations in oil prospecting.
Direct application of Poisson's relation is diffi
cult, so we proceed as follows. We assume a dike
with dip E and strike p, and we take the axis
along the strik:e direction. We assume that magnetic
polarization is in the F,. direction, that is, M F,..
The geometry is illustrated in Figure 3.1 Ba. from
which we havc the following relations:
directly above the center. in the case of the
dipole, the depth to tbe center of the dipole linc is
approximately equal to the width of thc prole at
Figure 3.l 7b displays totalfield principal proles
for a line of dipoles striking NW (or NE) for inclina
tions I - 90,60,30.0. When 0 s l s 15 and
45 s l s 90, the full width of Fmax/2 is roughly
the depth, whereas Cor 15 s l s 45, the depth is
approximately the distance between Fmax and Fmia
These proles are also more diagnostic of dip than
measurements are.
Magnetic
Two principal profiles for the vertical component
are sbown in Figure 3.17a, one wbere the dipole line
strikes EW (P 'fr/2) and one for a strike N30W
({J '11/6). When thc dipole line is in the magnetic
meridian, {J - O, the curve is symmetrical with Zmu
F - ( H sin /l) cos l + Z sin 1
1 4)
x { ( cos2 sirt p - sin2 !}
sin cos sin P} ( 3.4Ic)
H - iJ A/iJ x'
(2m1/r4)
X { ( 2 - 1 )cos sin fJ - 1 sin I}
(3.41b)
The field components are found from ths (noting
that H is along the axis so that the component
along the axis is H sin fJ - see Eq. (3.41c))
z
1 4)
X { ( 1 - 2 )sin I - 1 cos sin {J }
(3.41a)
1 1 )( 2yp1)(x'i' z'"1k)/r2} a1
1 2 )( x'i' z1911k)
(cos I sin/H' +sin /k)
1 2 )( cos sin 1 sin /)
The magnetic potcntial becomes
where p1 is the density per unit length and 1 is the
dipole moment per unit length in the direction u1 =
cos sin {J i' +sin /k. Then ln(r) for in
finitesimal cross section [Eq. (2.8)); we now write
r2 ... x'2 +( 2, differentiate, then set O.
This gives
A - ( m m VU a1
(e) U ne ot dipoles (rbbon). The opposite extreme
to a line of poles is a magnetic stringer of limited
depth extent, which can be modeled by a line (rib
bon) of dipoles, sometimes called a thin horizontal
cylinder, We take the axis along the strike and
derive the magnetic response along the principal
prole (in the l' direction) using Poisson's relation
(Eq. (3.26)):
Then,
4yp/ j - , dx' dz '
2
( z2 + x'l)
l x+r'c;otEfl
2(z'2 + x'l)}
.1:+1'c:ot(
z '
2yp
l z'2 csc2 ( + cot ( + x2
z ' }
z'2 csc2 ~ + 2z'( )cot ( + ( 2
We change and to and the coordinates of
a point inside the dike. r2 becomes (x'2 + z'2).
U . u =4yp j J (
Bccause U satises Laplace's equation (2.1 la), U u -
Differentiating Equation (2.9), we get
Figure 3. 1 8. Proles for dike model. L - ee except ior (d). (a) F. Z profiles for 1 60.
45. E 45 , b 2d, D - 3. 5d. (b} Z profiles for I - 75, O , E 45 E
91 . b 2d, D - oo.
(b)
0.4
(a)
D
1
0.8
Magnetic effects of simple shepes
0.4
~
N
'
e
S45E
~ o
- 4 - 2 2 4 6
0.4
Fisure 3. 1 8. (Continued) (e) F, Zproflfes for I 6Q, fl - 90, E - 45N and S , b -
D oo. (d) F,Zprofites I 75 , fl - 90, E 90, 2d, D ee, 2L oo and 1 6d.
(d)
Surfac:e
0.4
(e)
- 2 - 4
0.8
~
0.4
~
~
u
:
o
6 Mag. N
0.4
l.2
Magnetic
(e) Dike of limited /ength. Dike anomalies rarely
satisfy the criteria for twodimensionality (strike
length ~ lOb, where is width). For a more realis
tic model (sometimes called a prism model) having a
strike length 2l, D - ec, and ~ 90, equation
(3.44c) becomes
F- kF, (sin 21sin11! 1n{ ( + L2)1/2 + L}
In{ ( + L2)112 L}
+In{ (r32 + L2)112 L}
In { ( TJ 2 + L 2) 11 + L} 1
( cos2 I sin fJ - sin1 1)
X [ 1an( ~) tan( X~ b)
tan{ Ld }
'12 + L2)112
+tan{ b)(~ + L2)1/2}])
(3.49)
(d) Discussion of principal profiles. Principal pro
files are shown in Figure 3.18. In Figure 3.18a, the
Z - 2kF, sin/( 4> 1 ch + +,) {3.48a)
F - - 2kF, sin2 l{ cp1 + 4'> .)} (3.48b)
For a vertical dike with NS strike, Equations (3.47)
beco me
Z sin (sin cos ~ln( r2r3/r4r1)
sin (( +1 4>z +, + 4>4)} (3.47a)
F- 2kF,sin(sin2 I{(cos()J n(r2r3/r4r1)
sin((cf1 4> 2 +, + 4> 4)} (3.47b)
For NS strike, {J - O, so
Z - 2k.F,{ cos l ln( r2r3/r4r1)
+ sin 1( 4> 1 ~+et> .. )} (3.46a)
F- 2kF,{ sin 21 ln( r2r3/r4r1)
+cos2/(4> 1 4> 2 ~ + cf4)} (3.46b)
lf, in addition. the dike has vertical sides, t 90
and Equations (3.45) are simplied to
(b) E- W or N- 5 strike. When the dike strikes
EW, {J 90 and Equations (3.44a, e) become
sin H sin( +(}In( r2r3/r4r1)
+cos( / + ~)( 4> ~ ~ + +4)) (3.45a)
F 2kF, sin E{ cos( ( + 2T)ln{ r2r3/r .. ri)
+sin((+ 21)
X ( +1 ~ c#J i + cp4) } ( 3 .4Sb)
The parameter values in these equations may
sometimes be found Irom thc interpretation of
ground surveys, but generally this cannot be done
for airbome work. Monopolc and dipoleline ap
proximations (3.6.4b, e) may occasionally be distin
guished from dikelike models of considerable width
in mineral exploration, but usually basement is so
far removed from the aircraf t in oil reconnaissance
work that discrimination is impossible. The vertical
dike is often used for basement depth determinations
in the latter case.
= sin E sin fJ
x { (sin 1 sin E cos 1 cos E sin )In( r2r3/r4r1)
+( cos sin (sin + sin cos n
X ( 1 ~ ~ + f> .c) } ( 3 .44b)
ur; sin E [ {sin sin E sin {J
cos E( cos2 I sin2 fJ - sin2 1)}
X In( r2r3/r.cr1)
+ {sin 21 cos (sin
+sin E( cos2 sin1 p - sin21)}
x( cl> a ~ ch + 4> .c) J (3.44c)
.,
.
Substitution of the values of the derivatives in Equa
tion (3.43) gives
sin E {( cos sin E sin P
+sin 1 cos ()In( r2r3/r4r1)
+( cos cos E sin P - sin sin f)
X ( cfl1 ~ 4> J + 4> .c) }
(3.44a)
2yp sin E ( cos E In( r2 r3/r1 r4)
sin E( 4> 1 4> i - <l>J + 4> .c)}
The value of can be found the same way; the
result is
After sorne manipulation, this becomes
2yp sin {sin (In{ r2 r3/r1 ')
+cos(( 4> 1 ~ ~+et> .)}
Magnetic effecrs of simple sruoes
It is uscful to provide a qualitativc explanation
for the charactcr of the profiles in Figure 3.18 bascd
on pole distribution where thc magnctization is
mainly indueed by the Earth's field. For a dike
dipping east and striking NS as in Figure 3.19a, Z,
wilt produce N and S poles along thc footwall and
hanging wall, respectvely, as can be seen by resolv
ing z. into componcnts parallcl and nonnal to thc
dip. A similar explanation (Fig. 3.19b) accounts for
tbe more pronounced asyrnmetry in tbe prole of thc
EW di.Ice dipping south in Figure 3.18c; the N poles
on thc f ootwal) are produced by components of both
H, and In fact, such a dike tends to be magnc
tized transversely because F, is practically normal to
tbe dip axis.
Depth estimates based on width of profilcs are
not particularly usef ul unless the profilcs are sym
metrical and the width is no greater than thc depth
to the top face. Under thcsc restrictions the rule for
halfwidth at halfmaximum gives the depth to within
20%, that is, x112 at Zn .. /2. Several tech
niques based on prole slopes are effcctive for depth
detcrmination for the dke model (3.7.llb). Direc
tion of dip is usually f airly obvious from the proftles
dike strikes NESW and dips 45SE. In high mag
netic latitudes where the H component is small,
strike direction is relatively unimportant. This is
clear from Equations (3.44a, e) where, if I 90, the
expressions for Z and F are practically independent
of thc strike direction /l
Figure 3.18b disp]ays proles for NS strike.
Figure 3.18c is for E.cW strike. The asymmetry is
less pronounced for north dips than for south di,s.
In higb latitudes a dike with EW strike and dip I
gives a symmetrical profilc and a nearly symmetri
cal prole.
Figure 3.18d is for a vertical dike with NW or NE
strike; Z and F profiles for L - oo are similar
because the vertical component of predominates.
Gay (1967) takes advantage or these similarities and
obtains a single f amily or curves for F, Z, and H
over the whole range of dips and inclinations, de
fined by an index response parameter.
Figure 3.l 8d also shows the total field response
over a short dke. The shape is similar to the other
two curves but with lower amplitude. This profile is
across the eenter of tbe dke and normal to tbe
strike.
Figure 1 9. Po/e distribution in a dike.
F,
(b)
(a)
Soutb
wVZ
Eut
Magnetic methods
Weat
= 1 ) { cos sin sin(~ + 61)
+sin l cos( E+ 91)} (3.54a)
r: (2kt~/r1){sin2/ sin P sin( E +
( cos2 1 sin2 sin2 cos( ~ + 6) } ( 3 .S4b)
The profiles in Figure 3.20 for I - 60 are similar
to, althougb. sharper than, those for the dike. Rough
dip estimates are possible when the strike of the
body and totaleld direction are known. Depth
estimates from curve widths are fairly good when tbe
curves are roughty symmetrical, but not practical
wben the sheet extends lo great depth. The half
widths at Zmu:/2 for the short vertical and dipping
models give x112 l. 7d.
When the depth extent is very great, r e:: oo and
the sheet is effectively a half plane. Thcn for and
we have from Equations (3.50),
(3.53c)
= {1/r2){ sin 21 sin cos 82
+ ( cos2 sin2 {l - sin2 1 ) sin
(1/r1) {sin 2/ sin P cos 91
+( cos2 / sin2 fJ - sin2 I) sin Di} )
(3.53a)
F- sin(21 + 02)/r2 sin(2l +
{3.53b)
If the sheet is vertical, E 90 and Equations (3.50b),
(3.Slb), and (3.52b) simplify to
{3.52b)
sin2 1{ cos( E + 82)/r2
cos(t + 81)/ri}
cos(E + 82)/r2 cos(E + 61)/ri}
(3.52a)
and when the strike is NS ({l =0),
F- cos(E + 2! + 82)/r2
cos( E+ 21 + B1)/r1} {3.Slb)
( 3 .5la)
= sin( E+ + 02)/r2
sin(t + 1 + 91) /ri}
Wben the strike is EW (/1 90), Equations
(3.50) become
97
Xcos( E+ 61)}] (3.50b)
Z 2kt~((1/r2){cos/sin/hin(E + 82)
+sin 82)}
(1/r1) { cos I sin P sin( E+
+sin leos((+ 81)}] {3.50a)
F 2krF,[(1/r2){ sin2lsin/hin(E + 02)
( cos2 1 sin2 sur 1)
xcos( E+ 92)}
(1/r1){ sin21sinfl sin(E + 91)
( cos2 sin2 {l - sin2 1)
3.6.6. Dipping Sheet
Thc cxprcssions for Z and F proles over a thin
sheet may be derived from Equation (3.44) by replac
ing thc horizontal width of the dipping dike model
with 1 ese whcrc 1 is tbe thickness of thc sheet. The
principal reason for considcring tbe thin sheet is that
thc cxprcssions are simpler than f or the dike and are
sufficiently accurate provided the thickness is not
greater than the depth to the top The thin sheet
geometry is also common in mineral exploration
arcas.
For the geometry shown in Figure 3.20a, the
rcsult is
since wc know the totalfield direction. The situation
is complicated, bowevcr, because wc cannot deter
mine in advance the presence and direction of rema
nent magnetism and we bave difficulty in determin
ing the zero line for a eld profile, that is, in isolating
a single anomaly. A solution to finding the zero line
for the dike modeJ is given in Section 3.7.8.
The dike profiles in Figure 3.18b, e, and d were
simplified by assuming infinite depth extent ( D -
ee). They are not greatly changed for finite depth
extent unless the depth extent is less than ve times
the width of the top Iace, in which case the positive
tails are pulled down sligbtly. For very shallow dip,
sbort strike lengtb, and smalt deptb extent (eec
tively a ftatlying plate magnetized transversely), tbe
profiles becomes more symmetrical, with a broad
maximum of small magnitude above the plate and
ncgative tails at thc ftanks.
Magnetic effects of simple shepes
Figure 3. 20. Thin sheet; I J , D/d . 3. (a) F, Zprofiles for ( - 45 NE, P - J . (b) Z
profiles for N- S stril< e (~ - O ), ' - '4S N and ~. and effect depth e: xtent.
1
t
'\:~~
'
'
"
"
l
11
11
"
11
11
" 11
11
..
(b)
E
(o)
04
21rtZ,
1 '\
Ol
I '
venic:al '""'""'
1
'
l
\
1
1 . /Oip4S"E
02
1
\ D- flO
'
'
I
'
,
'
'
01
I
'
.....
I
'
.

,
' I
\ x/rl
o
__ ,
- 4 - 2 .6 .. ..

01
w
0.2
0.4
0.6
\
\
\
\\
\
\
\
(
I
I
I
I
I
I
I
I
//
o . . _ . . . . . . 1 . ~- J . . . - " - l ! ! : l : : ; _ _ ~p ~. 1 - - - ' - ~. . . . . . _ ~L \ - - Y - ~- ' - ~L - - " - -
6 o 4 6
Magnet;c methods
For tbis limiting case we can determine the depth F curve, we havc
and dip uniqucly. Setting dZ/dx dF/dx - O, we
x: ,,/d - 2( sin2 2/ sin2 {J
obtain the vaJ ues for the maximum and minimum
values for Z and F. This yields the f ollowing rela
2} 1/2
tions for Z:
+ ( sin2 1 - co52 1 sin2 P)
2( cos2 sin2 {J + sin2 I)
112
/{sin 21 sin fJ sin E
+ ( sin2 cos2 sin2 {J ) cos ( } (3.S6a)
/(coa 1 sin sin ( + sin I cos O (3.SSa)
x12/ d { sin2 2 / sin2 {J
X112/ cos2 sin2 fl + sin2 / }
112
2}1/l
+ ( sin2 / cos2 1 sirr {J )
, .
/( cos I sin fJ cos ( + sin I sin O (3.SSb)
/ { sin 2 / sin fJ cos (
wbcre Xm1n and x112 is the full width
+ ( sirr cos2 sin2 fJ ) sin f } {3.56b)
at balfmaximum. Likewise, writing xj12 for the Combining Equations (3.SS), we obtain the dip
Figure J . 20. (Continued) (e) Zprotes for f W 45"N snd 45 5. P - <J O.
s
Mag. N
'/'\
1
'
"
'
D
1
(e)
,
' ,,,.:.'.. __.
,_, .
.....
1
1
'
,
I
1
I
,
I
02
01
.,. . ..,. .
'
'
'
\/ DipWN.D/d 3
,
'
'
'
'
\
t " - ' . = = = = : : ' i ' ! ~: - : : ~- : : ~" ' t t " ~~- i " - - , . - - - : ~~- t - : = = = = -
' '
' '
, '
I 1
OJ
04
99 Magnetic effects of simple shapes
OS
3.6.8. Semilnflnlte Horizontal Sheet:
Fault Approximation
If r2 (or r1) approacbcs infinity, I becomes very large
and Equations (3.S9) reduce to one term for a semi
infinite sheet. Profiles for vertical component and
total field are illustrated in Figure 3.22. In Figure
3.22a, the strike is NS with 1 - 60; botb curves
Totalfield proles Cor the thin plate, illustrated in
Figure 3.21b for infinite strike length and for
4/ t are remarkably similar. (This is also true for
gravity profilcs over a thin plate.) Unless L < 1 , the
finite length does not affcct the curve more than 20!{, ,
Use of both the bottomless dike and the thin
sheet models yiclds mnimum and maximum depths,
respectively, f or Sbarpness of the peak is charac
teristic of a shallow plate compared with the dike.
F- 2ktF,[( dsin2/sin/j
- x( cos2 1 sin2 fJ - sin2 1) }
/r1{1 + r.2/L2)112
sin sin fJ - (
x ( cos2 J sir? fl - sin2 1) }
/r2(1 + /L 1)112] (3.60)
Figure 3.21 shows profiles for horizontal plates,
In Figure 3.2la, a shallow platc striking north south
produces a symmctrical Z profile. The F profile (not
shown) is the same as the Z prole but reduced in
magnitude by the multiplier sin I, lt is not possible
to make good depth estimares from the width of
these curves, but other techniqucs are available (see
3.7.8 and 3.7.11).
Over a horizontal thin sheet of finitc strike J ength
Equation (3.59b) becomes
F- 2krF.((1/r1)2{ dsin2/sin,8
cos2 / sin2 ,8 - sin2 I)}
(1/r2)2(dsin2lsinP (x /)
x( cos2 / sin2 sin2 I)}) (3.59b)
(l/r1)2( cos sin~+ x sin/)
{l/r2)2 { cos sin ,8 + ( /)sin I}J
(3.S9a)
1 , cos 81 1 , sin 82 2, and cos 82
(x /)/r2 Eliminating 91 and 82, we get
Magnetic metbods
3.6.7. Horizontal Sheet (Plate)
When the sheet is horizontal (see Fig. 3.2la), E O,
d becomes the depth of the sheet, and F.quations
(J .SO) give
2kt~((l/r1)(coslsin/hin91 +sin /cos81)
(l/r2)( cos 1 sin /hin82 + sin 1cos92)]
F - 2ktFr [ ( l/r1){sin2/ sin sin 11
( cos1 l sin2 1) 005 li}
(1/r2) {sin 21sin{J sin12
( cos2 l sirl p - sur 1) cos 92 } ]
where 112 x2 + d2, (x /)2 + d2, sin81
(3.58c)
Q - sin( ( + 92)/{ 12( + L2)1/l}
sin((+ 81)/{ r1( '12 + L2)l/1}
sin2 E 005 E/( x2 sin2 E +
2)
}
X ( cot( ( 01)/( + L2)112
cot(( B i)/( + L 2)1/l}
cos(E + 82)/{ + L2)1/l}
005{( + 91)/{ r1( r2 + L2)112
cos2 (sin V( x2 sin2 ( + 2) }
X { cot( ( 92) /( + L
2)
112
cot(( 81)/( + L2)1/l}
1be profile is reduccd in magnitude but otherwse
uncbanged in shape.
where
When the sheet is not two dimensional, wc can
modify Equations (3.50) ror a length 2L. Then the
principal profiles Cor Z and F become
+ R)oos/sin/j + (S + T)sin /}
(3.58a)
+ R)sin 21 sin P)
(S + T)(oos2 /sin2 P- sin21)} (3.58b)
(3.57b)
angle E.
tan ( .. ( 2x1/l cos 1sin13 + sin J ]
/( 2x111 sin/ cos sin /J ) (3.57a)
and the dcpth,
100
where Zm =Zmax Zmin and (dZ/dx)mu is the
maximum slope.
Thc fault approximation in Equations (3.61) is
accurate to within a f ew percent provided is larger
(3.62)
=:: X1;2 [ zm( 1 +
/{ 2x112(
z, ..(1 + xt/l/d2)}.
F or a vertical Iault, Z and profiles are shown
in Figure 3.23a. Responses from normal faults are
similar to these examples regardless of strike, but a
profile over a reverse fault reftects the concentration
of anomalous material in the area of overlap and
resembles the response of a steeply dipping target.
An estmate of the depth to the upper bed can be
made in terms of x half the distance between
zmax and zmin: 0.5 ~ s 1.25. The lower limit
applies when the lower bed is only slightly displaced;
the upper limit applies when D/d is large. When the
Iaultplane dip ~ is fairly steep, ~ == 90 and can
be estimated from
d2 + ( + d cot ~)2
= D2 + + Dcot~)2
where
+ d cot O ( cos2 T sirt sin2 /) }
( 1 / D sin 21 sin p
(x + D cot O( cos2 1 sin2 sin2 l)}]
(3.6lb)
(3.6la)
are antisymmetrical. The response in Figure 3.22b is
similar to those from targets with stcep dip. Both
EW profiles provide good deptb estimares; equals
half tbe horizontal distance between Z and z .
mu mm
(or and F mill). This measurement is 25% too
large for the NS profile (Fig. 3.22a).
1t is necessary to traverse a considerable distance
before the magnetic background is reachcd. For ex
ample, when d =30 rn, the survey should extend al
least 800 m either way from the edge of the sheet. In
practical situations this often cannot be done be
cause there are likely to be other magnetic f eatures
in the vicinity.
H we introduce another serniinfinite sheet at a
different depth, as shown in Figure 3.23, we have an
approximation to a fault. When the fault plane has
dip ~. Equations (3.59) give
z =2kt~[(1;rn{ /sin/3
+( x + d cot O sin I}
( 1/rf )( D cos l sin f3
+ ( + D cot E) sin I}]
F = ua); [ ( 1/ri2 ){ d sin 2I sin f3
(a)
Figure 321 . Thin horiz ontel plste ot finite width, 0.125. (a) Profiles fer/ 60,
O and 9J . d/! = 0.25.
1

,,.C: r1 j_
1
I I . . T
0.8
1 01 Magnetic effects of simple shspes
wbcrc Q, and are as in Equations (3.58c).
[Note tbat E in Eqs. (3.63) and in the factors Q,
and T is thc slope of thc fault plane.]
EW strike and strikc lcngth 2L, we gct
Z - 2k~ sin U Q J n( r2r7/r1r,)
+ R( 4'1 ~ ct7+ cf> a)) {3.63a)
sin E[ In( r2r,/r1r1)
+ T( 4>1 ~ 4t7 + ~) J (3.63b}
than When this assumption is not valid, it is
necessary to use the dike modeJ (3.6.S). With r3 r
oo and 4(>4 ", thc Z and F cxpressions for
the horizontal slab are identicaJ to Equations (3.44)
to (3.48) with the terms in r r lf>J , and 44 omitted
(see Fig, 3.24). Figure 3.24 shows F proles over a
single horizontal bed with NS strike and face an
gles of 45 and 135. The curves are reftections of
each other in both axes.
To simuJ ate a fault, we add a similar slab at a
difl'erent dcpth (as in Fig, 2.32). for example, for
Figure 3. 21 . (Continued) (b) Profi/es I - 75, fJ - 45 , d/l 0. 5. (e) Profiles for
I 1 5 , fl - 45, 2
(e)
.i..
~- - - TI
(b}
Ol
1 02 Magnetic
os
F
04
2k1 F.
lnfini1e \trikc lcngth
03
07
OI
o

01
1
"'
1
02
- 2 1 o 2 J
~ ~<~.
+ {sin 2 / cos ~sin fl
+sin H cos2 sin2 p - sin2 I)} ( <l> i _ ~) J ~k = and = Lo
(3.64b) (3.66)
The maximum slope occurs over the contact and is
given by (dF/dx> x- o""' 2 sin2 l/d. We can
calculate the susceptibility contrast and depth of the
bed:
=Fmu Fmin =2w~k sin2
The diff erence between the limiting value of at
xoois
(3.65)
F- 2 ~k ~ sin2 l{ rr/2 tan1(d/x)}
Figure 3.25 shows profiles over vertical and dip
ping contacts of considerable depth extent. When
ch= w/2, Equation (3.64b) becomes, for a vertical
contact striking NS
3.6.9. Contad between Beds of
Oiff erent Susceptibilities
A common magnetic structure results from the con
tac: of two slabs with contrasting susceptibility val
ues (see Fig. 3.25). Because a uniform continuous
slab creates no magnetic anomaly, the effect of the
contact is the same as that of the serniinfinite slab to
the left with susceptibility J lk . . . k' - k, From Equa
tion (3.44), we have
Z =2 Ak ~ sin ( { ( cos l sin (sin
+sin l cos O In( r2/
+ ( cos cos t sin f1
sin I sin E)( cf> 1 ch)) ( 3 .64a)
F - 2 ~k F,. [sin H sin 21 sin E sin ,B
cos H cos2 I sir? fJ - sin2 !) } In( r2/ r1)
Figure 3. 22. S emiinfinite boriz ontsl sheet. F, Z profiles for (a) N- S strike, I =60; (b}
W f - 60
(b)

Surface
t
- 0. 4
1 03 Magnetic ettects of simple
0.8
::
0.4
.O(
......
'
12 "O

:
o
8
8 12
0.4
(a)
0.8
3.6.10. Demagnetization
In the preceding examples, we assumed that the
induced magnetization is the product of k, the vol
ume susceptibility of the body, and the extemal field
F~. In fact this is true only for rodlike shapes
magnetized along the axis and having a cross section
small in comparison to their length, such as the
dipo1c of Equations (3.37). In general, the resultant
eld inside the body is diff erent f rom F.. This is
profiles. The depth is equal to the separation of
maximum and minimum values.
This calculation is not simple in practice because
long traverses are required to locate Fmu. and .mi.n
and they are usually obscured by other anomalies.
The curve in Figure 3.25a Ior EW strike is
similar to profiles ovcr dikes dipping east (Figures
3.18b and 3.20b) and it would be difficu1t to recog
nize that it represents a steeply dipping contact. In
Figure 3.2Sb. a contact with a slant f ace produces a
curve that gives some indication of the model. One
curve (broten line) also resembles a dipping sheet;
lhe other resembles a dipping dike.
A gradiometcr profile of dF / dx defines the loca
tion of a vertical contact better than either Z or F
figure 3. 23. Thin- sheet fault approximation; D - 2d, / - 60 . (a) F, Z profiles for vert
cal faults striking N- S and S E- NW O. 45). (b) F profiles for normal and reverse
faults striking E W; fault dip 45, 135.
(b)
T
l t
t
Surface
- 1 \
~- 1
el
14.Q .. /
- 4 - 6
0.4
0.6
Surface
f
l..:r~~
D
r
1
t
1
,,/
(a)
Magnetic
r
Figure 3. 25. F protiies ior contsct between two horiz ontal of ditterent
ties: Me .. k - k', O - 1 0d (a) Vertical 90). I =60. fJ and 90 . (b
S loping contsct (~ - 45") tor I - J O and 60. p =45.
(b)
0.8
1.6
(a)
24
24 16 ;

........_ ,
1.6

~ ..
0.8
<I
N
'
:S?
J :
o
0.8
Figure 3. 24. F profiles semiinfinite horiz ontal slsb striking N- 5. I - . J 5. dip of
sleb terminus 45 and 1 35". O - 3d.
0.2
12 B 4
I

I
' I
,_,
- 4 - 8
, , . ......
.... ,
...........


0.2
1 05 Magnetc effects of simple siupes
3.7.2. Crude lnterpretation and
Strudural Aspeds
Because of the erratic and complex character of
magnetic maps, interprctation is often only qualita
tive. Indeed, interpretation is something of a fine art.
An interpreter experienced in magnetics can usuaJ ly
see structure merely by looking at a magnetic map,
much as one can visualize surf ace features from the
contours of a topographic map. Frequently magnctic
features are rather directly related to surface out
crops and a magnetic map may be a f air substitute
f or a surface geology map whcre surface features are
obscured by al1uvium. Often there is a connection
between magnetism and topography, as well as with
buried gcologic structures, particularly in mineral
exploration arcas. A visual study of the magnetic
maps can be fruitful for preliminary intcrpretation.
In this regard, experience is essential. Remanent
magnetization, however, can produce significant ef
fects and lead to incorrect interpretation if over
looked.
In sedimentary regions, particularly where the
basement depth cxceeds 1,500 m, the magnctic con
5. Analyzing the proles for the locations and orien
tations of anomalies.
6. Intcrpolating profilcs normal to strike and cen
tered on anomalies, for more detailed analysis.
7. Comparing proles with curves developed from
models.
Comparing field measurements with the results
expected for simple models, such as discussed in
Section 3.6, is done to determine the location, depth,
sze, shape, attitude, and, possibly, the susceptibility
of the magnetic bodies responsible for the anomaJ ies.
Although simplified both geometrically and with re
gard to magnetization, matching curves (
with model curves provides reasonably
rapid analysis and may be sufficient, especially where
data are poor and/or incomplete.
Considerable eff ort has been expended to devel
ope workable inversin procedures for magnetics,
and, in spite of the nonunique nature of the problem,
several effective algorithms have been developed, to
which numerous references in the literature attest
(Bhattacharyya, 1964; Hartman, Teskcy, and
Friedberg, 1971; AlChalabi, 1971; McGrath and
Hood, 1973; Bamett, 1976; Teskey, 1980). Wemer
deconvolution (3.7.10) is one such algorithm. The
conventional starting point for magnetic data inver
sion might be a leastsquares fit, ridge rcgression
(Leite and Leao, 1985), and so on, as in Section
2.7.9, but magnetic inversin is more complex than
gravity invcrsion because there are more variables.
Magnetic metbods
3.7 .1. General
Magnetic survey results are displayed as a set of
profiles ora magnetic contour map. In sedimentary
arcas there may be some similarity between magnetic
and gravity maps, but in general magnetic anomalies
are more numerous, more erratic, less persistent, and
of larger magnitudc than gravity anomalies. Conse
quently, regionalresidual scparation is mucb more
eomplex, Considerable success has been achieved
with bandpass, matching, and nonlinear lter opera
ton. Downward continuation is not suitable in arcas
of complex sballow magnetics, charactcristic of min
eral exploration regions, J t migbt be used for esti
mating tbe thickness of scdiments in petroleum sur
veys, but it is not used mucb for this purpose,
lnstead, dcpths are detcrmined by semicmpiricaJ
depth rules or tecbniques like Wemcr deconvolution.
Seconddcrivative anaJ ysis is usef u) in mineral
prospecting to cnhance smallscale features near the
surface, whcreas upward continuation may be used
to suppress them. Upward continuation may also be
used reduce topographic effects in ground mag
netic work. Equation (3.33) is a crude form of up
ward continuation.
Aeromagnetic data are often treated as follows:
1. Reduction of data to a uniform grid by onc
dimensionaJ interpolation perpendicular to ftigbt
direction.
2. Preprocessing, which might involve continuation,
calculation of dervtives, extraction of the verti
cal component, and so forth.
3. Summation of severa] profiles to attenuate back
ground noise.
4. Filtcring and setting a thresbold to locatc anoma
lous arcas.
3.7. PROCESSING ANO INTERPRET ATION
Maximum demagnetization occurs in thin sheets
magnetized normal to the lace; in this case, 41r.
For the sphcre N . . 4w /3. The effect is quite small
unless k ~ 0.1 SI units. Demagnetization is signi
cant only in massive pyrrbotite and in rocks contain
ing >5101> magnetite. Sometimes an additional
ractor of (1 cos2 D cos2 AJ ) is includcd to allow
for the resultant magnetization being in a diff erent
dircction from I; is the difference in inclination
and D is the declination.
O< < 4tr (3.67) k k/(l + Nk)
usuaJ ly called The effect can be
accommodated by replacing the susceptibility k by
an 0:
y', - h)
j y.O)
2,,. {(x _ x')2 + + h2}1;2
(3.70)
Field continuation was discussed in Section 2.6.7 as
it applies to gravity. Adapting Equation (2.48) to
magnetics, we have. for upward continuation (whcre
is positive downward).
3.7.5. Continuation
3.7.4. Derivatives
Derivatives tend to sharpen the edges of anomalics
and enhance shallow Ieatures, First and second vert
cal derivative maps are the most common ones made,
Dcrivative maps may be made by the same tech
niques used for gravity data (2.6.S). The first verti
cal derivative is also measured in gradiometer sur
veys.
The relations betwecn sources and their potentiaJ
elds may be considered convolution operations, and
transforms can be uscd to determine source charac
teristics rom field operations, as discussed in the
following sections.
(3.69b)
where w( a, /1) is the convolution operator (also
called a or In the wave
number domain, this becomes simple multiplication:
J
(3.69a)
(x, y) are spatial coordinates and (u, u) are wave
number coordinates.
The important characteristic of transformations is
that information is not lost in the process, and in
many cases operations are easier to perform in the
transfonn domain. For example, the outputdnput.
expression g(x, /(x, y) w(x, y) for convolu
tion in the spatial domain is (A.10)
(1/2tr)2j
(3.68a)
J (3.68b)
We usually write relations using twodimeosional
Fourier transforms /(x, y) ++ (Eq. (A.57))
in the form
1 07
3.7.3. Data Processing Operations: The
Fourier Transform
As with gravity and seismc data processng, mathe
matical operations, such as convolution and correla
tion, can aecomplish filtering, residualizing, continu
ation, and so on. Operations can be perf ormed in the
spatial, or wavenumber, domain (often called the
frequency domain because wavenumber is spatial
frequency). Fourier transfonns (A.9) are particu
luly useful in magnetics for (i) resoluton of specic
anomalies by downward or upward continuation, (ii)
changing the eft'ective field inclination (reduction to
the pote) or conversion of totalfield data to
verticalcomponcnt data, (iii) calculation of deriva
tives, (iv) general filtering separating anomalies
caused by sources of different size and deptb, and (v)
modeling (Bbattacharyya and Navolio 1976). For
literature on transformations of potential eld data,
see Dean (1958), Bhattacharyya (1965, 1966), Gunn
(1975), and Spector and Grant (1985).
tours are normally smooth and variations are small,
reftecting the basement rocks rather than nearsurf ace
f eatures. The larger anomalies usuaJ ly are caused by
susceptibility variations rather than basement relief,
Consequently, anomaly magnitudc is not of much
value in finding basement depth, and depth calcula
tions are usually based upon anomaJ y shapc mea
surements, especially sharpness.
Regions where igneous and metamorphic rocks
predominate, like thc Precambrian Canadian Shield
and thc AppaJ achians, usually exhibir complex mag
netic variations. Deep features are frequently camou
ftaged by higber frequency magnctic effects originat
ing nearer the surface. Techniqucs for separating
deep and shallow anomalies are similar to those
discussed for gravity features in Scction 2.6.
Magnetic anomalics often lie in trcnds. From a
study of aeromagnetic maps of primarily sedimen
tary arcas in western and central North America and
Venezuela, Atneck (1963) found that the dominant
direction within single magnetictectonic provinces
is usually NESW or NWSE and the trend nor
mally tenninates at the provincc boundaries. Moder
ate to wealc features trending EW or NS oftcn are
superimposed on these. These weaker trends fre
quently extend across province boundaries and are
probably of more recent origin. A cursory study of
the Canadian Appalachian region (the Maritirne
provinces and eastem Quebec south of thc St.
Lawrence) and northem Saskatchewan and Alberta
appears to confino the trends. Largescale northeast
trends are obvious on the east and west ftanks of the
Canadian Shield.
.
,.
Processing ;nterpretation
3.7.6. Spectral Analysis
The Fourier transform expresses a magnetic eld as
an integral of sine and/ or cosine waves, each defin
ing a wave of amplitude (x) and phase 4> ( ic ), whcre
ic/2" l/A is the wavenumber. PJ otting A(ic) gives
the amplitude spectrum and A2(") gives the power
spectrum.
The expression for the feld of an anomalous
body of ten can be written as the product of three
functions in the wavenumber domain (their convolu
tion in the spatial domain):
(i) / amplitude factor, which is equal to (4trM)2,
whcre M is the magnetic momentjunit volume.
(i) / the depth factor, exp{2h(u2 + u2)112},
wherc h is the pole dcpth.
(iii) / tbe fieldorientation factor, { / cos fJ +
sin /j)2 + n1 (/u + mll)2 /(u2 + v2) + n2,
where /j is the angle bctween thc body's orien
tation and magnetic north and /, and are
the direction cosincs of tbe field F,.
Thus the calculation of , , . (u, v) is straightforward.
Hanson and Miyazk.i (1984) use a continuation
method that is etrective where surface rclief is large
and rocks are highly magnetic. As in gravity, upward
continuation smooths the data, whereas downward
continuation emphasizes higb frequencies.
, 11(u,v} Fo(u,v)(h/2w)
Xexp{ - 2fl'h(u2 + v2)112} (3.7S)
For upward continuation, we know Fo(x, y, O)
and the unknown is F,,(x, h), wbose trans
fonn is
(3.74)
X dx dy
( h/2rr)e211(u2+u2)'12
[compare witb Eq. (3.33)). For a first approximation,
the ftrst two terms involving F and a F /a are often
sufficient, and the simultaneous measurement of to
tal field and vertical gradient provides these values.
The < J F/dz term can be found from maps of F, as
indicated in Scction 2.6.S, and, if necessary, tbe
a
3
/a : term can be found in a similar manner
from maps of oF/az .
The Fourier transform provides another tech
Dique for fic)d continuation. The integrand in Equa
ton (3.70) is the product of 0) and
(h/2){(x x')2 + (y y')2 + h2 }3/2. Using
the symmetry theorem [Eq. (A.60)) and the convolu
tion thcorem [Eq. (A.67a)], we get
X( t) Y (t).,. 21tx( <&J ) w)
2wx( w)
wberc K(r1, h) are wcighting coefcients (listed in
Table 3.2 for h - 1 to S). These coefficients give the
upward contnued field wthin 2%.
Another solution for continuation (in either direc
tion) is by means of a Maclaurin expansin [Eq,
(A.40)):
y, h) y,O) + aF( x,
+( h2/2!) F(x, y,0)/J z2
+( /3!) y,O)/az3 +
(3.72)
h) 1 , - h) (3.71)
The lcft sidc is the total eld at the point
h) above thc surface on which Ft y,O)
is known. The calculanon procedure is to replace the
integral wth a weighted sum o vales taken on a
regular grid.
The empirical formula of Henderson (1960) gives
the eld at the elevation h above the surface in terms
of values F( 1 ), the average value F( r) over a circle
of radios r1 centered at the point (x, y,O):
1 08 Magnetic
Table 3.2. Coefficients for upward
Thus, applying this result and Equation (A.62}, we
obtain for the transfonn of Equation (3. 70),
'; K (~. 1) K (';. 2) K(r;.3) K(~. 4) K(r;. 5)
o o 0.11193 0.04034 0.01961 0.01141 0.00742
,,;(u,v) (2w/uv)Fo(u,v)$"(u,v)
1 1 0.32193 0.12988 O.Ob592 0.03g()6 0.02566
(3.73)
2 006062 0.07588 0.05260 0.03566 0.02509
3 . s 0.15206 0.14559 0.105&3 0.07450 0.05377
where ~11(u, v) . . . . . Ft x', h), Fo(u, v) ....
4 . s 0.05335 0.07651 0.07146 0.05841 0.04611
5 /13 0.06586 0.()9CX)2 0.10226 0.09173 0.07784
Ft x, y,O), and 1 f""(u, v) . . . . w(x, y) . . (h/211)(x2 +
6 s 0.06650 0.11100 0.12921 0.12915 0.11986
yl + hl)l/2.
v'SO 0.05635 0.10351 0.13635 0.15474 0.16159
The continuation lter is
8 /136 0.03855 0.07379 0.10322 0.12565 0.1410b
9 ./274 0.02273 0.04464 0.0bSOO 0.08323 0.09897
10 2 5 0.03015 0.05998 0.08917 0.11744 0.14458
#'"(u, v) (h/2rr) J J ( x2 + + h2)3/2
In [ {<X1 B)
2
+ i ] / {( X1 +
2
+ 1} ]
+ In [{ ( X B)
2
+ i]
I { ( + B)
2
+ 1 } ) ""' O
{ ( X1 B)2 + 1} { ( X2 8) + l}
.. { ( X1 + B)
2
+ i] { ( X2 + B)
2
+ 1}
For the antisymmetric component,
we have
Then, because
(3.78)
The first term is the symmetric componcnt and the
second term is the antisymmetric component A.
Over the center of the dike, has a mximum and
O.
U wc chose two conjugare points X1 and on
the dike profilc such that the sum of the Z valucs is
equal to Z,. the valu of Z at X - O, that is,
N /2 2 k~ sin H ( cos J sin sin ~+ sin / cos n }
and
M- 2kF.,sin( {(cos/sinPcos~ sin /sin~)}
whcrc
Z M{ cot1( B) cot1( B)}
+ ( /2) In [ { ( B)
2
+ i]
/ { ( X + B)
2
+ l}) ( 3. 77)
wherc d + (x + b)2, d + b)2,
4'1 cot1(x + b)/d, and cl> J cot1(x b)/d.
Dividing by d to obtain the dimensionless parame
ters X - x/d and B - b/d, wc obtain
2 sin H ( cos sin sin P
+sin cos Oln( r3/r1)
+ ( cos l cos ~sin fJ
sin l sin ( )( lf)1 ~) }
point above the midpoint of the top of the dike. Ir
we assume infinite depth extent so that r2 oo,
~ = cf> 41, we have
3.7.8. Use of Master Curves for Dikes of
Great Depth Extent
lt is oten difficult to establish a background or
datum level for magnetic measurements and to lo
cate a dike with respect to the profile. Hence, match
ing field results with profiles obtained from simple
shapes can be difficult. An analysis (Koulornzine,
Lamontagne, and N adeau, 1970) for prisms and
dikes of infinite depth extent solves this problem
witb master (or characteristic) curves tbat give depth,
dip, and width of a prism or dike.
We rewrite Equation (3.44a) ma.king the width 2b
instead of (Fig, 3.18a) and shifting lhe origin to a
This operation changcs the actual inclination to the
vertical. lt can be performed (Baranov, 1957;
Spector and Grant, 1985) by convolving thc mag
netic field with a filler whose wavenumber response
is the product of a polarizationorientation factor
and the fieldorientation factor in items (v) and (ili)
of Section 3.7.6. Tbis transformation simplifies
'
tetaleld maps and is a relatively easy operation at
high magnetic latitudes where Z =F, but becomes
more difficult near the magnetic equator (Silva, 1986).
3.7.7. Reduction to the Pole
The eff ect of any of these factors may be removed by
deconvolving in the space domain or by dividing in
the wavenumber domain.
Numcrous other filter operations can be used to
separate deep from shallow, large from small, and
threedimensional from twodimensional eff ects.
Gunn (1975) discusses a solution for the magnetized
rectangular prism. Hahn, Kind, and Mishra (1976)
estmate depth from Fourier spectra.
1 1
(3.76b)
or, in the wavenumber domain,
F( O) /(X, Y } /{J ( Y }
(3.76a)
For bodes that are large comparcd to their depth,
we require two additional factors:
(iv) The size factor: for a rectangular prism,
(sin ua sin vb)/(ua vb) where 2a and 2b are
the prism dimensions,
(v) The polarizationorientation factor involving the
direction cosines of the polarization vector.
We thus write the field of a point pole or dipole in
terms of factors (i) to (ili) as
where + 1 D- (t/t11)/4,
P) - (1 11)2/4., w - (4 (+2 1)1)112/2,
.,._ (1/2){42 (1 #')4)112, X112 X1 1 d,
and so forth. Finally, we can find thc clip anglc (
(3.83)
S( X) (1/2){ Z( X) + Z(X)}} ( )
and A( X) (1/2) { Z( X) Z( X)}
382
wc can plot S( X) and X) by taking points that
are cquidistant eithcr side of X - O. On these pro
files we mark points ~14, S112, and A112 with corre
sponding abscissae X X112, and X, also X,
(Fig. 3.26b). By a development similar to that used
Cor Equation (3.81), it can be shown that
xv2( 4'2 1)/2 1 2
2b - .x112{ 4 ( 1/)2 1)2}
112
2x112W
x,( 1 I' )1 /2p.
lb .x,{ 4.2 (1 11)4}
112/p.
= 2x,Y
follows from F..quation (3.78) if we pul
and Z2 Zm1n, that is, Zinm Zawi Z.,.
The analysis may now be carried further to estab
lish the dike paramcters. First, from the deftnitions
of symmctric and antisymmetric functions,
We can now J ocate thc point X - O and get Zo For
the best accuracy, X3 and X4 sbould be located closc
to the midpoint of the anomaly, and X1 and X2 near
the maximum and mnimum. Z(O) is thc point on
the profile located a horizontal distance X3 from Z,,
X1 from ~, and so forth. Thus thc datum linc
Z - O can be drawn at a dstanee abovc Zmia equal
to the vertical distancc betwccn z_ and z, . 'This
(3.81)
lnitially we do not know eithcr the location of the
dike eenter, X - O, or the datum level Z - O.
Two pairs of conjugate points, X1, X and X X
are cbosen such that Equation (3.78) is satised, and,
in addition, z_ Z2 Z1 Zm.a and ZDIU
Z4 Z3 Zm E (Fig. 3.26a). Writing I X3
- X X X X1 X3, and using thc
relation X1X2 X3X4 [Eq. (3.74)), we find that
X1 + /) /( I - + n)
X1 - m(/ m)/(1 + n)
X + n)
+ + n)
Solving for X1 X2, this gives
Figure J . 26. T~dimensional dilce of infinite depth extent. (a) Locstion of conjugate
poims end X - O .
(cr)
Zprolle
, ,
+
t
z
1 1 0
Figure 3. 26. (Continued (b) S ymmetric components. (e) 1 1 . 1 . ister
curves.
"
.,
]O
1 - 01 - 7 - . . : : : : : : - ~" " 7 ' ~~- - - - - - - -
!/
..... .,,.
" "
20
Cb)
o
OZ
06
An1isymmc1ric componen
A(X)
\
\
\
\
,.....1..~..t::;.._
\ z
\
' / ........
02
Ori1inal
profilc
1 1 1
Processing
where F( x) is given by Equation (3.86a) so that
+ 5) unknown quantities are involved. UsuaJ ly
the polynomiaJ is first or second order only, so that
six or seven stations are sufficient for a solution.
The scope of this analysis has been enlarged to
include models other than dikes: basement topogra
phy, magnetic interfaces (which use dF /dx rather
!F + 0 + 1 + {3.87)
where a0 - - Mx0 + a1 M, bii - x~ - z2,
and b1 we find that =h1/2 and
( 4f> o 112 /2. Thus we can determine 0 and
by measuring at four stations and solving Equa
tion (3.86b) for a1, f> o, and b1.
Extending the problem beyond an isolated
anomaly. Wemer assumed that the noise or nterer
ence caused by neighboring magnetic anomalies
could be taken into account by extending the poly
nomial, so that the measured field !F becomes
x2F(x) 0 + 1 + 0 + 1
(3.86b)
where is the surface point directly above the
center of the top of the dike, is the depth to the
top, is the point of measurement, and the axis is
normal to the strike. and are unknown func
tions of the dike geometry and mineralization. Rear
ranging Equation (3.86a) in the form
F(x) { x0) + Nz}/{(x x0)2 + z2}
(3.86a)
3.7.10. Wemer Deconvolution
Wemer (1953) proposed a method for isolating a
magnctic anomaly from the interf erence produced by
nearby anomalies. This led to automated procedurcs
for interpreting magnetic data, now known as
(Hartrnan, Teskey, and Friedberg,
1971; J ain, 1976; Kilty, 1983).
The rnagnctic anomaly for a dipping dike can be
written in empirical formas
spectrum of the soughtfor signa] is known. A
is a filter that has the same spectrum
as the soughtfor signal, One way to carry out
matched ltering involves using the Hilbert trans
form (Sheriff and Geldart, 1983: 10.3.11) to sepa
rate the symmetric and antisymmetric anomaJ y com
ponents (3.7.8). Both totalfield and gradient data
may be processed in this way (Naudy, 1971;
Nabighian, 1984).
Magnetic methods
3.7.9. Matched Filtering
Wherc the problem is locating a signal in a data set,
matched filtering provides a powerful method if tbe
gi. In( (1 'IV)/{l + ~)} (3.85b)
Becausc Equation (3.83) gives W, D, 'IV, and !1 J in
terms of ~ and "' and these in tum can be found
from the curves, and 9' are fully detcrmined.
We can solve for the totaleld anomaly in simi
lar f ashion. The analysis for thc dike of infinite
depth extent has been extended to cover a prism of
finite length and depth extent. Because of the extra
terms, it is necessary to provide more master curves,
but the proccdurc is similar. Master curves for other
models are also availabele (Grant and Martin, 1966;
Martin, 1966; Gay, 1967). The second reference con
tains tbe following models, in addition to the prism:
horizontal slab, plate, rod, and dipping sheet. (The
method can be extended to dikes or prisms in gravity
interpretation.)
Six master curves involving functions of d, b, and
~ for the symmctric and antisymmctric components
of thc dike prole are shown in Figure 3.26c. The
dipanglc functions and gi in Figure 3.26 are
rclatcd to known quantities in the following expan
sions of Equation (3.84):
( ( " cot1( H, sin P/Z.,)}
tan1( or tan1( lgJ ), (3.85a)
where A(X)mu:/S(X)aiu,
4tan1(
In{ (1 W)/(1 + W)}
4 tan1( ~? 2))
or
( " cot "! ( H, ;,
1{ A(X)mu
+ tan ( )
S X mu
4 tan1
'1n{(x112 + B)} (J .S4b)
from the relation
( " coi1(
+ tan1{ A{X)1t1u
S(X)max
4 tan1
In{ (x., + B)} (J .S4a)
1 1 2
The use of both rnethods provides a check on the
(3.89a)
(3.89b)
h =k1 S 1.67 s k1 s 2.0
(generally k1 1.82)
h . . k P (generally k ""0.63)
(b} Empiricet depth rules. A number of ruleso
thumb for depth estimation have developed from
practica! experience in magnetic interpretation. These
relate to profiJ e shapes; for example, thcy often use
horizontal widths at sorne fraction of the peak value
for symmetrical curves and horizontal distances from
peaktozero values for asymmetric curves. Peters
(1949) was probably the first to relate depth to the
horizontal extent of portions of sloping flanks, and
variations of slope techniques are among the most
popular. The vertical gradient is also used in such
rules (Barongo, 1985). A summary of such rules is
given in Table 3.3.
S)ope rnethods are wideJ y used, especially Ior
aeromagnetlc interpretation. Graphical techniques
use the sloping flanks of proles to estimate depth
(Nettleton, 1971; Spector. 1979). In Figure 3.27a,
is the horizontal extent of the portion of the curve
that is nearly linear at the maximum slope. Two
additional line segments have been drawn tangent to
the profile at half the max.imum slope: the distance
between the ponts of tangency is P. The depth of
tbe source beneath these portions of the curve is
given by
Where the body is uniformly magnetized by induc
tion, we may rep)ace Mmax by kF, or k~/(l + Nk)
as in Equation (3.67).
Because we do not normally have a vale of
estimates obtained by combining the two lim
its are even cruder than the equivalent relations Cor
gravity. For a semiinnite thin sheet, thc rcsult is
within 50%, but it appears to be even poorer for
threedimensional featurcs.
and
For Z proles, where M is everywbere vertical and
in the same direction (down or up), the numerical
factors are reduced to 2.6 and 3.1. respectively.
For twodimensional magnetic fe atures having in
nite length in the . v direction, in which the total
magnetization is parallel throughout, the equivalent
expressions become
;f.
,
3.7.11. Depth Estimates
(a) S mith rules far meximutn depth. As in Section
2.7.12, which dealt with depth estimates for gravity
anomalies, there are corresponding limiting vales in
magnetics derived by Smith (1961). lf the magnetiza
tion is parallel throughout a body, though not
necessari)y uniform or even in the same sense, and if
IMlmu 1az;axlmuand 1az2;ax21mu are abso
lute values of the mxima of M and the rst and
second derivatives of F or Z along the profile,
then the depth z,. to the upper surface is given by
than F), faults, and contacts. This type of anaJ ysis is
also suitab]e Cor gravity interpretation.
There are lmitations to Wemer deconvolution,
such as resolution between neighboring bodies and
J ack of discrirnination among parameters, leading to
a re)ation between, say, dip angle and susceptibility.
The data are sensitive to geological and measure
ment noise (signa]jnoise 100 produces 20% scat
ter in depth and position estimates). However, the
technique is attractive because of ease of access to
the computer and consequent speed in handling large
quantities of data.
Note: is the full width at halfpeak amplitude, is the horizontal distance
from peak to zerocrossing. is an emprcally determined index factor, and 1 is the
horizontal distance over which slope is straight line.
Method Halfwidth PeaktoZero Vertical grad. Flank slope
Component F. aFaz . F. Z
measured F. Z F,Z,H z . az ;az
Model
Monopole 1.3x12 - - 2F/(iJ F/iJ z )
Dipole 13/xpo s 2 - 3F/( aF/iJ z )
Monopole line
"112
- F/( aF/8z )
Oipole line
2x1/2 Xp0 - 2F/(iJ F/iJ z )
General 0.7 s s 1.3
- nF/(aFaz 0.5 s 1.5
Table 3.3. deprh
1 1 3
3.8.1. Ground Surveys
(1) The first example shows tbe inhcrent complexity
of ground magnetic data and the difticultics in accu.
rately interpreting them. Figure 3.28 displays mag
netie contours and two vertical component profiles
3.8. F IEL D EXAMPL ES
should be multiplied by a cosine factor. Correction
also has to be made for tbc ftight elevaton to give
values with rcspect to sea lcvcl (or to an arbitrary
datum).
depth eseimates and tbe care with whicb thc graphi
cal analysis is done. Th.is method generally yields
rcasonable resulu for horizontal bascmcnt modcls
with steeply dipping contacts; thus, it is suitablc in
tbe analysis of airbome data. It is much simplcr and
faster and provides more dcpth cstimatcs than analy
sis by modcl curve fitting. It can be carried out on
original field profilcs and so need not wait on map
preparation; it can also be applied to analysis of
maps (Rao and Babu, 1984).
Use of slope techniques requires corrections.
When ftight lines are not normal to tbc local geologi
cal strike, horiz.ontal distances are too large and
Fisure 3. 27. Determinins anomaly depth from tbe of a magnetic profile. (a)
Maximum- slope (5) and half- slope (P) measurements. (b) Maximum slope
ments on a thin plete anomaly.
PI ate
(b)
Fliaht line
(a)
I
I
- ,, I
" .. ,'
, . . " I I 1 1 ; ' 1 ,. . _,. p H
: /'; : : I
I
I
I
' ,
'o I , . . ,
I I ; '
~,' ~/
..,, /
' "~
'
1 1 4 Magnetic methods
+ Q
"'.
+
e
1 E
: ~ g
1 E "
i:, c:i"..... ...
"I c:i
E g a ..;
. . : '- =
.. 1 1
. e'G. N N
....... :i:, ~
l&I
, i .:
1
, ' - +
/' 86 ...
,.,.#." + ::~ !.
" . . . .. ,,., ..
.!! g
& : =
~.!I
l i
~]
~
70 m, 90 m, k1 1.3 SI, and k2 1.9
SI. This results in a reasonablc match of the central
trough and the northeast ftank, but the southwest
ank is much too largc. A better t (sbown in Fig.
3.28b) was obtained with the two vertical sheets
illustrated, but the trough between thcm is too deep,
Also, the depth extent must be less than 120 m
because the bodies wcre not encountered in boles
T1 and T3. Although this interprctation is certainly
not denitive, it is clear that the magnetic sourees
are shallow, bave limited strike J ength, steep dip, and
large susceptibility contras t. This last f act indica tes
high magnetitc contcnt and possibly large rema
nence, which may be responsible for the dsagree
ments (Green, 1960).
(2) The rnagnetic methocl is particularly useful in
cxploring for asbestos because of its occurrcncc
in ultrabasic intrusive rocks rich in magnetite.
When olivine (Mg2Si04) is aJ tered to serpentine
(Mg3SiO~(OH)4) and magnesite (MgC03) by the
addition of water and carbon dioxide, the asbestos is
associated with high magnetic susceptibility and
massive serpentinite. Figure 3.29 shows a vertical
component protile over an asbestos prospect near
Matheson in northem Ontario and the gcologic scc
tion under a 15 m overburden. High magnetic re
sponses corrcspond to the asbestos and massive ser
pentine zones with lows over the volcanics and highly
carbonized serpcntite. A rcasonable match to the
ficld profile was obtained by assuming dik.cs of con
siderable depth extent using Equation (3.44a) with
r2 as r4, ~ ct> 4, fJ - w /2, and ali con tacts vertical
except the left one, which dips 30. The presencc of
asbestos in tbe massive serpentine zoncs can only be
established by drilling.
normal to the strike of pyrite mineralization. There
are two parallcl pyrite zones in acidic flows, near a
contact between the lattcr and rhyolite porphyry.
B oth havc a strike length greater than 300 m and the
zone ncarer the contact appears to pinch out on
line 75.
Although the pyritc mineralization is c1ear1y asso
ciated with a magnetic trcnd in the area, the large
magnctic anomalies on lines 73 and 75 could only be
due to magnetite or possibly pyrrhotite, since the
susceptibility of pyrite is relatively low (Table 3.1).
However, there is no specific indication of thcse
minerals in thc drill logs of boles 1 to 4.
Because the overburden near the diamond drill
holes was generally quite thick (25 m at T1, for
example), it was originally assumed to be at least 1
m throughout the grid. Howcvcr, a sballow seismic
refraction survey carried out later on line 75 showed
bedrock only 1.5 to 3 m below the surface in the
vicinity of the pyrite zones, dropping off abruptly to
15 to m northeast of the acidic ftowrhyolite
contact. Thus thc magnetite sources may be very
close to the surfacc and or small depth extent.
The sourcc for the single 13 T peak on line 75
appears to be a finite steeply dipping sheet at very
shaUow depth. Using Equation (3.58a) with p - 45,
1 - 70, 90, and Z, 36 J tT, and fitting tbe
profiles at three points (ineluding the maximum), we
obtain a reasonable t with d t= 8 m, D 25 m,
2l 30 m, and 3 SI unit (See Fig. 3.28b).
However, whcn we try to match the double peak
prole on line 73 by assuming two vertical sbeets of
identicaJ cross section separatcd by SO m and induc
tively magnetized in the earth's fic]d, Equation (3.58a)
produces the following parameters: 2.5 m,
Fi1 ure 3. 29. Vertical componen! ground megnetk: proiite in an ere ot asbestos miner-
aliz ation near Matheson. O ntsrio.
Owerbvrden
0.39
1
F ield profile .. __ ,
Composi1e theoretieal
pro file
~VolcaniCJ
E Asllestos libre zone
~ Serpentmite. H1ghly .
~ carbonatized
~Serpenlinite. Moderately
D
carbona1ittd
Massiwe serpenline
Magnetic methods
5000
1 1 6
also see two welldefined magnetic highs that are not
topographic features: one about 5 km west of Mt. St.
Gregoire and a larger one 11 km northwest of Mt.
Bruno. One assumes that they are igneous plugs that
failed to reach the eminence of the Monteregian
bilis.
These two fea tu res and Mt. St. Gregoire provide
excellent examples of the verticalprism model com
monly ernployed in aeromagnetic interpretation. Fig
ure 3.30a shows the totalfield contours, whereas
profiles are displayed in Figure 3.30b. Two mcthods
(l) The Monteregian hills of the St. Lawrence low
land regin near Montreal were formed by igneous
intrusions into scdirnentary rocks. These hills are
magnetic as weU as topographic anomalies because
of their contrast with the low susceptibility sedi
ments. Aeromagnetic maps (Canadian Govemment
Aeromagnetic Series, St. J ean and Beloeil) show this
clearly for Mt. Bruno, Mt. St. Hilaire, Mt. Rouge
mont, and Mt. St. Gregoire. On the sarne sheets we
3.8.2. Airborne Surveys
figure Mdgnetic tor reas in towlends.
(a) Maps, 700 nT
(ii) lleloril Aeromaanetic Sheet 1674Ci
~?' O 3 km
'l> .tf" 1 1 1 1
~i . . . - ~_ _ ; : o . . . . : ; : - - ~~~~~~~: : i o - . c. . _ . ' - - ~~~~~~~ miks
P- - P Principal prole
T.N.
1
2 km
o
1 1 7
(i) St J ean Aeroma1nctic Sheet l 678G
Field exemptes
thc curves have steeper slopcs on the south or soutb
east ftanks and it is necessary to inerease the lateral
dimensions to match the field profilcs. In practical
interpretations, the dcpth to thc top of the prism is
the most significant dimension. Because ftight eleva
tion was 300 m, Table 3.4 puts the St. Gregore plug
about 150190 m above ground (Mt. St. Grcgoirc
rises to a height of 180 m abovc ground). The top of
were employed to assess the magnetic cbaracteristics.
One used tbe models of Vacquicr et al. (1951) and
the other used Equation (3.44c) for a vertical prism.
1bis allowed calculation of the susceptibility con
trast the depth z, strike J engtb and width
by matching thc principal profiles. The results, which
ftt rcasonably weD, are shown in Table 3.4 and in
Figure 3.30b. If we assume l 75 instead o 60,
I k
Ano mal y (deg) (SI) (m) (m) ( m) Source
St. Gregoire 60 0.04 110 880 880 Vacquier (F ig A60)
75 o.os 110 880 670 Vacquier (fig. A70)
60 0.06 150 79:) 79:> Equation (3.44c)
Anomaly near &O 0.025 230 1, 700 1, 700 Vacquier (F ig. AO)
St. Gregoire 75 0.03 230 1,700 1,250 Vacquier (fig. A70)
60 0.04 300 1, 220 1, 220 Equation (3.44c)
Anomaly near 75 0.08 1,130 3.660 9,140 Vacqier (F ig. A75)
B runo &O 0.055 1, 220 4.880 12.200 Equation (3.44cl
Table 3.4. lnterpretation of anomalies in the S t. lawrence lowlend
Figure J . 30. (Continued) (b) Principal profiles. I 60 snd F,. 60 T.
(b)
IS 20 km 10
NW
Principal profile NWSE
Vertical pr1m
~ 135", 1200 m
0.06 6 4900 m
21. 12000 m
Prilm 1CC1lon
km
N
Principal profile NS
Vertical prism
~ 90", 3(l0 m
lt0.04, h21.12oom
4 km
N
Principal profile approx. Ns
Vcrtieal pri1m
~90, 1 ISOm
0.06,b 21. 790 m
2500
Anomaly norihwctt
ofMt lruno
2
s
2D
Mt St Grqoirc
...
..5
s
2
2700
Pri1m r.ectlon
Anornaly west
: ;
o St Ore1oir1
e
._,

1 1 8
Figure 3. 31 . High reso/ution eeromsgnetic survey. Timmins eres, O nterio. (From Bhat
tacharyya, 1 971 . } (a) Geotogicsl map. (b) Cround vetticsi- intensitv map.
(b)

E) :. ::::::~.!!~ .. _ .
. ....,.,,,, .
. .
E] :. =!~ .. ~!."".
" '"' . .. ..
.. ..
a=!.~"' .... ".
E] =~~~
~, , , ......... _ ., .
D :. =::'..
..,.a&ICS
~ ; =~=: . . . .
: a. .,.,,,
, ., ..
"' " ~ .
--r l
l e
,.. l 'f l l J 'TY
" .. ,...,, ..
T ........ I
..... ., ............
...............
.......
1 1 9
la)
Field examples
Figure 3. 31 . (Continued) (e) Convemionet eeromsgnetic map. (d High- resolution eero-
m. J gnetic nup.
(d)
100
., .
,
Magnetic methods
1 20
3. Terrain clearance in units of 60 cm.
4. Barometric altitude in units of 3 m.
5. Dopplerradar alongtrack and cross distances in
units of SO rn.
6. Time in seconds.
Data compilation involved the following:
l. Check of inight digital data and necessary cor
rections.
2. Calculation of coordinares.
3. Location of traverse and baseline intersections.
4. Adjustment of intersection points.
5. Calculation of, and correction for, drift.
6. Reduction of data to a common datum.
7. Reduction of corrected values for contouring.
A map of a 1 O x l O km portian of this survey is
shown in Figure 3.3ld.
Figure 3 .31 a, a provisional geological map, was
prepared with help from an earlier ground vertical
l. Total magnetic eld in units of 0.02 nT.
2. Total eld vertical gradient in units of 0.005 nT.
the anomaly near Mt. St. Gregoire is just above tbe
surface whereas the one near Mt. Bruno is about 870
m below tbe surface.
(2) The use of high sensitivity aeromagnetic data
has been described by Bhattacharyya (1971 ). In 1969,
the Geological Survey of Canada arranged an experi
mental highresolution survey in the Precarnbrian
shield of northern Ontario near Timmins that used a
cesiumvapor magnetometer with a sensitivity of
0.02 nT. Control of the survey was much tighter than
in conventional work at the time. Line spacing was
300 m at an average altitude of 250 m and flight
paths were straight within 100 mover 24 km. Double
baselines perpendicular to these were own in oppo
si te directions every 8 km. The total field was contin
uously recorded at a ground station. The following
were recorded on the aircraft:
Figure- 332. Cenereliz cd flow cbstt of compoteriz ed interpret stion. (From ;md
Hood. 1 973)
Calculate
model curve
lnilialize
parameters
1 21
Printout
bestfit values
Calculate residual
!
Compare and
determine errors
In rerpreter
chooses model
Input data
by gridding
Anomaly
Fefd exemples
lnitialize
List pararneters
Calcula te
Marquardt's).
to be varied
partial derivatives
by Powell algorithm
Search for mnimum
Calcula te
correction vector
j
Replace
e
partial deriva tives
9
in Powell alorithm
Genera te Calcula te ~
new model curve
..
new errors s

Calculare derivatives
~
near mnimum o
a.
Restar! Powell Calcula te
algorithm or end residuals
cession instrument (sensitivty 0.1 nT) at 300 m with
800 m line spacing, obviously shows much less detail
than the highresolution magnetic map in Figure
3.3ld. SeveraJ pronounced anomalies, probably due
to gabbro, are obvious on both aeromagnctic maps,
whereas the ground map does not show them clearly.
The cost of the hghresolution survey was about six
times greater than the conventional aeromagnetic
survey, but the difference would be much less today:
an equivalent ground survey on 120 m spacing would
cost five to six times as much.
intensity map (Fig, 3.32b). The bedrock in this arca,
cut by numerous NS diabase dikes, is an Archean
complcx of gabbro, granite, and mafic and felsic
voJ canics. There are three major fault systems: The
one striking N30W is the main control for thc
diabasc dikes, whereas the other two, trending WNW
and NE, appcar to have atrccted the dikes by shear
ing and deection,
The ground survey map shows much detal, but
the trends are broken up. The map f rom a convn
tonal survey, Figure 3.3lc, flown with a protonpre
Figure 3. 33 Belec lake snomslv, O nterio. (From McCrtJ th and 1 973. ) (a) Ar'ro
magnetic mep of the snomslv. (b) Profile AB end the enomslv from the
model. (e} lnferred geological ero~~ section.
{e)
- - . . - - - . P AL E OZOI C S E OI ME N TS - - - - - - - .
11
(b)
~ 30~
~ 600
"'
2 900
A~~- . - - - ' F _ L _ t _ G _ H _ T_ L ~I N _ E ; . _ ~v~B
300m
1 km
/
/
,
(a)
5km o
.i.
~SURVEY
"DATA
BE S T
FIT
DATA
<,
300
u..


400
500
Magnetic 1 22
Figure J .J .J . Vertical magnetic ground S U T\'L'~'. Noranda sros: C I =50 n T (AitC'r 5Pigc> I.
1 957. )
IOOm o
Powrrhnc
Building
Fcncc
Road

N
t
l
Stn. (nT) Sin. (nT)
ON 275 8N 40
1 220 9 10
2 224 10 IS
3 230 11 + 100
4 185 12 150
5 185 13 2 2 0
6 155 14 220
7 35
Table 3.5. verticet- cornponent readings
in serpentine z one.
drilling measurements). The bestfit model showed
an intrabasement magnetic zonc at a depth of 900 m,
which corresponds to 600 m of sediment. The mag
netic body is 1.730 m thick with a strilce length of 14
km and dips 82 north: the susceptibility contrast is
0.029 SI. which is typical of igneous rocks. The body
has its polarization vector dipping 64 with declina
tion 107. The local magnetic indination is 79,
which means that the body possesses significant re
manent magnetization.
r
1
1
1
~ c::r:::>
? ~;~e
l
1
' ,

i

The minimum of may be found by the method of
Gauss, least squares, steepest descent, or other tech
niques. Here it was found by a combination of the
Marquart and Powell algorithms (McGrath and
Hood, 1973). [See 2.7.9 for a similar gravity proce
dure.]
This example is o modeling a basement anomaly
in the Moose River basin of the Hudson Bay low
lands in northem Ontario, which is shown in Figure
3.33. The model was a thick, stccply dipping platc.
The Moose River basin contains about 600 m of
nonmagnetic Paleozoic sediments overlying a Pre
cambrian crystalline basement (based on seisrnic and
E( q' q2, ... q,.)
E(D(x,y)
(3) An example o computcr modeling followed
the procedure diagrammed in Figure 3.32. This mod
eling minimizes E. the difference between observed,
Dt x, y). and model anomalies T(x. y), at m points
through an iterative adjustment of 1 1 model pararne
ters, q1, q2, . , q,.:
Field eamples
pro(ile, Esst
SOON 400N 200N 600N
O IOO m
300N IOON
SOOO
10.000
IS,000
15,000nT
10,000
5000
eralization, estmate the depth and crosssection
of the body causing magnctic anomaly from this
pro file.
2. Figure 3.34 shows Zcomponent contoun ob
tained f rom a detaiJ ed ground magnetic survey
in the Noranda district, an area ricb in sulfides
and graphite. The rocks to the south are rhyo
lites, rhyolitic breccias, and tuffs. To the north
we find basic volcanics and tuffs with occasional
rhyolite. Where would you expect to ftnd anoma
lous suldes and/or graphite? Are they shallow
or decp? If they are suldes, what varietics can
be ruled out? Have the fence and power line
aff ected tbe readings to any extent?
3. Table 3.6 shows Zcomponent readings made 3
m apart on an E W traverse in the Quebec
Eastern Townships region. The large anomaly
was originally detected during a paeeandcom
pass exercise wben tbe magnetic declination sud
denly changed by about 110, an effect that did
not persist for any appreciable distance. Inter
pret the source with regard to location, depth,
dip, lateral extent, and possible mineral charac
ter.
4. Use a dipping sheet model of considerable strike
length to calculate and plot the total eld F and
vertical gradient aF given that D - b - lOd
and tbat the structure strikes (a) NS and (b)
EW, in a region where l . . . 60. B y neglecting
small tcrms, calculate the approximate maxi
rnum and minimum values of and in
(a) to check the NS profile.
The vertical component magnetometer readings
in Table 3. 7 are Irorn a detailed survey of an old
mining property where the primary metal was
zinc, witb sorne lowgrade coppcr and minor
silver. Assume that the source of the weak mag
L inr 450W
l. Chromite is found in serpentine in a certain
district. A magnetic analysis of several speci
mens indicates that tbere is less magnetite in the
chromtebearing serpentine than in barren ser
pentine. The Zcomponent readings in Table 3.5
were taken on a NS line at 8 m stations,
Assuming a twodimensional E W zone of rnin
3.9.
Table 3.7. Vertical- component readings
z inc propenv.
Stn. Stn.
(ft) (nl) (ft) (nT)
6 +005 4S o+ sos 130
s +00 so o+ 00 95
4 +50 52 O+SON 75
4 + 00 so 1+00 40
3 + 50 45 2 + 00 8
3 + 00 so 3 + 00 o
2 + 00 90 4 + 00 - 5
1 + 00 1 3 5
aw
7
6
5
4
3
2
1
o
7, CXX)
1,060
2.720
5,140
4, 260
2, 680
2.220
2.240
1,940
2.040
2, 320
2,080
2.080
1,800
3, 280
9,200
3,400
9.500
17W
16
15
14
13
1 2
11
10
9
(nT) Stn. (nT) Stn.
Table 3.6. Traverse in Quebec Eastern Townshps.
1 24
zones were resurveyed in followup work on SO ft
grids with occasional readings 5 to 10 ft apart
because of the small lateral exteot of tbe anoma
lies. Three of these detaiJ ed grids are sbown in
Figure 3.36b, e, d.
The differences, particularly in lateral extent,
strike axis, magnitude, and anomaly loeation, are
very apparent. What are they? Other surveys
have ndicated massive suldes at the centers of
Figure 3.36b, d but not at 164N, 103E. Given
thls additional information, would you conclude
that the magnetic anomalies are directJ y or indi
rectJ y related to the sulfides? Do they have any
association with the sulfides? WouJ d you rcach
the same conclusin if only Figure 3.36a were
available? Estmate the depth, lateral extent, atti
tude, susceptibility, and probable content of a
f cw of the anomalies in all four diagrams. Do
they show evidence of strong remanent magneti
:z.ation?
8. The verticalcomponent ground magnetic con
tours in Figure 3.37 were from a survey over a
netic anomaly is a dke of large strike length and
depth extent and use the method of Section 3. 7 .8
to determine its parameters.
6. The verticalcomponent ground profile shown in
Figure 3.35 is from a largescale multiplemethod
survey for base metals in Tanzania. The geology
is generally Precambrian metamorphosed sedi
ments and volcanics associated with granite. In
this regin, ~ 0.35 .T, the declination is 4W,
and the inclination is I ~ 31 N. Sulde bodies
with large EW extent are known in the vicinity.
lnterpret the anomaly.
7. The magnetic contours in Figure 3.36 show the
eect of line spacing (see also Bhattacbaryya,
Sweeney, and Godson, 1979). A small section
from what would normally be considered a de
tailed ground survey is shown in Figure 3.36a.
Lines were spaced 200 ft apart NS with station
readings every ft (reduced to 20 ft near
anomaJ ous vales). The contours of Figure 3.36a
are elongated NS because the line spacing is
greater tban the station spacing. Sorne ot the
Figure J . 36. Effect of line specing on megnetic data. (a) Conventiorut rrugnetic mep.
I 1 ,000 (o) O eteiled megnetic map lt 1 56N. 1 0TE.
(b)
IS8
LIDIE
1 25
N
UO~E 103E
' 1
1
Problems
LIS6N
m.
300 ft
1
LIS2N
IS4
100
(o)
ments and tbick overburden; the mineralized
zones frcquently extend for miles.
Diamond drilling has established that the
mineralization associated with two of these pro
files is pyritc and pyrrhotite, a third is graphitc
and pyrite, and the fourth graphitc and
pyrrbotite, and that they are located at f our
different dcpths. With this information, use the
magnetic data to locate the mineralized sections
as precisely as you can.
11. The two ground magnetic contour maps in Fig
ure 3.40 illustrate the effect of irregular topogra
phy on magnetic measurements (Oliver and
Hinze, 1985). The arcas surveyed in Figure 3.40
are only 60 miles apart. The terrain is quite
rugged in both, as can be seen f rom the dotted
contours, but the geology is entirely diff erent. In
the former, the rocks are sedimentary to great
depth; at the latter site, there are granites in
most of the north and west parts of the map and
sediments in thc lower ground at the south. lt is
niclcel prospect in nortbem Manitoba. Zone C is
approximately 4 km NE of zone A. There is
considerable ovcrburden tbroughout the area.
Electromagnetic survcys showed that both zones
wcre good conductors. One of thcm contained
oregrade nickel suldes, the other was barren of
suUides anci graphite. Is it possible to distinguish
the economic mincralization solely from the
magnetic results? Estimate the depths of the
main magnetic anomalies in the two zones,
9. A copper deposit of limited extent in the Rouyn
district of Quebec produced thc Z profiles in
Figure 3.38. Match the profiles toan appropriate
model given that the EW strike is very limited
and the inclination 7S. Wbat is the proba
ble magnetic mineral?
10. Figure 3.39 shows four vertcalcomponent mag
netic profiles from the Manitoba Nickel Belt,
obtained during largescale basemetal explo
ration programs. Sulfides and graphite oceur in
the Precambrian rocks below Paleozoic sedi
Figure 3. 36. (Continued) (e} Detailed msgnetic map centered st 1 64N, WJ E.
(d) Detailed magnetic map centered at 1 60N, 1 05E.
fd)
106
)"
.. ,
G
L160N
o IOOfi
LI05E
1
1
162
Magnetic metbods
(e)
104
LI J E
16l
101
1oon o
16l
1 26
southeast to indicate this clearly. However, a
pronounced magnetic low follows the topogra
phy from north to south, thcn west to east,
starting in the upper left arca. This is a clear
reftection of the terrain eff ect on ground magnet
ics.
Using Equations (3.33) and (3.71), apply thc
terrain correction at a few selected points on the
map or Figure 3.40a. For exarnple, the 100 nT
low on the steep slope near the bottom of the
map, midway between the eastwcst boundaries,
lies on the 425 ft contour; if we choose the 600 ft
elevation for O, the value of h will be 175 ft.
Reasonable values for circle radii would be r1
100 ft, r2 =200 ft, and so forth. For other
stations. one might select one at the top of the
thought that the sedments extend for sorne dis
tance up the bill in the lower lef t comer and the
assumed contact between the granites and vol
canic formations is somewhere in the upper le
portion of the map.
The uniform magnetic response over the sedi
ments in Figure 3.40a (about 300 nT maximum)
is to be expected becausc of the low susceptibil
ity of sandstone and limestones. There is no
particular correlation between the topographic
and magnctic contours, hence no need for a
topographic correction.
In Figure 3.40b the situation is quite differ
ent. A definite magnetic contrast exists between
the granites and the sedimentary area, altbough
tbe map does not extend far enough south and
Figure 3. 37. Cround contours. northern 1 > . 1 ,m1 tob< 1 . C T(X) n
2400
1 27
ZoneC
,,,~
= - - - ~~- - - - ~- - - - - - i " t S / l
~ ~
lkm
fects of smoothing with altitude. The ground
survey indicates ve distinct maxima, compared
with a single symmetrical anomaly in the air
bome contours. Furthermore, the largest of thesc
ground maxima is displaced about 650 m from
the center of the totaleld anomaly. The alti
tude of the aircraft was 300. m and the ground is
ftat. The magnetic inclination in the area is 75.
CaJ cuJ ate the depths and approximate lateral
extents of thc 6,000 nT and the larger of the
2,000 nT ground anomalies as well as the air
borne anomaly.
With the aid of Equation (3.71) and Table
3.2. carry out an upward continuation of the
ground data to 300 m by choosing r1 100 m
and h - S .
13. Figure 3.42 shows a 9.5 x 6.5 km section from a
Canadian govcmment aeromagnetic survey in
hill, one in thc northwcst comer, and one to the
southeast. Do these modified values aid the
magnetic interpretation in any way?
Repeat the procedure for severa! strategically
located stations on thc map of Figure 3.40b,
particularly in the area of the magnetic lows
following the stream gorge, (Obviously the best
method for handling analysis of this typc would
be to digitize thc contoured data and use a
computer.) Are the terrain corrections sign
cant? Would they be more reliabte if vertical
gradients had been mcasurcd? Why?
12. Tbe two sets of contours shown in Figure 3.41
illustrate tbc differences between airbomc total
fleld and ground verticalcomponent survey re
sults. Only the relative values are significant
because tbere is no relation bctween the absolutc
magnitudes. The airbome results show the et
figure verticsl- compooent megnetic Rouyn- Norsnd O uebec.
ISN H
100 rt
ION
Line 4 + 00 E
.,.... 11 1
J ON ISN
!to~ 1
Lne 3Z+ 00 E
Llne 2 + 00 E
~~
t 1
S N
~~ b
1
S N
Line 1 +OOE
1 1
ISN
A.
+e 1
ION
1 adro
SN
4000o~t...:t:=::::t:::~~
Line O+ OOE
.. 1
SN J ON ISN
Magnetic metbods 1 28
part of the section. Could it be entirely or partly
the result of topography? Is it the reftection of a
single magnetic structure? Is it possible that the
larger magnetic low area, contained in the
200 nT contour striking roughly easiwest, might
represent a distinctly different structure?
14. The Zcomponent data in Table 3.8 constitute
airbome and ground profiles along approxi
mately the same line crossing a strong magnetic
anomaly of great length, stri.king roughly EW.
The airborne magnetometer was a heavily
darnped verticalcomponent instrument at eleva
tions of 300 to 400 It above ground. The ground
instrument was a conventional ftuxgate magne
tometer. Station intervals are in feet and air
bome station 800N corresponds approximately
to a point between 400 and 500N on the ground
traverse, Determine the depth, cross section, ap
proximate susceptibility, and dircction of mag
northwest Newfoundland. The rocks in the area
are sedimentary, consisting of sandstones, shales,
and limestones with sorne dolomite. In the upper
half and the lower lef t quarter of the section the
topography is flat; the average elevation is 100
m. A steep escarpment, in the shape of an in
verted U with apex to the north, occupies the
lower middle portion of the figure. It follows the
closed 200 nT contour on the J ert, continues
north and east to overlap the east half of the 100
nT low, and then tums southeast between the
300 nT contours on the lower right, Ths scarp
rise about 200 m, in places having a slope of
nearly 30. As a result, the magnetic high in the
lower part of the diagram is on a 300 m plateau.
Flight lines were eastwest, 300 m above ground
level.
With this inormation, make an interpretation
of the magnetic anomaly in the lower central
100
200nT
O 200 m
( 2 )
o
1 1
o IOOm
200nT
(3)
100
1!0
300nT
200
100
o
1 1
. ..... '
o IOOm
Figure 3. 39. verticel- component megnetic Manitoba Niclr. el Belt.
1200nT
1100
_ . _ ~, _ _ _ . _ ~. . . . . _ _ , _ ~. . _ . . . . . . . , , ~. . . _ . . . . . . . . . ~. . . _ - ' - ' O
( 1)
1 29 Problems
F1 6ure 3. 40. Effect of topography on magnetic meesurements. (a) Elevation (dashed)
and Z- component (so/id) contours in sn ilreil o( sedimentary rocks. (b) Elevation end
Z component contours in en area of granitic snd sediment. J ry rocks.
(/> )
IOE B.L.
"
"'
/
/ J OW
......
ION ... ' ... 400 _ _
'
....
~
'
'
4fo 20w,
20N 20N

400
...
,
(ol
and flight altitude 300 m.
; .
/
,
'
1
' ..
. ., , ., , , , ,
- -
: ...


, , , ,
, , ,
,
; ,,
, ,
I
1
1
....

...
. ..
Figure J . 41 . of ground verticsi- component data ~1 /ith eirbome
data.
Vrrticalcomponenl 1round
contours. lntc:f"al ~IOOOnT
Totalfield airborne
contours. lntrnal ~IOOnT

..... .,,
, , .....
<t
I f
/ '
, . '
,, I
- . . I
/. .,
I I
I I
\ I
,,,.,
........... _
o
l
I
,
/
;
,,,
,.
/
I
/
/
I
I
I
I
I
I
1
I
I
I
I
1
1
l
\
1
\
'
~~',
~ .,
'
'
,_
1
1
1
1
1
I
1
I
I
I
/
,
/
,,
/
,,,''
/
/
/
"
/
/
,,
!:!
s
1
1
'
\
'\
' \
'
, , ...
/
, ,
,,
/
/
I
/
I
I
'
1
'
I
I
I
I
t
1
1
l
_ . . .

, , ...
, ,
/
,,
/
/
I
I
I
' I
1
,
I
I
1
1
1
1
l
'
l
~
1
1
\
1
1
1
1
\
1
1
1
1
1
1
1
1
' 1
1
1
,
I
I
,
1
1
1
\
' \

1 31 Problems
anomalies produced by structures in the sedi
ments or in the underlying basement rocks? Ana
lyze this aeromagnetic section as precisely as
possible, with particular emphasis on depths to
the sources. Altitude of the aircraft was 300 m
above ground level and the ftight lines were
EW. The topography is essentially ftat througb
out.
16. Figure 3.44 shows a portion of an EW aero
magnetic prole. The fiduciaJ marks (numbers at
the bottom) are 2 km apart and the aircraft was
netintion of thc source rrom cach traverse, As a
check on the results, continue the ground profile
upward to find out if it matches the airborne
pro file.
15. A scction from Canadian government aeromag
netic maps of thc St. Lawreocc lowlands sedi
mentary regin is reproduocd in Figure 3.43.
There are at least three large structural fe atures
producing the magnetic anomalies. Two are well
documcnted geologically. Can you distinguish
any fault zones? Any domes or pJ ugs? Are these
Figure 3. 43. Total fiefd aeromagnetic contot1 rs, S t. tawrence lowlands. C. I. 1 00 nT.
'J ao /
170o _
IOo
lJ oo~~
- - - - = ~
36,000
28, 600
16,000
6, CXX>
4,600
6 00N
700
800
900
1,000
4,600
7,400
13,700
28,600
40.000
40, CXXJ
Stn. (nT) (nT)
Airborne
Stnl. (nT) Stn.
ON 600 ON
400 2.600 100
600 4,100 200
800 S,700 300
1, (XX) 4,050 400
1.200 2,760 500
1,500 750
Ground
Magnetic
Table 3.8. Ground and airborne survey.
1 32
Clark, A. J . 1986. Archaeological geophysics in Brirain.
sr. 140313.
Dean, W. C. 1958. Frequency analysis Ior gravity and
magnetic interpretauon. 23. 97127.
Gay. S. P. 1967. Standard curves for interpretation of
rnagnetic anomalies over long tabular bodies, In
vol. 2. pp. 51248. Tulsa: Soeiety
o{ Exploration Geophysicisrs.
Grant, F. S .. and Martn. L. 1966 lntcrprctation of
acromagnetic anomalies by the use of eharacteristic
curves. 31. 135414.
Grauch, V. J . S . and Campbcll. D. L . 19M4. Does draping
aeromagnctic data reduce terraininduced eects?
49, 7580.
Green. R. 1960. Remancnt magnetization and the
interpretation of magnctic anomalies.
8, 98110.
Gunn, P. J . 1975. Linear transformations gravity and
magnetic fields. 23. 30012.
Gupta, V. K., and Fitzpatrick. M. M. 1971. Evaluation of
terrain effects in ground magnetic surveys,
36. 5829.
Hague, B. 1957. New
Y ork: Putnam.
Hahn, A .. Kind, E. G., and Mishra. D. G. 1976. Depth
estimation of magnetic sources by mcani; o Fourier
amplitude spectra, 24. 287308.
Hanson, R. D. and Miyazki. Y . 1984. Continuation of
potential fields bctwecn arbitrary surfaces. Geophysics
49, 78995.
Hartman. R. R. Teskey. D. J .. and Fricdberg. J . L . 1971. A
system for rapid digital acromagnctic intcrprctation.
36, 891918.
Hendcrson. R. G. 1960. A comprchensive systcm o
automatic computation in magnetic and gravity
intcrpretation. 25. 56985.
Hood. P. J . 1965. Gradient measurements in aeromagnctic
survcying. 30. 891902.
Hood. P .and McClure. D. J . 1965. Gradient measurements
in ground magnctic prospecting. 30. 40310.
J ain, S. 1976. An automatic method of direct intcrpretation
magnetic profi1cs. 41. 53141.
Kilty. K. T. 1983. Werncr deconvolution or profilc potential
field data. Gft1 physics 48. 234 7.
Kip, A. F. 1962.
Ncw York: McGrawHiJ I.
K.oulomzine. T., Lamontagnc. Y . and Nadeau. A. 1970.
New mcthods for dircc1 intcrprctation of magnctic
anomalics caused by nclincd dikes of infinitc lcngth.
35, 81230.
Lcite, L. W. B . and Lcao. J . W. D. 1985. Ridge regrcssion
applied to the invcrsion of twodimensionaJ
aeromagnetic anomalcs. 50:1294306.
Atneck. J . 1963. Magnetic anomaly trend and spacing
pauerns. 28. 37995.
AlChalabi, M. 1971. Sorne studies relating to non
uniqucncss in thc gravity and magnctic nverse
problem. 36, 83S54.
Baranov, V. 1957. A new method for interpretation of
acromagnctic maps: Pseudogravimetric anomalies.
22. 35983.
Bamctt, C. T. 1976. Theoretical modc1ing of the magnetic
and gravitational ficlds of an arbitrarily shapcd 3D
body. 41, 135364.
Barongo, J . O. 1985. Method for dcpth estimation on
acromagnctic vertical gradienl anomalies.
so, 9638.
Bhattacharyya, B. K.. 1964. Magnetic anomalies due to
prismshapcd bodies with arbitrary polarization.
29, 51731.
Bhattacharyya. B. K.. l 96S. Twodimensional harmonic
anaJ ysis as a tool for magnetc intcrpretation.
30, 82957.
Bhanacharyya, B . K. 1966. Continuous spectrum of total
magnetieeld anomaly due to a rectangular prismatic
body. 31, 97121.
Bhattacharyya, B. K. 1970. Some importan! considerations
in the acquisition and treatmcnt of highresolution
acromagnetic data.
12, 2144, 456.
Bhattacharyya, B. K. 1971. An automatic mcthod of
compilation and mapping of highresolution
aeromagnctic data. 36, 695 716.
Bhauacharyya. B. K., and Navolio, M. E. 1976. A fast
Fourier transform method for rapid computation or
gravity and magnetic anomalies duc to arbitrary bodies,
24, 63349.
Bhattacharyya, B. K., Swecncy, R. E., and Godson, R. H.
1979. Integration or aeromagnctic data acqured at
ditTcrcnt times with varying elevations and line
spacings. Geophysics 44, 74252.
R E F E R E N CE S
flown ar a constant barometric elevation o 750
m. Analyze the profile using the methods of
Section 3.7.11 and Equations (3.89).
17. The schematics shown in Figure 3.45 represent
two relatively common basement structures: (a)
a contact between beds of grcat strike length and
depth extent and (b) a uniforrn bed with a step,
Assume D/d - 1.1 and a NS strike for both
fe atures. Plot both pro files and compare the
maximum anomalies.
Figure 3. 45. Bssemen: structures. (a) Contect between sldbs of different
(b) U niform bed with a step.
f !

Surfaee
Magnetic
(g)
d

Surface
1 34