Вы находитесь на странице: 1из 365

Solved Problems in Geophysics

Solving problems is an indispensable exercise for mastering the theory underlying the
various branches of geophysics. Without this practice, students often nd it hard to
understand and relate theoretical concepts to their application in real-world situations.
This book is a collection of nearly 200 problems in geophysics, which are solved in
detail showing each step of their solution, the equations used and the assumptions made.
Simple gures are also included to help students understand how to reduce a problem to its
key elements. The book begins with an introduction to the equations most commonly used
in solving geophysical problems. The subsequent four chapters then present a series of
exercises for each of the main, classical areas of geophysics gravity, geomagnetism,
seismology and heat ow and geochronology. For each topic there are problems with
different degrees of difculty, from simple exercises that can be used in the most elementary courses, to more complex problems suitable for graduate-level students.
This handy book is the ideal adjunct to core course textbooks on geophysical theory. It is
a convenient source of additional homework and exam questions for instructors, and
provides students with step-by-step examples that can be used as a practice or revision aid.
Elisa Buforn is a Professor of geophysics at the Universidad Complutense de Madrid (UCM)
where she teaches courses on geophysics, seismology, physics, and numerical methods.
Professor Buforns research focuses on source fracture processes, seismicity, and seismotectonics, and she is Editor in Chief of Fsica de la Tierra and on the Editorial Board of the
Journal of Seismology.
Carmen Pro is an Associate Professor at the University of Extremadura, Spain, where she has
taught geophysics and astronomy for over 20 years. She has participated in several
geophysical research projects and is involved in college management.
Agustn Udas is an Emeritus Professor at UCM and is the author of a large number of papers
about seismicity, seismotectonics, and the physics of seismic sources, as well as the
textbook Principles of Seismology (Cambridge University Press, 1999). He has held
positions as Editor in Chief of Fsica de la Tierra and the Journal of Seismology and as
Vice President of the European Seismological Commission.

Solved Problems in Geophysics


ELISA BUFORN
Universidad Complutense, Madrid

CARMEN PRO
Universidad de Extremadura, Spain

AGUSTN UDAS
Universidad Complutense, Madrid

cambridge university press


Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, So Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by
Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9781107602717
# Elisa Buforn, Carmen Pro and Agustn Udas 2012
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2012
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloging-in-Publication Data
Buforn, E.
Solved problems in geophysics / Elisa Buforn, Carmen Pro, Agustn Udas.
p. cm.
Includes bibliographical references.
ISBN 978-1-107-60271-7 (Paperback)
1. GeophysicsProblems, exercises, etc. I. Pro, Carmen. II. Udas Vallina, Agustn. III. Title.
QC807.52.B84 2012
550.78dc23
2011046101
ISBN 978-1-107-60271-7 Paperback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third-party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.

Contents

Preface

1 Introduction
Gravity
Geomagnetism
Seismology
Heat ow
Geochronology

2 Gravity
Terrestrial geoid and ellipsoid
Earths gravity eld and potential
Gravity anomalies. Isostasy
Tides
Gravity observations

3 Geomagnetism
Main eld
Magnetic anomalies
External magnetic eld
Main (internal), external, and anomalous magnetic elds
Paleomagnetism

4 Seismology
Elasticity
Wave propagation. Potentials and displacements
Reection and refraction
Ray theory. Constant and variable velocity
Ray theory. Spherical media
Surface waves
Focal parameters

5 Heat ow and geochronology


Heat ow
Geochronology
Bibliography
v

page vii
1
1
4
6
10
11
13
13
25
53
95
116
121
121
142
156
174
201
208
208
211
224
243
277
307
324
335
335
345
352

Preface

This book presents a collection of 197 solved problems in geophysics. Our teaching
experience has shown us that there was a need for a work of this kind. Solving problems
is an indispensable exercise for understanding the theory contained in the various branches
of geophysics. Without this exercise, the student often nds it hard to understand and relate
the theoretical concepts with their application to practical cases. Although most teachers
present exercises and problems for their students during the course, the hours allotted to the
subject signicantly limit how many exercises can be worked through in class. Although
the students may try to solve other problems outside of class time, if there are no solutions
available this signicantly reduces the effectiveness of this type of study. It helps, therefore, both for the student and for the teacher who is explaining the subject if they have
problems whose solutions are given and whose steps can be followed in detail. Some
geophysics textbooks, for example, F.D. Stacey, Physics of the Earth; G.D. Garland,
Introduction to Geophysics; C.M. Fowler, The Solid Earth: An Introduction to Global
Geophysics; and W. Lowrie, Fundamentals of Geophysics, contain example problems, and,
in the case of Staceys, Fowlers, and Lowries textbooks, their solutions are provided on
the website of Cambridge University Press. The main difference in the present text is the
type of problems and the detail with which the solutions are given, and in the much greater
number.
All the problems proposed in the book are solved in detail, showing each step of their
solution, the equations used, and the assumptions made, so that their solution can be
followed without consulting any other book. When necessary, and indeed quite often, we
also include gures that allow the problems to be more clearly understood. For a given
topic, there are problems with different degrees of difculty, from simple exercises that can
be used in the most elementary courses, to more complex problems with greater difculty
and more suitable for teaching at a more advanced level.
The problems cover all parts of geophysics. The book begins with an Introduction
(Chapter 1) that includes the equations most used in solving the problems. The idea of
this chapter is not to develop the theory, but rather to simply give a list of the equations
most commonly used in solving the problems, at the same time as introducing the reader to
the nomenclature. The next four chapters correspond to the division of the problems into
the four thematic blocks that are classic in geophysics: gravity, geomagnetism, seismology,
and heat ow and geochronology. We have not included problems in geodynamics, since
this would depart too much from the approach we have taken, which is to facilitate
comprehension of the theory through its application to specic cases, sometimes cases
which are far from the real situation on Earth. Indeed, some of the problems may seem a bit
articial, but their function is to help the student practise with what has been seen in the
vii

viii

Preface

theory. Neither did we want to include specic problems of geophysical prospecting as this
would have considerably increased the length of the text, and moreover some of the topics
that would be covered in prospecting, such as gravimetric and geomagnetic anomalies, are
already included in other sections of this work.
Chapter 2 contains 68 problems in gravity divided into ve sections. The rst section is
dedicated to the terrestrial geoid and ellipsoid, proposing calculations of the parameters
that dene them in order to help better understand these reference surfaces. The second
corresponds to calculating the gravitational eld and potential for various models of the
Earth, including the existence of internal structures. Gravity anomalies are dealt with in the
third section, with a variety of problems to allow students to familiarize themselves with
the corrections to the observed gravity, with the concept of isostasy, and with the Airy and
Pratt hypotheses. The fourth section studies the phenomenon of the Earths tides and their
inuence on the gravitational eld. The last section is devoted to the observations of
gravity from measurements made with different types of gravimeters and the corrections
necessary in each case. We also include the application of these observations to the
accurate determination of different types of height.
Chapter 3 contains 42 problems in geomagnetism divided into ve sections. The rst is
devoted to the main (internal) eld generated by a tilted dipole at the centre of the Earth. It
includes straightforward problems that correspond to the calculation of the geomagnetic
coordinates of a point and the theoretical components of the magnetic eld. This section
also introduces the student to the use of the principal units used in geomagnetism. The
second considers the magnetic anomalies generated by different magnetized bodies and
their inuence on the internal eld. The third section is devoted to the external eld and its
variation with time. In the fourth section, we propose problems of greater complexity
involving the internal eld, the external eld, and anomalous magnetized bodies at the
same time. The last section is devoted to problems in paleomagnetism.
Chapter 4 contains 69 problems in seismology divided into seven sections. The rst
presents some simple exercises on the theory of elasticity. The second addresses the problem
of the propagation of seismic energy in the form of elastic waves, resolving the problems on
the basis of potentials, and calculating the components of their displacements. We study the
reection and refraction of seismic waves in the third section. The fourth is devoted to the
problem of wave propagation using the theory of ray paths in a plane medium of constant and
variable velocity of propagation. The fth studies the problem of the propagation of rays in a
spherical medium of either constant or variable propagation velocity, with the calculation of
the travel-time curves for both plane and spherical media. The sixth section contains
problems in the propagation of surface waves in layered media. The seventh section is
devoted to problems of calculating the focal parameters and the mechanism of earthquakes.
Chapter 5 includes 11 problems in heat ow with the propagation of heat in plane and
spherical media, and seven problems in geochronology involving the use of radioactive
elements for dating rocks.
Finally, we provide a bibliography of general textbooks on geophysics and of specic
textbooks for the topics of gravity, geomagnetism, and seismology. We have tried to
include only those most recent and commonly used textbooks which are likely to be found
in university libraries.

ix

Preface

In sum, the book is a university text for students of physics, geology, geophysics,
planetary sciences, and engineering at the undergraduate or Masters degree levels. It is
intended to be an aid to teaching the subjects of general geophysics, as well as the specic
topics of gravity, geomagnetism, seismology, and heat ow and geochronology contained
in university curricula.
The teaching experience of the authors in the universities of Barcelona, Extremadura,
and the Complutense of Madrid highlighted the need for a work of this kind. This text is
the result of the teaching work of its authors for over 20 years. Thanks are due to
the generations of students over those years who, with their comments, questions, and
suggestions, have really allowed this work to see the light. We are also especially grateful
to Prof. Greg McIntosh who provided us with some problems on paleomagnetism, to
Prof. Ana Negredo for her comments on heat ow and geochronology problems, and to
Dr R.A. Chatwin who worked on translating our text into English.
The text is an extension of the Spanish edition published by Pearson (Madrid, 2010).

E. BUFORN, C. PRO AND A. UDAS

Introduction

Gravity
As a rst approximation the Earths gravity is given by that of a rotating sphere. The
gravitational potential of a sphere of mass M is:
V

GM
r

where r is the position vector (Fig. A) and G the universal gravitational constant.
If the sphere is rotating with angular velocity o the centrifugal potential at a point on the
surface is given by
1
F o2 r2 sin2 y
2
where y is the angle that r forms with the axis of rotation.
The gravity potential is their sum U V F.
The value of the acceleration due to gravity (the gravity force) is given by the gradient
of the potential:
g rU
The radial component of the gravity force is given by
gr

GM
ro2 sin2 y
r2

The potential of the Earth to a rst-order approximation corresponds to that of a rotating


ellipsoid, and is given by


 m  r 2 2
GM a J2 a3 
2
U
3sin 1
cos
a r 2 r
2 a

where 90 y is the geocentric latitude and a the equatorial radius.


The coefcient m is the ratio between the centrifugal and gravitational forces on the
sphere of radius a at the equator:
m

a3 o2
GM

Introduction

North Pole

P
q

r
a
Equator

Fig. A

The dynamic form factor J2 is dened as


J2

C A
a2 M

where C and A are the moments of inertia about the axis of rotation and an equatorial axis.
The attening of the ellipsoid (the shape of the Earth to a rst-order approximation) of
equatorial and polar radius a and c is:
a

c
a

In terms of J2 and m,
3
m
a J2
2
2
The dynamic ellipticity is
H

A
C

The gravity attening is


b

gp

ge
ge

where gp and ge are the normal values of gravity at the pole and the equator, respectively.

Gravity

The gravity at a point of geocentric latitude = 90




g ge 1 bsin2

y is

The geocentric latitude of a point is the angle between the equator and the radius vector of
the point. The geodetic latitude is dened as the angle between the equatorial plane and the
normal to the ellipsoid surface at a point. Astronomical latitude is the angle between the
equatorial plane and the observed vertical at a point.
The normal or theoretical gravity at a point of geocentric latitude referred to the
GRS1980 reference ellipsoid is
g 9:780327 1 0:0053024 sin2

0:0000059 sin2 2 m s

The effect of the Sun and Moon on the Earth is to produce the phenomenon of the tides.
If one considers more generally the tidal effect due to an astronomical body of mass M at a
distance R from the centre of the Earth, one must add the corresponding potential, which, in
the rst-order approximation, is given by
c

GMr2 
3cos2 #
2R3

where r is the geocentric radius vector of the point, and # is the angle the position vector r
forms with the distance vector R.
Gravity anomalies, dened as Dg g g, are the effects of the existence of anomalous
masses inside the Earth. The gravity anomaly along the Z (vertical) axis at a point distance
x along the horizontal axis produced by a sphere of radius R, density contrast Dr, and
buried at a depth d, is given by
gx; z

@Va
GM z d
h
i3=2
@z
x2 z d2

where Va is the potential produced by the anomalous spherical mass DM 4/3pR3 Dr.
For problems in two dimensions, one uses the anomaly produced by an innite horizontal cylinder at depth d, perpendicular to the plane under consideration. The anomalous
potential is given by
0
1
and the anomaly by

1
B
C
Va 2pGra2 ln@qA
x2 z d2

gx; z

@Va 2pGra2 z d

@z
x2 z d2

To correct for the height above sea level at which measurements are made, one uses the
concepts of the free-air and Bouguer anomalies. The free-air anomaly is
gFA g

g 3:086h

Introduction

where g is the observed gravity, h the height in metres, and the anomaly is obtained in
gu (gravity units) mm s 2.
The Bouguer anomaly is
g B g

g 3:086

0:419rh

with r being the density of the plate of thickness h.


To account for isostatic compensation at height in mountainous areas, one adds an
isostatic correction which can be calculated assuming either the Airy or Pratt hypotheses.
With the Airy hypothesis, the root t of a mountain is given by
t

rc
rM

rc

where rc and rM are the densities of the crust and mantle, and h is the height of the
mountain. For an ocean zone, with water density ra, the anti-root is
t0

rc
rM

ra 0
h
ra

With the Pratt hypothesis, the density contrast in a mountainous area is


r r

r0

h
r
Dh 0

where D is the level of compensation, h the height of the mountain, and r0 the density at
sea level. For an oceanic zone of depth h0 :
r0 D ra h0
D h0
0
r r r0
r0

The isostatic correction can be calculated using a cylinder of radius a and height b, whose
base is located at a distance c beneath the point, and with density contrast Dr:

q p

I
a2 c 2
C 2pGr b a2 c b2
For mountainous zones, with the Airy hypothesis: b t, c h H t (H crustal
thickness, h height of the point); and with the Pratt hypothesis: b D, c D h.

Geomagnetism
To a rst approximation, the internal magnetic eld of the Earth can be approximated by a
centred dipole inclined at 11.5 to the axis of rotation. The potential created by a magnetic
dipole at a point distant r from its centre and forming an angle y with the axis of the dipole is
F

Cm cos y
r2

Geomagnetism

GNP
90 fB
l lB
180 l

90 f

GMNP
q = 90 f

Fig. B

where C m0/4p with m0 4p  10 7 H m 1, and m is the dipole moment in units


of A m. The product Cm is given in T m3.
The components of the magnetic dipole eld B are:
Br
By

@F
2Cm cos y

@r
r3
1 @F
2Cm sin y

r @y
r3

In the centred dipole approximation for the Earths magnetic eld, the geomagnetic
coordinates (f*, l*) of a point (y 90
f*) in terms of its geographic coordinates
(f, l) and those of the Geomagnetic North Pole (GMNP) (fB, lB) can be calculated using
the expressions of spherical trigonometry (Fig. B):
sin f sin fB sin f cos fB cos f cosl
sin l

sinl

lB

lB cos f
cos f

The vertical and horizontal components of the eld, the geomagnetic constant B0, and the
total eld are given by:
Z  2B0 sin f
H  B0 cos f
Cm
B0 3
a
q
p

F H 2 Z 2 B0 1 3sin2 f

The units used for the components of the magnetic eld are the tesla T and the nanotesla
nT 10 9 T. The NS (X*) and EW (Y*) components are
X  H cos D
Y  H sin D
and the declination and inclination are given by

Introduction

cos fB sinl
cos f
tan I  2 tan f

sin D

lB

The radius vector at each point of the line of force is:


r r0 cos2 f r0 sin2 y
where r0 is the radius vector of the point of the line of force located at the geomagnetic
equator.
Magnetic anomalies are produced by magnetic materials within the Earth. The anomalous potential due to a vertical dipole buried at depth d is
FA

Cmcosy
Cmz d
h
i3=2
r2
x2 z d2

The vertical (z) and the horizontal (x) components of the magnetic anomaly at the surface
(z 0) produced by a vertical magnetic dipole at depth d are:
Z
X

Cm2d 2

x2

x2 d 2 5=2
3Cmxd
x2 d 2 5=2

The Earth is affected by an external magnetic eld produced mainly by the activity of the
Sun. This eld is variable in time, with distinct periods of variation. The most noticeable is
the diurnal variation (Sq) with a maximum at 12 noon local time. The most important nonperiodic variations are the so-called magnetic storms.

Seismology
Earthquakes produce elastic waves which propagate through the interior and along the
surface of the Earth. Using the plane-wave approximation, the displacements of the
internal P- and S-waves (uiP and uiS) can be obtained from a scalar potential and a vector
potential:


ui ui P ui S ri r  cj
i



A exp ika gj xj at


cj Bj exp ikb gj xj bt

where A and Bj are the amplitudes, xj the coordinates of the observation point, ka and kb the
wavenumbers, gj are the direction cosines dened from the azimuth az and angle of
incidence i of the ray as:

Seismology

g1 sin i cos az
g2 sin i sin az
g3 cos i
and a and b are the P- and S-wave velocities of propagation, respectively, dened from
the Lam coefcients (l and shear modulus m) and the density r:
s
l 2m
vP a
r
r
m
vS b
r
Units used are: displacement amplitudes (u) in m; potential amplitudes (A, Bi) in 10
wavenumber (k) in km 1; and wave velocity (a, b) in km s 1.
Poissons ratio is dened in terms of the Lam coefcients as
s

m 2;

l
2 l m

The angle of polarization of S-wave e is dened as


e tan

uSH
uSV

where uSH is the amplitude of the SH component, and uSV that of the SV component. SH
and SV are the horizontal and vertical components of the S-wave on the wavefront plane.
The coefcients of reection V and transmission W are given by the respective
ratios between the amplitudes of the reected or transmitted potentials and the incident
potential:
A
A0
A0
W
A0
V

where A0 is the amplitude of the incident wave potential, A that of the reected potential,
and A0 of the transmitted potential.
Snells law for plane media is expressed as
p

sin i
v

and for spherical media


p

r sin i
v

where p is the ray parameter, i the angle of incidence, v the propagation velocity of the
medium, and r the position vector along the ray.

Introduction

In the case of plane media with propagation velocity varying with depth v(z), the
epicentral distance and the travel time of a ray for a surface focus are given by
h
pdz
x 2 p
 2 p2
0
h
2 dz
t 2 p
 2 p2
0

where  v 1 and h is the depth of maximum penetration of the ray. The variation of the
epicentral distance x with the ray parameter p is given by
dx

dp
where

2
p 2
B0 20 p2
B

dB
dz
pdz

 2 p2
0 B2

1 dv
v dz

In spherical media with velocity varying with depth v(r), the epicentral distance,
trajectory along the ray, and travel time are given by
r0
p
dr
p
2
2
r

r2
rp
r0
dr
s 2 p
2 r2
rp
r0
dr
p
t2
r2
rp v  2

where  rv 1, r0 is the radius at the surface of the Earth, and rp is the radius at the point
of maximum penetration of the ray.
The variation of the distance from the epicentre D with the ray parameter p in a spherical
medium is
dB
B
dr
d
2
drp
p 2

dp
1 B0 20 p2
B 2  2 p2
0 1
where
r dv
B
v dr
The radial and vertical components (u1 and u3) of surface waves can be obtained from
the potentials and c. The transverse component (u2) is kept apart
@ @c
;1 c;3
@x1 @x3
u2 C exp iksx3 ikx1 ct
@ @c

;3 c;1
u3
@x3 @x1
u1

Seismology

where c is the wave propagation velocity and


A exp ikrx3 ikx1

ct

c B exp iksx3 ikx1


r
c2
r
1
a2
s
c2
1
s
b2

ct

For surface waves, c < b < a, and hence r and s are imaginary.
For dispersive waves, the relationship between the phase velocity c and the group
velocity U is
dc
U ck
dk
where k is the wavenumber.
The position of the seismic focus is given by the coordinates of the epicentre (0, l0)
and the depth h. The time is that of the origin of the earthquake t0. The size is given by
the magnitude which is proportional to the logarithm of the amplitude of the recorded
waves. For surface waves this is:
Ms log

A
1:66 log  3:3
T

where A is the amplitude of ground motion in microns, T is the period in seconds, and the
epicentral distance in degrees.
The magnitude of the moment is given by
2
Mw log M0
3

6:1

where M0 is the seismic moment in N m (newton metres). The seismic moment is related to
the displacement of the fault u and its area S:
M0 muS
The mechanism of earthquakes is given by the orientation of the fracture plane (fault)
dened by the angles (azimuth), d (dip), and l (slip angle or rake), or by the vectors n
(the normal to the fault plane) and l (the direction of slip).
The elastic displacement of the waves produced by a point shear fault is

 @Gki
uk xs ; t mut S li nj lj ni
@xj

where Gki is the mediums Greens function which, for an isotropic, homogeneous, innite
medium, and P-waves in the far-eld regime, is given by

1
r
GPki
gi gk d t
2
4pra r
a

10

Introduction

The P-wave displacements are given by:


uPk xs ; t


u_ t S 
m l i nj l j ni g i g j g k
3
4pra r

This equation can be expressed also in terms of the moment tensor Mij
uPk xs ; t

M_ ij t
ggg
4pra3 r i j k

Mij is a more general representation of a point source.

Heat ow
The Fourier law of heat transfer by diffusion states that the heat ux q_ is proportional to the
gradient of the temperature T:
q_

KrT

where K is the thermal conductivity coefcient. The units of heat ow are W m 2.


The heat diffusion equation, assuming that K is constant, is given by
kr2 T

e
@T

rCv
@t

where Cv is the specic heat, r the density, e the heat generated per unit volume and unit
time (heat sources), and k the thermal diffusivity:
k

K
rCv

If there are no heat sources, the diffusion equation is


kr2 T

@T
@t

In the case of one-dimensional ow with periodic variation of temperature over time,


one has:
 r
r


o
o
T z; t T0 exp
zi
z ot
2k
2k
where z is the vertical direction (positive towards the nadir) and o the angular frequency.
In the case of stationary one-dimensional solutions (T constant in time) one obtains from
the diffusion equation:
T

e 2 q_ 0
z z T0
2K
K

11

Geochronology

where T0 and q_ 0 are the temperature and ow at the surface (z 0).


For a spherical Earth, assuming that the thermal conductivity is constant, and that
the amount of heat per unit volume depends only on time, the diffusion equation takes the
form:
2

@ T 2 @T
@T
K

et rCv
@r2 r @r
@t
where r is the radial direction.
For the stationary case, the above equation reduces to

1 d
e
2 dT
r

r2 dr
dr
K
Integrating twice, one has
T T0

e  2
R
6K

r2

where T0 is the temperature at the surface (r R).

Geochronology
Geochronology is based on determining the age of a rock by measuring the decay of its
radioactive elements. In a sample of radioactive material, the number of atoms that have
yet to disintegrate after time t is given by
nt n0 e

lt

where n0 is the initial number of atoms, and l the decay constant. The rate of decay dn/dt is
the activity R, so that
R R0 e

lt

where R0 is the initial activity (at t 0).


The half-life (or period) of the sample is the time it takes for the activity R to fall to half
its initial value. It is given by:
T1=
2

0:693
l

The mean life-time t of one of the atoms that existed at the start is given by:
t

1
l

If a sample consists of NR radioactive nuclei and NE stable nuclei, the time to arrive at the
propotion NE/NR is given by

12

Introduction

1
NE
1
l
NR

If the rubidiumstrontium (Rb-Sr) method is used to date a sample, a correction must be


made for the contamination of the stable 86Sr isotope relative to the radioisotope 87Sr:



87
87
87

Sr
Sr
Rb  lt

e
1
86 Sr total
86 Sr initial
86 Sr


This expression corresponds to a straight line (isochrone) of slope e lt 1 and intercept

87
Sr
corresponding to the initial content 86 initial .
Sr

Gravity

Terrestrial geoid and ellipsoid


1. Calculate the geodetic and geocentric latitudes of a point P on the ellipsoid
whose radius vector is 6370.031 km, given that m 3.4425  103, GM 39.86005 
1013 m3 s2, and 6356.742 km is the polar radius. Determine J2 and b.
The major semi-axis (equatorial radius) is a and the minor (polar radius) is c, the geocentric
latitude is , and the geodetic latitude is d (Fig. 1).
The coefcient m is given by the equation
m

o 2 a3
GM

where G 6.671011 m3 kg1 s2 is the gravitational constant, M the Earths mass, and
the angular velocity is o = 2p/T, where T is the rotation period (T 24 h). We obtain for
the semi-axis a the value
a

1=3
mGMT 2
6378:127 km
4p2

The Earths attening a can be obtained directly since we already know a and c so
a

a  c 6378:127  6356:742

3:3529  103
c
6378:127

The radius vector to the point P is given by the equation r a(1  a sin2 )
From this equation we can calculate the geocentric latitude :


6370:031 6378:127 1  3:3539  103 sin2
37 580 2200

The relation between the geocentric and geodetic d latitudes is given by


tan d

1
1  a 2

tan

Substituting the already obtained values for a and


d 38 090 3500
13

14

Gravity

jd

j
a

Fig. 1

The dynamic form factor J2 can be obtained from the equation


3
m
a J2
2
2
Then
J2

2
m
1:0878  103
a
3
2

From this value we can determine the gravity attening b using the equation
5
b m  a 5:2533  103
2
2. Taking the rst-order approximation, let two points of the ellipsoid at 45 N and 30 S
be situated at distances of 6367.444 km and 6372.790 km from the centre, respectively.
If the normal gravity values are 9.806193 m s2 for the rst and 9.793242 m s2 for
the second, calculate: the attening, gravity attening, coefcient m, equatorial radius,
polar radius, dynamic form factor, and the Earths mass.

Data
r1 6367.444 km
r2 6372.790 km

w1 = 45 N g1 = 9.806193 m s2


w2 = 30 S g 2 = 9.793242 m s2

The normal or theoretical gravity at a point can be expressed in terms of the normal gravity
at the equator ge, the gravity attening b, and the latitude of the point :




g1 ge 1 b sin2 1
g2 ge 1 b sin2 2

If we divide both expressions we obtain:

g1 1 b sin2 1

g2 1 b sin2 2

15

Terrestrial geoid and ellipsoid

From this expression we can obtain the gravity attening, since we already know g1, g2, 1, 2:
b

g1  g2
5:297  103
g2 sin2 1  g1 sin2 2

The distance r from the centre of the ellipsoid to points on its surface can be given as a
function of the attening a, the equatorial radius a, and the latitude :
r1 a1  a sin2 1
r2 a1  a sin2 2
If we divide both expressions
r1 1  asin2 1

r2 1  asin2 2
Thus we obtain the value of the attening,
a 3:353  103
From this value we nd the equatorial radius,
a

r1
6378:137 km
1  asin2 1

The polar radius c can be found from this value and the attening:
a

ac
a

and

c a1  a 6356:751 km

The coefcient m is obtained from a and b:


5
ab m
2

2
m a b 3:460  103
5

From this value we can obtain the value of the Earths mass M from
m

o 2 a3
GM

2p
p where T 24 hours.
T
Therefore

with o

4p2 a3
5:946  1024 kg
T 2 Gm

3. Obtain the value of the terrestrial attening in the rst-order approximation, given
that the normal gravity values for two points of the ellipsoid are:
Point 1: w1 42 200 g1 980.389 063 Gal
Point 2: w2 47 300 g2 980.854 830 Gal
Take the equatorial radius to be 6378.388 km.

16

Gravity

For this problem we use the equations of Problems 1 and 2:


g1 ge 1 bsin2 1
g2 ge 1 bsin2 2
If we divide these expressions:
g1 1 bsin2 1

g2 1 bsin2 2
and we can solve for the gravity attening, b:
b 5:288 2675  103
Using the gravity attening b we can determine the value of gravity at the equator, ge:
ge

g1
978:043 614 Gal
1 bsin2 1

Using the following equations


5
ab m
2
3
m
a J2
2
2
o2 a3
m
GM
we derive the expression
ge

GM
1  b o2 a
a2

and substituting the values we obtain GM 3.986 5415  1014 m3 s2


From this value, taking T 24 hours, we obtain
m

o 2 a3
4p2 a3
2
3:442 5698  103
GM
T GM

And nally the Earths attening is


5
a m  b 3:318 1575  103
2
4. P is a point of the terrestrial ellipsoid at latitude 60 S and distance to the centre of
6362.121 km. The Earths mass is 5.9761  1024 kg and the ratio between the polar
and equatorial semi-axes is 0.9966. Taking the rst-order approximation, calculate:
(a) The attening and the coefcient J2.
(b) The value of normal gravity in mGal at P.
As in the previous problems we use the equations given in Problems 1 and 2.

17

Terrestrial geoid and ellipsoid

(a) From the equation for the attening (Problem 2)


a1

c
3:4  103
a

The value of the equatorial radius a is obtained from the equation


a

r
6 378 386 m
1  asin2

The coefcient m is given by


m

o 2 a3
4p2 a3
2
3:4429  103
GM
T GM

The gravity attening b is given by


5
b m  a 5:2072  103
2
The dynamic form factor J2 is found from the relation
J2

2a  m
1:1190  103
3

(b) The normal gravity at that point is given by


g ge 1 bsin2 981 856:3 mGal
5. At a point P on the ellipsoid at latitude 50 S, the value of normal gravity is 9.810
752 m s2 and the distance to the centre of the Earth is 6365.587 km. Given that
the mass of the Earth is 5.976  1024 kg and the ratio between the minor and major
semi-axes is c/a 0.996 6509, calculate:
(a) The attening, equatorial radius, gravity attening, dynamic form factor, and
coefcient m.
(b) The normal gravity at the equator.
(c) The centrifugal force at P.

Data
50 S r 6365.587 km

g 9.810 752 m s2.

(a) According to Problem 1, the attening is given by


a1

c
3:349  103
a

and the equatorial radius a is


a

r
6378:122 km
1  asin2

18

Gravity

Taking T 24 hours, the coefcient m is then given by


4p2 a3
3:4425  103
T 2 GM

b and J2 can be obtained from the equations


5
b m  a 5:2571  103
2

J2

2
m
a
1:0852  103
3
2

(b) The normal gravity ge at the equator is given by


ge

g
9:780 579 m s2
1 bsin2

(c) The centrifugal force at point P is given by its radial and transverse components
f fr er fy ey
where, since 90  y = ,
fr o2 r sin2 y o2 r cos2 0:013 909 m s2
fy o2 r sin y cos y o2 r cos sin 0:016 576 m s2
6. Taking the rst-order approximation, calculate the Earths attening a, gravity
attening b, dynamic form factor J2, and polar radius c, given that:
ge 978.032 Gal (normal gravity at the equator)
a 6378.136 km (equatorial radius)
GM 39:8603  1013 m3 s2
and that for a point on the ellipsoid at latitude 60 N the normal gravity value is
981 921 mGal.
Calculate also the radius vector of this point and the gravitational potential.
Assuming a rst-order approximation, the expression for the normal gravity is
g ge 1 b sin2
The value of b is given by
b

g  ge
5:302  103
ge sin2

The coefcient m, taking T 24 hours, is


m

4p2 a3
3:442  103
T 2 GM

a and J2 are determined from the equations


5
a m  b 3:303  103
2

J2

2
m
a
1:055  103
3
2

19

Terrestrial geoid and ellipsoid

The polar radius c is determined from the attening


a

ac
) c a1  a 6357:069 km
a

The radius vector at the point of latitude 60 N is given by


r a1  asin2 6362:335 km
The gravity potential at the same point, in the rst-order approximation, is found using
Mac Cullaghs formula,


  r 3 m 2
GM
J2 a2 
2
1
cos 6:263 57  107 m2 s2
U
3 sin  1
r
a 2
2 r

7. Assuming that the Moon is an ellipsoid of equatorial radius 1738 km and polar
radius 1737 km, with J2 = 3.8195  104 and a mass of 7.3483  1022 kg, calculate its
period of rotation.
First, we calculate the lunar attening a,
a

ac
5:7537  104
a

The coefcient m is found from the values of a and J2:


3
m
a J2 ) m 2a  3J2 7:5900  106
2
2
From the value of m we nd the period of rotation T:
2 3
1
4p2 a3
4p a 2
)T
m 2
27:32 days
T GM
mGM
8. Calculate, in the rst-order approximation, the latitude and radius vector of a point
P of the terrestrial ellipsoid for which the value of normal gravity is 979.992 Gal, given
that the Earths mass is 5.976  1024 kg and that normal gravity for another point Q at
50 S latitude is 981.067 Gal, for the equator is 978.032 Gal, and that J2 is 1.083  103.
The gravity attening b is found from the normal gravity at Q with latitude 1 = 50 :
g1
1
ge
5:288  103
b 2
sin Q
Using the same expression and the normal gravity at P we calculate its latitude
gA
g 1
sin P e
;
b
2

P 37 590 46:5700

The Earths attening is given by


a

15
b
J2 3:353  103
8
4

20

Gravity

and the value of the coefcient m by


m 2a  3J2 3:457  103
From the value of m we obtain the equatorial radius a:

1
mGM 3
6 387 062:758 m
a
o2
where we have substituted o = 2p/T, taking T 24 hours.
Finally, we nd the radius vector of point P
r a1  asin2 6 378 946:678 m
9. At a point P on the terrestrial ellipsoid of latitude 70 S and radius vector 6359.253
km, the value of normal gravity is 982.609 Gal. If the mass of the Earth is 5.9769  1024
kg and the equatorial radius is 6378.136 km, calculate the value of normal gravity at
the Pole, the dynamic form factor, and the centrifugal force at the Pole and the equator.
The attening a is given by
a

1 
r
3:3528  103
1

a
sin2

Putting T 24 hours, the coefcient m is obtained from the equation


m

o 2 a3
3:4425  103
GM

From a and m we nd the gravity attening b:


5
b m  a 5:2535  103
2
Normal gravity at the equator is found from the value of the gravity at point P:
ge

g
9:780 72 m s2
1 b sin2

From this value we nd the normal gravity at the Pole:


gp ge 1 b 9:832 10 m s2
The dynamic form factor is found from the values of a and m:
2
m
a
1:0877  103
J2
3
2
The centrifugal force is given by the expression:

f o2 r sin2 yer o2 r sin y cos yey


At the pole, y = 90  = 0 ! f = 0.
At the equator, y 90 ! f = o2 a er = 0.033 73 er m s2

21

Terrestrial geoid and ellipsoid

10. Let two points of the ellipsoid be of latitudes w1 and w2, with radius vectors
6372.819 km and 6362.121 km, respectively. The ratio of the normal gravities is
0.997 37, the attening 3.3529  103, and the gravity attening 5.2884  103.
Calculate:
(a) The Earths mass.
(b) The latitude of each point and the dynamic form factor.
(a) The equatorial radius a can be obtained from the ratio of the two normal gravities
a  r 
1
2
1

b
g1 1 bsin 1
aa

l

g2 1 bsin2 2 1 b a  r2
aa

where l 0.997 37. We solve for a and obtain


a

blr2  r1
6382:94 km
a b l  1

We calculate the mass of the Earth from the attening and gravity attening as in Problem 2:
2
m a b 3:4565  103
5

4p2 a3
5:9653  1024 kg
T 2 Gm

(b) The latitudes at each point are calculated from the radius vectors
a  r1
sin2 1
! 1 43 270 5800
aa
a  r2
! 2 80 330 4600
sin2 2
aa
The dynamic form factor is obtained from the values of a, b, and m:
9
b 2a m  J2
2
2
J2 2a m  b 1:0831  103
9
11. Let a point A have a value of gravity of 9793 626.8 gu and a geopotential number
of 32.614 gpu. Calculate the gravity at a point B, knowing that the increments in
dynamic and Helmert height over point A are 271.116 m and 271.456 m, respectively.
Take g45 9.806 2940 m s2. Give the units for each parameter.
The dynamic heights at points A and B are given by:
CA
g45
C
B
HDB
g45
HDA

where C is the value of the geopotential at each point


gravity for a point on the ellipsoid at 45 latitude.

N
P

j1

gj dhj

and g45 the normal

22

Gravity

Subtracting both equations,


HDB  HDA

CB  CA
g45

Solving for CB,




CB CA g45 HDB  HDA 298:478 gpu

If heights are given in km and normal gravity in Gal, geopotentials are in gpu (geopotential units)
1 gpu 1 kGal m 1 Gal km
The Helmert orthometric height H is given by
H

C
g 0:0424H

11:1

where C is in gpu, g in Gal, and H in km.


Solving for H:
p
g  g2 4  0:0424C
H
2  0:0424
Since the point A is above the geoid (CA > 0), we take the positive solution,
HA 33:301 m
Then, the Helmert height at point B is
HB HA HAB 304:757 m
The gravity at point B is calculated using Equation (11.1)
gB

CB
 0:0424HB 979:382 75 Gal
HB

12. Calculate the value of gravity in gravimetric units and mGal of a point on the
Earths surface whose orthometric (Helmert) and dynamic heights are 678.612 m and
679.919 m, respectively, taking g 45 9.806 294 m s2.
The geopotential is calculated from the dynamic height,
HD

C
) C HD g45 666:748 gpu
g45

where HD is given in km and g45 in Gal.


Knowing the geopotential, we calculate the gravity from the orthometric (Helmert)
height H, using its denition,
H

C
C
) g  0:0424H 982:489 34 Gal 9 824 893:4 gu
g 0:0424H
H

13. If at a point on the surface of the Earth of Helmert height 1000 m one observes
a value of gravity of 9.796 235 m s2, calculate the average value of gravity

23

Terrestrial geoid and ellipsoid

between that point and the geoid along the direction of the plumb-line, and
the points geopotential number.
The mean value of gravity between a height H and the surface of the geoid is given by

1 H
g
gzdz
H 0
where g(z) is the value of gravity at a distance z from the geoid along the vertical path to a
point of height H. This value can be obtained using the Poincar and Prey reduction from
the value of g observed at the Earths surface at a point of height H,
gz g 0:0848H  z
Then

1 H
1 H
gzdz
g 0:0848H  zdz
g
H 0
H 0
H
1
gz 0:0848 Hz  0:0424z2 0

H
g g 0:0424 H 979:6659 Gal
where g is given in Gal and H in km.
The geopotential C can be obtained from the formula for the Helmert height,
H

C
) C g 0:0424 H H 979:666 gpu
g 0:0424H

14. For two points A and B belonging to a gravity measurement levelling line, one
obtained:
gA = 9.801 137 6 m s2
CA = 933.316 gpu
Gross increment elevation: DhBA 20:340
B
Increment in dynamic height: H A
D  H D 20:340 m.
Given that the normal gravity at 45 latitude is 9806 294 gu, calculate the Helmert
heigth of point B.
As in Problem 11, the dynamic heights at A and B are given by
CA
g45
CB
HDB
g45

HDA

Subtracting both equations:


HDB  HDA

CB  CA
g45

Solving for CB:




CB CA g45 HDB  HDA 913:371 gpu

24

Gravity

The geopotential at B can be obtained from the gross increment in elevation between A and B,
g g 
A
B
CB CA
hBA
2

and, solving for gB,

gB

2CB  CA
 gA 980:103 08 Gal
hBA

Finally we calculate the orthometric Helmert height at point B,


H

C
g 0:0424H

Substituting the values for point B, and solving for H we obtain


p
gB  gB 2 4  0:0424CB
H
931:875 m
2  0:0424
15. A, B, and C are points connected by a geometric levelling line. Given that the
normal gravity at a latitude of 45 is 980.6294 Gal, complete the following table:

Station

Gravity
(Gal)

Height
Increment (m)

Geopotential
Number (gpu)

Dynamic
Height (m)

Helmert
Height (m)

A
B
C

979.88696
?
979.88665

0.541
?

664.982
?
?

?
677.577
?

?
?
657.134

Station A
Dynamic height:
HDA

CA
678:118 m
g45

Helmert height:
C
g 0:0424H
p
g  g2 4  0:0424  C
H
678:611 m
2  0:0424
H

Station B
The geopotential number is found from the dynamic height:
CB g45 HDB 664:452 gpu

15:1

Earths gravity eld and potential

25

From this value and the difference in height with respect to station A we nd the gravity at B:
g g 
A
B
CB CA
hA
B
2
from which we get gB = 979.877 84 Gal
The Helmert height of B is found as in station A:

HB 678:077 m

Station C
From the known values of gravity and Helmert height we nd the geopotential number
(Equation 15.1)
CC gHC 0:0424HC2 661:574 gpu
To calculate the difference in height of C with respect to B we begin with the expression
g g 
B
C
hCB
CC CB
2
from which

hCB

2CC  CB
2:937 m
gB gC

The dynamic height is found directly from the geopotential number:


HDC

CC
674:642 m
g45

The complete table is:

Station

Gravity
(Gal)

Height
increment (m)

Geopotential
number (gpu)

Dynamic
height (m)

Helmert
height (m)

A
B
C

979.88696
979.87784
979.88665

0.541
2.937

664.982
664.452
661.574

678.118
677.577
674.642

678.611
678.077
657.134

Earths gravity eld and potential


16. Suppose an Earth is formed by a sphere of radius a and density r, and within it
there are two spheres of radius a/2 with centres located on the axis of rotation. The
density of that of the northern hemisphere is 5r and of that of the southern hemisphere is r /5. The value of the rotation is such that m 0.1. Determine:
(a) The potential U in the r3 approximation.
(b) The values of gr and gu for a point on the equator in the r2 approximation.

26

Gravity

(c) The error made in (b) with respect to the exact solution.
(d) The deviation of the vertical from the radial at the equator.
(a) The gravitational potential is the sum of the potentials of the three spheres
V V1 V2 V3

GM GM1 GM2

0
r
q
q

16:1

where r is the distance from a point P to the centre of the sphere of radius a and mass M,
where M is given by
4
M pra3
3
q and q0 are the distances to the centres of the two spheres in its interior in the northern and
southern hemispheres which have differential masses M1 and M2, respectively (Fig. 16).
The differential masses are those corresponding to the difference in density in each case
with respect to the large sphere:
4
a3 M
M1 p5r  r
3
8
2
differential mass of the sphere in the northern hemisphere
 a3
4 r
M

M2 p  r
3 5
8
10

differential mass of the sphere in the southern hemisphere


The distance q can be calculated using the cosine law

r
r
a2
 a 2
a
a
2
q r
cos y
 2 r cos y r 1
2
2
2
2r
2r
Considering this expression, 1/q corresponds to one of the generating functions of the
Legendre polynomials. Then 1/q, in the rst-order approximation, is given by



1 1
a
1  a 2 
2
1 cos y
3cos y  1
q r
2r
2 2r

Since cos y0 =  cos y, 1/q0 is given by





1 1
a
1  a 2 
2

1

cos
y

3cos
y

1
q0 r
2r
2 2r
P
5

q
r

r2

a/2

a
r3

/5

Fig. 16

q'
'

Earths gravity eld and potential

27

If we substitute in Equation (16.1), the potentials for each sphere are given by


GM
GM 1
a
a2 
V2
2 cos y 3 3cos2 y  1
V1
r
2
r 2r
8r


GM 1
a
a2 
2
V3 

cos y 3 3cos y  1
10 r 2r2
8r

Then, the total gravity potential is the sum of the three gravitational potentials plus the
potential of the centrifugal force due to the rotation:


1 1 1
1 1 a
U GM 1 

cos y
2 10 r
4 20 r2




1
1 a2 
1 2 2 2
2


3 cos y  1 r o sin y
16 80 r3
2
In terms of the coefcient m, given here by m

o2 a3
,
GM



 1  r 2 2
GM 7 a 3 a2
1 a3 
2

cos y
3cos y  1 m
sin y
U
a 5 r 10 r2
20 r3
2 a
(b) Using this rst-order approximation of the potential, the radial and tangential
components of gravity at the equator, r = a and y = 90 , putting m 0.1, are
@U
GM
1:3 2
@r
a
1 @U
GM
0:3 2
gy
r @y
a

gr

(c) To calculate exactly the value of gr at the equator we have to calculate


the exact contribution of each of the three spheres plus the centrifugal force (m 0.1):
gr gr1 gr2 gr3  m

GM
a2

GM
a2
GM
gr2  2 cos a
2r2
GM
cos a
gr3 
10r32

gr1 

where
r
a2
r 2 a2 r 3
4
and a is the angle which forms r2 and r3 with the equator (Fig. 16)

28

Gravity

sin a
Then
GM
gr  2
a

a=2
1
p
r2
5

4
4
GM
1 p  p  m 1:19 2
a
5 5 25 5
0
1

GM 1 C
B GM 1
p
pA
gy gy1 gy2 @
5 2 5
5
2 a
10 a2 5
4

4
GM
2
2
GM
2  p p 0:14 2
a
a
5 5 25 5

The error made in the rst-order approximation with respect to the exact solution is
GM
GM
0:11 2
a2
a
GM
GM
gy 0:3 0:14 2 0:16 2
a
a
gr 1:19 1:3

(d) The deviation of the vertical with respect to the radial direction is given by the angle i
which is determined from the gravity components gr and gy. At the equator this angle is:
Using the rst order approximation
tan i

gy 0:3
) i 13:0

gr 1:3

Using the exact values


tan i

0:16
) i 7:6
1:19

17. A spherical planet is formed by a sphere of radius a and density r, and inside it a
sphere of radius a/2 and density 5r centred at the midpoint of the radius of the
northern hemisphere. There is no rotation.
(a) Determine J0, J1, and J2.
(b) What is the deviation of the vertical from the radial at the equator?
(a) The total gravitational potential is the sum of the potentials of the two spheres
(Fig. 17) where g, is the attraction due to the potential V1 and g2 that due to the
potential V2:
V V1 V2

GM GM 0

r
q

where r and q are the distances from a point P to the centres of the large and small spheres,
respectively.

Earths gravity eld and potential

29

5r
r
a /2
q

g2
i

g1

Fig. 17

As we did in Problem 16, for the small sphere of radius a/2 we take the differential
mass M 0
4 a3
16
M
M 0 p 5r  r pra3
3 8
24
2
where M is the mass of the sphere of radius a and density r.
For 1/q we take the rst-order approximation of the Legendre polynomial, as we did in
Problem 16:


1 1
a
1  a 2 
2

1 cos y
3cos y  1
q r
2r
2 2r
Then, the expression for the gravitational potential V is:


GM GM 1
a
1 a2 
2

2 cos y
3cos
y

1
V
r
2
r 2r
4 4r3


3
a
a2 
2
GM

cos y
3cos y  1
2r 4r2
32r3

(b) We know that the potential can be expressed by an expansion in zonal


spherical harmonics (Legendre polynomials) given in the rst-order approximation by
   

a3 1 

GM
a
a 2
J0

J1 cos y
J2 3cos2 y  1
V
a
r
r
r
2
Comparing the two expressions we obtain,

30

Gravity

3
2
1
J1
4
1
J2
16
J0

The components of gravity at the surface of the large sphere (r a) are:



@V GM
3 1
3 
2   cos y 
3cos2 y  1
@r
a
2 2
32

1 @V GM
1
3
gy
2  sin y  cos y sin y
r @y
a
4
16

gr

and at the equator, y 90 :


45GM
32a2
GM
gy  2
4a

gr 

At the equator the deviation of the vertical with respect to the radial direction is
gy
8

gr 45
i 10:08

tan i

18. Suppose an Earth is formed by a sphere of radius a and density r, and within it
there are two spheres of radius a/2 and density 2r with centres located on the axis of
rotation in each hemisphere. If M is the mass of the sphere of radius a, calculate:
(a) The potential U(r,u) and the form of the equipotential surface passing through the Poles.
(b) The component gr of gravity in the rst-order approximation for points on the surface.
(c) Calculate gr directly at the Pole and the equator, and compare with the rst-order
approximation.
(a) This problem is similar to Problem 16, but now the density of the two spheres is the
same. The total gravity potential is the sum of the gravitational potentials of the
three spheres (V, V1 and V2) plus the potential due to the rotation F:
where

U V V1 V2 F
GM
r
GM 0
V1
q1
GM 0
V2
q2
1 2 2 2
F o r sin y
2
V

18:1

Earths gravity eld and potential

31

q1

2r

r
r

q2
a/2

a
a

2r

Fig. 18

and where M is the mass of the large sphere of radius a and M 0 the differential mass of each
of the small spheres of radius a/2, r is the distance from a point P to the centre of the large
sphere, and q1 and q2 the distances from P to the centres of the small spheres (Fig. 18). As
in Problem 16 the differential mass is given by the difference in density between the large
and the small spheres:
 a 3 M
4
M 0 p2r  r

3
2
8
The inverse of the distance 1/q can be approximated by


 a 2 1 

1
1
a
2
1 cos y
3cos y  1
q1 r
2r
2r 2

and since cos y0 =  cos y



 a 2 1 

1
1
a
2
3cos y  1
1  cos y
q2 r
2r
2r 2

The potential of the rotation can be written in terms of the coefcient m = a3o2/GM,

GM 1 1 a3 o2 3
GM m  r 3 2
r
sin2 y
sin y
F
3
r
2 a GM
r 2 a
Substituting in Equation (18.1)


  r 3 m 2
GM 10 1  a 2 
2

sin y
U
3cos y  1
r
8 8 2r
a 2

18:2

32

Gravity
At the Poles, r a and y = 0 , and the potential is


GM 10 2
21 GM
U poles

a
8 32
16 a
The form of the equipotential surface which passes through the Poles (r a) is obtained
from Equation (18.2)
r



  r 3 m 2
GM 10 1  a 2 

3cos2 y  1
sin y
Upoles 8 8 2r
a 2

Making the approximation r a in the right-hand side:


r




32 10 m 2
1 
a
sin y
3cos2 y  1
42 8
2
32

and substituting cos2 y = 1  sin2 y, we obtain





32 3 m
2

r a 1
sin y
42 32 2

This is the equation of an ellipse with attening a (32/42)(3/32m/2). Since there is


symmetry with respect to the axis of rotation, the equipotential surface is an ellipsoid of
revolution.


At the poles: y 0 ) rp a. 
32 m 3
At the equator: y 90 ) re a 1

42 2 32
Depending on the value of m, we have the following cases,
m
3

) re a ) sphere
2 32
m
3
<
) re < a ) prolate ellipsoid
2
32
m
3
>
) re > a ) oblate ellipsoid
2
32

3
< a ) prolate ellipsoid
m0)r a 1
42
(b) For the gravity at the Pole, in the rst-order approximation, we take the derivative
of the potential (18.2) and substitute y 0 ) rp a:


@U GM
10 6
GM
gr
2  
1:4375 2
@r
a
8 32
a
(c) The exact solution for the gravity at the pole is the sum of the attractions of the
three spheres:
gr 

GM GM GM
GM
 2
1:5555 2
a2
2a
18a2
a

Earths gravity eld and potential

33

At the equator we take the derivative of the potential and substitute r = a and
y = 90 :


@U GM
10 3
gr
2  m
@r
a
8 32


GM 37
GM
 m  2 1:1562  m
 2
a 32
a
For the exact solution we write
gr 

GM 2GM

cos a o2 a
a2
8q2

From Fig. 18 the distance q is given by


a2
5
a2 a2
4
4
r
4
cos a
5
q2

Therefore
"
#
r
GM
8 4
GM
gr  2 1
 m  2 1:1789  m
a
40 5
a
The approximated values are smaller than the exact solutions.
19. For the case of Problem 18, if GM = 4  103 m3 s2, a = 6  103 km, and v = 7 
105 s1, calculate the values of J2, a, m, H, and b.
From the denition of m we obtain
m

a3 o2 216  1018  49  1010

2:6  103
GM
4  1014

The value of J2 is obtained by comparing the two expressions for the potential U
(Problem 18):


a2 1 
 m  r 3 2
GM
U
1
J2 3cos2 y  1
sin y
r
r
2
2 a




2


5 GM
1 a
m  r 3 2
3cos2 y  1
sin y
1
U
4 r
40 r
2 a
Then, J2 = 0.05.
The attening is obtained from the relation
3
m
a J2
2
2
a

3
2:3  103
 50  103
0:0765
2
2

34

Gravity

X3

2r

r
a/2

X1

a
2r

Fig. 19a

The gravity attening is given by


b

gp  ge 1:555  1:175

0:323
ge
1:175

where we have used the values of gravity at the Pole and equator obtained in Problem 18,
and in the latter we have substituted the value obtained for the coefcient m.
The dynamic ellipticity H is dened as the ratio of the moments of inertia with respect to
the polar and equatorial radius (Fig. 19a):
H

CA
C

where A and C are the moments of inertia of a sphere respect to the polar and equatorial
radi: (axes x1 and x3). The moment of inertia of a sphere of radius R is
2
Isph MR2
5
We have to add to the moment of inertia of the sphere of radius a the moments of inertia of
the two internal spheres of radius a/2. For the C-axis (x3) we have
2
2 M a2
0:425Ma2
IC Isph a 2Isph a2 Ma2 2
5
58 4
For the A-axis (x1) the moment of inertia of each of the small spheres is given by
(Fig. 19b)

Earths gravity eld and potential

35

R
h

Fig. 19b

I ICM Mh2
since in this case the A axis does not coincide with the centre of mass, where R a/2
and h a/2:
Isph a=2

2 M a2 M a 2

0:044Ma2
58 4
8 4

2
IA Isph a 2Isph a=2 Ma2 2  0:044Ma2 0:488Ma2
5
Finally
H

C  A IC  IA 0:425  0:488

0:147

C
IC
0:425

20. Suppose an Earth is formed by a sphere of radius a and density r, and within it
there is a sphere of radius a/2 and density 5r centred at the midpoint of the northernhemisphere polar radius. If m 1/8 and M is the mass of the sphere of radius a,
determine:
(a) The form of the equipotential surface passing through the North Pole.
(b) For latitude 45, the astronomical latitude and the deviation of the vertical from
the radial.
(a) The gravitational potential is the sum of the potentials for the sphere of radius a
and that of the sphere of radius a/2 (Fig. 20):
V V1 V2

GM GM1

r
q

As in the previous problems the potential of the small sphere is given in terms of
differential mass M1:

36

Gravity

gq

a/2

q
g

5r

gr

fa

Fig. 20

a3 M
4

M1 p5r  r
3
2
2

and for the inverse of the distance 1/q we use the approximation


 a 2 1 

1 1
a
2
1 cos y
3cos y  1
q r
2r
2r 2

Then, the total gravitational potential is





GM 3 a
1 a2 
V
cos y
3cos2 y  1
r 2 4r
16 r

The total potential U is the sum of the gravitational potential V plus the potential of rotation
F, where
1
F r2 o2 sin2 y
2
and using the coefcient m = o2a3/GM = 1/8, we have


  r 3 m 2
GM 3 a
1 a2 
U
cos y
3cos2 y  1
sin y
r 2 4r
16 r
a 2
At the North Pole, y = 0 and r a, and the value of the potential is


GM 3 1 1
15GM
Up

a 2 4 8
8a

Earths gravity eld and potential

37

The form of the equipotential surface is found by putting U Up:




  r 3 m 2
15GM GM 3 a
1 a2 

cos y
3cos2 y  1
sin y
8a
r 2 4r
16 r
a 2

Putting r a inside the square brackets and solving for r we nd




4
1
1
r a 1 cos y cos2 y
5
6
12

(b) The deviation of the vertical with respect to the radial direction is given by the angle i:
tan i

gy
gr

To nd this value we have to calculate the two components of gravity





@U
3 1 2a 1
3a2 1
2r
2
2
GM  2 
gr
cos
y

3
cos
y

1

sin
y
@r
2r
4 r3
16a3
16 r4


1 @U GM
a
a2
r2

 2 sin y 
6
cos
y
sin
y

sin
y
cos
y
gy
r @y
r
4r
16r3
8a3
For a point on the surface we put r a:


GM 19 1
11 2
gr  2
cos y cos y
a 16 2
16


GM 1
1
sin y sin y cos y
gy  2
a 4
4
and for latitude 45
GM
1:88
a2
GM
gy  2 0:30
a
0:30
) i 9:0.
Then the angle i is given by tan i
1:88
The astronomical latitude is
gr 

fa 90  y  i 36:0
21. If the internal sphere of Problem 20 is located on the equatorial radius at
longitude zero, nd expressions for the components of gravity: gr , gu , gl.
As in the previous problem the differential mass of the small sphere M1 is (Fig. 21a):
M1

M
2

The total potential U is the sum of the gravitational potentials V and V1, and the potential
due to rotation F. According to Fig. 21b, using the relations of spherical triangles, if and
l are the coordinates of the point where the potential is evaluated, then

38

Gravity

P
q

y
j

l = 0 A

a/2

Fig. 21a
B
90

90 j
P

Fig. 21b

cosc cos 90 cos(90) sin 90 sin(90) cosl


cosc cos cosl
Using the expression for 1/q as in Problem 16,


 a 2 1 

1 1
a
1 cos c
3cos2 c  1
q r
2r
2r 2

The gravitational potential of the small sphere is given by





GM
a
a2 
2
2
1 cos cos l 2 3cos cos l  1
V1
2r
2r
8r

The total potential U is given by:




 1 2 2 2
GM 3 a
a2 
2
2
U
cos cos l
r
o
cos

3cos
cos
l

1

r 2 4r
2
16r2

Earths gravity eld and potential

39

Using the coefcient m and sin y = cos , we obtain




 r2 m 2
3
a2
a2 
2
2
U GM

sin y cos l
3sin ycos l  1 3 sin y
2r 4r2
16r3
a 2

The three components of gravity are found by differentiating U with respect to r, y, and l:


 r
@U
3
a2
3a2 
2
2
2
gr
GM  2  3 sin y cos l 
3sin
ycos
l

1

msin
y
@r
2r
2r
16r4
a3
 2

1 @U
a
a2
r
cos
y
cos
l

6
cos
l
sin
y
cos
y

m
sin
y
cos
y
GM
gy
4r3
16r4
r @y
a3


1 @U
a2
3a2
gl
GM  3 sin l  4 sin y cos l sin l
r sin y @l
4r
8r
22. A planet is formed by a sphere of radius a and density r, with a spherical core of
density 5r and radius a/2 centred on the axis of rotation in the northern hemisphere
and tangential to the equator. The planet rotates with m 1/4. For the point at
coordinates (45 N, 45 E), calculate:
(a) The astronomical latitude.
(b) The deviation of the vertical from the radial.
(c) The angular velocity of rotation that would be required for this deviation to be zero.
(a) The gravitational potential is the sum of the potentials of the two spheres (Fig. 22):
V V1 V2

GM GM 0

r
q

22:1

5r

g
gr
r
j

Fig. 22

gq

a/2

40

Gravity

As in Problem 16, the inverse of the distance from a point P to the centre of the small
sphere, 1/q, can be approximated by



1 1
a
1  a 2 
1 cos y
3cos2 y  1
q r
2r
2 2r

As in previous problems we use the differential mass of the small sphere,


4 a3
2
M 0 p 5r  r pra3
3 8
3
4
and since M pra3 , then M 0 = M/2.
3
Substituting in Equation (22.1) we obtain



GM GM 1
a
1 a2 
2
V

2 cos y
3cos
y

1
r
2 r 2r
2 4r3



3
a
a2 
2
2 cos y
3cos
y

1
GM
2r 4r
16r3

The total potential U is the sum of V plus the potential due to rotation F:



GM 3 a
a2 
1
2
cos y
3cos y  1 r2 o2 sin2 y
U V F
r 2 4r
2
16r2

The components of gravity gr and gy are





@U
3
a
3a2 
2
GM  2  3 cos y 
3cos
y

1
ro2 sin2 y
gr
@r
2r
2r
16r4


1 @U
a
a2
gy
6 cos y sin y ro2 sin y cos y
GM  3 sin y 
r @y
4r
16r4

22:2
22:3

For a point on the surface of the large sphere and coordinates 45 N, 45 E, we have that
y = 45 , r a, and
a3 o 2 1

GM
4
Putting these values in (22.2) and (22.3), we obtain
m

gr 1:82

GM
GM
and gy 0:24 2
a2
a

To nd the astronomical latitude we rst have to nd the deviation of the vertical with
respect to the radial:
tan i

gy
0:13 ; i 7:5
gr

The astronomical latitude is, then, f =  i = 45  7.5 = 37.5 .


(b) The deviation of the vertical with respect to the radial direction, as already found, is i = 7.5 .
(c) If we want the deviation of the vertical to be null, i 0, this implies gy = 0.

Earths gravity eld and potential

41

writing Equation (22.3) in terms of the coefcient m, where


m
we have
GM
gy 0 2
a

a3 o2
GM

22:4

2 23 m

16
2

and solving for m gives m 0.73.


Substituting in Equation (22.4) we obtain
r
GM
0:85
o
a3

23. A planet consists of a very thin spherical shell of mass M and radius a, within
which is a solid sphere of radius a/2 and mass M 0 centred at the midpoint of the
equatorial radius of the zero meridian. The planet rotates with angular velocity v
about an axis normal to the equatorial plane. Calculate:
(a) The potential at points on the surface as a function of latitude and longitude.
(b) The components of the gravity vector.
(c) If M 0 10 M, what is the ratio between the tangential and radial components of
gravity at the North Pole?
(a) The potential U is the sum of the gravitational potentials due to the spherical shell V1, and
to the interior sphere V2, plus the potential due to the rotation of the planet F (Fig. 23):
U V1 V2 F
1
F o2 r2 cos2
2
GM
V1
r
GM 0
V2
q

23:1

where r is the distance from a point P on the surface of the planet to its centre, q is the
distance from point P to the centre of the interior sphere, and the latitude of point P.
Using the cosine law,
r
a2
q r2  ar cos c
4
where c is the angle between r and the equatorial radius, and its inverse can be approximated by (Problem 16)



1 1
a
a2 
2
1 cos c 2 3cos c  1
23:2
q r
2r
8r
Using the relation for spherical triangles
cos a cos b cos c sin b sin c cos A

42

Gravity

j
90
P
r
y
a

y
q

j
a/2

Fig. 23

putting b 90 , c 90, A l, and a c, where l is the longitude of P, then


cos c = cos cos l
Substituting in (23.1), the potential due to the small sphere is



GM 0
a
a2 
2
2
1 cos cos l 2 3cos cos l  1
V2 
r
2r
8r

The total potential U is




GM GM 0
a
a2 
2
2

1 cos cos l 2 3cos cos l  1


U
r
r
2r
8r
1
o2 r2 cos2
2
(b) The components of the gravity vector are obtained from Equation (23.3):
@U
@r



GM
1
a
3a2 
0
2
2
 2 GM  2  3 cos cos l  4 3cos cos l  1
r
r
r
8r

gr

o2 rcos2
1 @U
1 @U
gy

r @y
r @


0
GM a
a2
2
sin

cos
l

6
cos
cos
l
sin

r 2r2
8r3
o2 r cos sin

23:3

Earths gravity eld and potential

43



1 @U
GM 0
a
a2
2
gl
 2 cos sin l  3 6cos cos l sin l

8r
r cos @l r cos
2r
(c) At the North Pole, = 90 and r a. Putting M 0 10M and substituting in the
previous equations we obtain
gr 

GM GM 0 3GM 0 a2
GM
 2
7:25 2
a2
a
8a4
a
gy

GM 0
5GM
 2
2a2
a

The ratio between the radial and the tangential components of gravity at the North
Pole is
gr
1:45
gy
24. An Earth consists of a sphere of radius a and density r, within which there are two
spheres of radius a/2 centred on the axis of rotation and tangent to each other. The
density of that of the northern hemisphere is 4r and that of the southern hemisphere
is r/4.
(a) Express the gravitational potential in terms of M (the mass of the large sphere) up
to terms of 1/r3.
(b) What astronomical latitude corresponds to points on the equator (without
rotation)?
(c) What error is made by using the 1/r3 approximation in calculating the value of gr
at the equator?
(a) The total gravitational potential V is the sum of the potentials of the sphere of
radius a (V0) and of the two spheres of radius a/2 situated in the northern (V1) and
southern (V2) hemispheres (Fig. 24):
V V0 V1 V2
As in previous problems the large sphere is considered to have uniform density
r and the effect of the two interior spheres is calculated using their differential
masses

The potentials are

4
M pra3
3
4
a3 3M
M1 p4r  r
3
8
8
 a3
4 r
3M

M2 p  1
3 4
32
8

44

Gravity

4r

r1
a/2
a
a
r2

r/4

Fig. 24

GM
r



GM1 3GM 1
a
a2 

2 cos y 3 3cos2 y  1
V1
r1
8
r 2r
8r



GM2
3GM 1
a
a2 
V2
 2 cos y 3 3cos2 y  1

r2
32 r 2r
8r
V0

where r1 and r2 have been calculated as in Problem 16. Then, the total gravitational
potential in the 1/r3 approximation is

V GM


41 15 a
9 a2 

cos y
3cos2 y  1
2
3
32r 64 r
256 r

(b) The components of the gravity vector, taking into account that there is no rotation, are



@V
41
15a
27a2 
2
gr
GM 

cos
y

3cos
y

1
@r
32r2 32r3
256r4
24:1


1 @V
15a
27a2
gy
GM 
sin
y

cos
y
sin
y
r @y
64r3
128r3
At the equator, r a and y = 90 and we obtain

Earths gravity eld and potential

45

GM
a2
GM
gy 0:243 2
a
gr 1:175

The astronomical latitude (a) is the angle between the vertical and the equatorial plane. In
our case at the equator this is given by the deviation of the vertical from the radial
direction:
tan a

gy
0:207
gr

Then
a 11:68 N
(c) If we want to calculate the exact value of gr at the equator, we calculate the exact
attractions of each sphere and add them:
gr0 

GM
a2

gr1 

3GM
cos a
8r12

gr2 

3GM
cos a
32r22

24:2

where r1 and r2 are the distances from the centre of each of the two interior
spheres (Fig. 24):
r
p
a2 a 5
r 1 r 2 a2
4
2
and a is the angle which r1 and r2 form with the equatorial plane:
sin a

a=2
1
p
r1
5

The radial component of gravity is given by

grT gr0 gr1 gr2 1:335

GM
a2

The error we make using the approximation is


gapprox  gexact 0:160

GM
;
a2

that is; 16%:

25. An Earth consists of a sphere of radius a and density r within which there are
two spheres of radius a/2 centred on the axis of rotation and tangent to each other.
The density of that of the northern hemisphere is 2r and that of the southern
hemisphere is r/2.
(a) Express the potential V in terms of M (the mass of the large sphere), G, and r up to
terms in 1/r3.

46

Gravity

q1
a/2

q2

2r
x

gq

gr

r
q
ja

r/2

Fig. 25

(b) According to the value of this potential V, which astronomical latitudes correspond to the geocentric latitudes 45 N and 45 S?
(c) What must the rotation period be for the astronomical and geocentric latitudes to
coincide?
(d) What error is made by the 1/r3 approximation in calculating the value of gr at the
equator? And at the North Pole?
(a) As in previous problems the effect of the interior spheres is given in terms of their
differential masses (Fig. 25):
a3 M
4

M1 p2r  r
3
2
8




4 r
a 3
M
M2 p  r

3 2
2
16
The distances q1 and q2 from the centre of each sphere to an arbitrary point P are found
using the cosine law:
a2
ar
 2 cos y
4
2
2
a
ar
q2 r2 2 cos y
4
2
q1 r 2

Earths gravity eld and potential

47

Using the approximation for 1/q (Problem 16), the total gravitational potential V is the sum
of the potentials of the three spheres:
GM GM1 GM2

r
q1
q2



GM 17 3 a
1 a2 
2

cos y
3cos y  1

r 16 32 r
128 r

(b) The components of the gravity vector are given by





@V
1 17 6 a
3 a2 
2
gr
cos y
3cos y  1
GM 2
@r
r 16 32 r3
128 r4


1 @V
GM 3 a
1 a2

sin
y

6
cos
y
sin
y
gy
r @y
r 32 r2
128 r3

25:1

If the point P is at the surface, r a, then





GM 17 3
3 
cos y
3cos2 y  1
gr  2
a 16 16
128


GM 3
6
gy  2
sin y
cos y sin y
a 32
128
At geocentric latitude 45 N, y 45,
GM
a2
GM
gy 0:09 2
a

gr 1:21

The deviation of the vertical with respect to the radial direction i is given by
tan i

gy
0:074 ) i45 4:2
gr

According to Fig. 25, the astronomical latitude a can be determined from the deviation of
the vertical i,
a i 180  180 ) a  i 45  4:2 40:8 N
In the same way, for geocentric latitude 45 S (y = 135 )
GM
a2
GM
gy 0:04 2
a

gr 0:94

gy
Then tan i135 0:043 ) i135 2:5
gr
Then the astronomical latitude is
a 45  2:5 47:5 47:5 S

48

Gravity

(c) If we want the astronomical and geocentric latitudes to coincide, then the
deviation of the vertical must be null, i 0. This implies that gytotal must be
zero. To do this by means of the rotation, we have to make the tangential
component of the centrifugal force gyR be equal and of opposite sign to that of
the gravitational potential gyV :
gy total gy V gy R 0 ) gy V gy R
The tangential component due to rotation is
gyR

1 @F
r @y

1
where F o2 r2 sin2 y. Then
2
gyR o2 r cos y sin y
For a point on the surface at latitude 45 N, r = a and y = 45 , so
gy V gy R ) 0:09 GM =a2 o2 a=2
From here we can calculate the period of rotation
2p
2p
p
T
o
0:18

r
a3
GM

For a point at latitude 45 S, r = a and y = 135 , so


gyR

gyV

1
GM
2p
)  o2 a 0:04 2 ) T p
2
a
0:08

r
a3
GM

(d) The value of the radial component of gravity at the equator, r a, y = 90 , by
substitution in (25.1), is
gr 1:04

GM
a2

If we calculate the exact value by adding the contributions of the three spheres
(Fig. 25)
grexact grM gr1 gr2
grM 

GM
a2

gr1 g1 cos b


gr2 g2 cos b
where

Earths gravity eld and potential

49

a
2
cos b r p
5
5
a
4
GM 4
g1 
8 5a2
GM 4
g2
16 5a2
grexact 

GM
0:96
a2

The error in the approximation is:


grerror

GM
GM
GM
0:96 2  1:04 2 0:08 2
a
a
a

In a similar way, for a point at the North Pole, r a, y 0:


gr 1:30 GM =a2
grexact grM gr1 gr2
GM
a2
GM
GM
gr1   2  2
a
2a
8
2
GM
GM
gr2 

a2 36a2
16 a
2

grM 

grexact 

GM
1:47
a2

The error in the approximation is


grerror

GM
GM
GM
1:47 2  1:30 2 0:17 2
a
a
a

26. A spherical Earth of radius a has a core of radius a/2 whose centre is displaced a/2 along
the axis of rotation towards the North Pole. The core density is twice that of the mantle.
(a) What should the period of rotation of the Earth be for the direction of the plumbline to coincide with the radius at a latitude of 45 S?
(b) What are the values of J0, J1, J2, and m?
(a) As in previous problems we calculate the gravitational potential by the sum of the
potentials of the two spheres, using for the core the differential mass (Fig. 26):
GM GM 0
V V1 V2

r
q
2
4
a
M
M 0 p2r  r
3
8
8

50

Gravity

a/2

2r

q
r

a
45

i
gr

gq

Fig. 26

As we saw in Problem 16, we use for 1/q the rst-order approximation





GM GM 1
a
1 a2 
2
V

2 cos y
3cos
y

1
r
8 r 2r
2 4r3

26:1

The total potential U is the sum of the gravitational potential V and the potential due to
rotation
1
F r2 o2 sin2 y
2
U GM




91
a
a2 
1
2

cos
y

3cos
y

1
r2 o2 sin2 y
8 r 16r2
64r3
2

In order that the direction of the plumb-line coincides with the radial direction, the
tangential component of gravity, gy, must be null:

1 @U GM
a 1
a2 1
gy
sin
y

6
cos
y
sin
y


r @y
r
16 r2
64 r3
ro2 sin y cos y

For a point on the surface at latitude 45 S, the tangential component of gravity is, with
r = a, y = 135 ,
gy 0:003

GM ao2

a2
2

Earths gravity eld and potential

51

Putting this component equal to zero, we nd the value of the period of rotation T:
r
GM ao2
GM
2p
a3
2
gy 0:003 2 
0 ) o 0:006 3 ) T
a
2
a
0:077 GM
(b) The gravitational potential V of Equation (26.1) can be written as



GM 9
a
a2 
2
V
1
cos y
3cos y  1
r 8
18r
72r2

We obtain the values of J1 and J2 by comparison with the equation



a2 
GM
a
V
J0 J1 P1 J2
P2
r
r
r
Since the total mass is (9/8)M, we obtain
J0 1
1
J1
18
1
J2
72

27. Within a spherical planet of radius a and density r there are two spherical cores of
radius a/2 and density r 0 with centres located on the axis of rotation at a/2 from the
planets centre, one in the northern hemisphere and the other in the southern hemisphere.
(a) Neglecting rotation of the planet, calculate what the ratio r0 /r should be for the
gravity attening to be 1/8.
(b) If the planet rotates so that m 1/16, and the ratio of the densities is that found in
part (a), calculate the astronomical latitude which corresponds to the geocentric
latitude 45 N.
(a) Since there is no rotation the total potential U is the sum of the gravitational
potentials of the three spheres (Fig. 27). As in previous problems we use the mass
M of the planet with uniform density r and for the two cores the differential
masses M 0 . For 1/q we use the approximation as in Problem 16:

4 a3 0
4 a3 0
r M r0
0
1
27:1
M p r  r p r  r
3 8
3 8
r
8 r
The potential U is

 a 2 

GM GM 0
2
U

2
3cos y  1
r
r
2r

The radial components of gravity at the equator and the Pole are found by taking the
derivative of the potential U:


@U
GM
2 3a2 
 2 GM 0  2  4 3cos2 y  1
gr
@r
r
r
4r

52

Gravity
On the surface r a, and at the equator y = 90 and at the Pole y = 0 , so
GM GM 0 5
 2
a2
a 4
GM GM 0 14
p
gr  2  2
a
a 4
gre 

The gravity attening is given by


b

gp  ge 1

ge
8

By substituting the values of gravity we nd the relation between M and M 0 :


7 0
5 0
1 M  2 M M 4 M
67

) M M0
5 0
8
4
M  M
4
Putting M 0 in terms of M from Equation (27.1) we nd the ratio of the densities:

4
M r0
r0
 1 ) 1:48
M0 M
67
8 r
r
(b) For a rotating planet we add to the potential U the rotational potential, F:

 a 2 

GM GM 0
GM  r 3 m 2
2

sin y
3cos y  1
2
U
r
r
2r
r a 2

The radial and tangential components of gravity are now



@U
GM
2 3a2 
0
2
gr
 2 GM  2  4 3cos y  1
@r
r
r
4r
rm 2
GM 3 sin y
a
GM
 1:11 2
r
 

1 @U GM 0
a 2
2 
6 cos y sin y
gy
r @y
r
2r
GM  r 3
2
m sin y cos y
r
a
GM
 0:013 2
r

From Fig. 27 we see that the relation between the geocentric and astronomical latitudes is
a  i
where i is the deviation of the vertical with respect to the radial direction, which is given by
tan i

gy
0:012 ) i 0:7
gr

53

Gravity anomalies. Isostasy

gq

a/2
r

gr

i
r

j = 45

ja
a

Fig. 27

Then the astronomical latitude for geocentric latitude 45 is


a 45  0:7 44:3

Gravity anomalies. Isostasy


28. For two-dimensional problems, the gravitational potential of an innite horizontal
cylinder of radius a is

1
2
V 2pGra ln
r
where r is the distance measured perpendicular to the axis. Assume that a horizontal
cylinder is buried at depth d as measured from the surface to the cylinders axis.
(a) Calculate the anomaly along a line of zero elevation on the surface perpendicular
to the axis of the cylinder.
(b) At what point on this line is the anomaly greatest?
(c) What is the relationship between the distance at which the anomaly is half the
maximum and the depth at which the cylinder is buried?
(d) For a sphere of equivalent mass to produce the same anomaly, would it be at a
greater or lesser depth?

54

Gravity

gr
gz
gmax

d
r

(1/2)gmax

a
x

x1/2

Fig. 28

(a) The gravity anomaly produced by an innite horizontal cylinder buried at depth d, with
centre at x 0 (Fig. 28), is given by the derivative in the vertical direction (z-axis) of
the gravitational potential V:
0
1
1
B
C
V 2pGra2 ln@qA
x2 z d2
g gz 

@V
2prGa2 d

@z x2 z d2

28:1

For points on the surface (z 0):


g

2prGa2 d
x2 d 2

(b) To nd the point at which the anomaly has its maximum value, we take the
derivative with respect to x and put it equal to zero:
@g
0 ) 2prGa2 d2x 0 ) x 0
@x
Substituting x 0 in (28.1):
gmax

2pGra2
d

(c) The distance at which the anomaly has a value equal to half its maximum value
gives us the depth d at which the cylinder is buried:

55

Gravity anomalies. Isostasy

gmax
2pGra2 2pGra2 d
g )

) x1=2 d
2
2d
x2 d 2
(d) The gravitational potential produced by a sphere of differential mass DM buried at
depth d under x 0 is given by
V

The gravity anomaly is

GM
GM

1=2
r
x2 z d2

gz

@V
GM z d

3=2
@z
x2 z d2

and for a point on the surface z 0,

gz

GMd
x2

d 2 3=2

The maximum value for x 0 is


gmax

GM
d2

The distance at which the anomaly has half its maximum value is

GMd
GM
3=2
2d 2
x21=2 d 2

x1=2

p
22=3  1 d 0:766d

Therefore, the sphere is at a greater depth than the cylinder.


29. At a point at latitude 42 29 0 19 00 and height 378.7 m the value of gravity is
observed to be 980 252.25 mGal. Calculate in gravimetric units (gu):
(a) The free-air anomaly.
(b) The Bouguer anomaly if the density of the crust is 2.65 g cm3.
(a) We rst calculate the normal or theoretical value of gravity given by the expression


g 9:7803268 1 0:00530244sin2  0:0000058sin2 2 m s2

We substitute for its value 42 290 1900 and obtain

g 9:803 9299 m s2


The free-air anomaly, using the free-air correction, is
gFA g 3:086h  g 238:7 gu

56

Gravity

(b) The Bouguer anomaly is calculated from the free-air anomaly using the Bouguer
correction with a crust density of 2.65 g cm3:
g B g 3:086h  2pGrh  g gFA  2pGrh 659:3 gu
30. An anomalous mass is formed by two equal tangent spheres of radius R, with
centres at the same depth d ( d  R ) and density contrast Dr.
(a) Calculate the Bouguer anomaly at the surface (z 0) produced by the mass
anomaly along a prole passing through the centres of the two spheres.
(b) Represent it graphically for x 0 (above the tangent point), 500, 1000, and 2000 m
taking R 1 km, d 3 km, and Dr 1 g cm3.
(a) For one sphere the anomaly for points on the surface (z 0) is (Problem 28)
g

GMd
x2

d 2 3=2

For two spheres the anomaly is the sum of the attractions of the two spheres (Fig. 30a):
g

GMd GMd

r13
r23

where
q
x  R2 d 2
q
r2 x R2 d 2

r1

Then

GMd
GMd
g h
i3=2 h
i3=2
x  R2 d 2
x R2 d 2
P(x,0)

r2

R
x

Fig. 30a

r1

57

Gravity anomalies. Isostasy

(b) To represent graphically the curve of the anomaly (Fig. 30b), we rst nd the point
at which it is a maximum:
0
1
@g
@ B
@h
@x
@x

GMd

GMd
C
i3=2 h
i3=2 A 0
x  R2 d 2
x R2 d 2

h
i5
h
i5
x R2 x  R2 d 2 x  R2 x R2 d 2 ) x 0 maximum

Using the data given in the problem, we nd the values of the anomaly for the ve points,
with Dr = 1 g cm3, R = 1 km, d = 3 km
4
4
M pR3 r p  109  103 4:19  1012 kg
3
3

x (m)

Dg (gu)

0
500
1000
1500
2000

53.0
52.0
48.9
43.9
37.5

55

Anomaly (g)

50

45

40

2000

Fig. 30b

1000

0
Distance (m)

1000

2000

58

Gravity

31. At a point at geocentric latitude 45 N and height 2000 m the observed value of
gravity is g 6690 000 gu. Taking the approximation that the Earth is an ellipsoid of
equatorial radius a 6000 km, density 4 g cm3, J2 103, and m 103,
calculate for that point:
(a) The free-air and the Bouguer anomalies.
(b) The distance from the free surface to that of the sphere of radius a (precision 1 gu).
(a) The volume of an ellipsoid is:
4
V pa3 1 2a
3
The attening is
a

3J2 m
2  103
2
2

and the mass is


4
Me V r pa3 1 2ar 3:624  1024 kg
3
Using G = 6.67  1011 m3 kg1 s2
GM 6:67  1011  3:624  1024

6:732 994 m s2


a2
36  1012
The value of gravity at the equator in the rst-order approximation is given by

GM
3
ge 2 1 J2  m 6:736 361 m s2
a
2
For a point at latitude 45 N the radial component of gravity is


gr ge 1 b sin2 6:738 045 m s2

where we have used the value of the gravity attening b given by


5
b m  a 0:5  103
2
The free-air correction is
C FA 

2GM
h 2:24  106 h m s2 2:244h gu
a3

Then, the free-air anomaly at that point is


gFA g  g C FA
6 690 000  6 738 045 2:244  2000 43 557 gu
In order to calculate the Bouguer anomaly, we rst calculate the Bouguer correction
C B 2pGrh 1:676  106  h m s2 1:676h gu

59

Gravity anomalies. Isostasy

N
h
a
r

Fig. 31

Then, the Bouguer anomaly is:


gB g  g C AL  2pGrh gFA  C B
45 557  1:676  2000 46 909 gu
(b) If we call N the distance at the given point between the free surface and the surface
of the sphere of radius a (Fig. 31), this is given by:
N arh
where r is the radius of the ellipsoid at latitude 45 N which to a rst approximation is

1
3
r a 1  asin 6000 1  2  10 
5994 km
2


Then,
N 6000  5994  2 4 km
32. Beneath a point A at height 400 m there exists an anomalous spherical mass of
radius 200 m, density 3.5 g cm3, whose centre is 200 m below the reference level.
A point B is located at a height of 200 m and a horizontal distance of 400 m from A,
and a third point C is at a height of 0 m and at a horizontal distance of 800 m from
A. The density of the medium above the reference level is 2.6 g cm3, and below the
reference level it is 2.5 g cm3. The theoretical value of gravity is 980 000 mGal.
Calculate:
(a) The values of gravity at A, B, and C.
(b) The Bouguer anomalies at these points.

60

Gravity

Precision 1 gu.
(a) The gravity at each point is given by
g g  C FA C B C am
where
Normal gravity: g 9800 000 gu
Free-air correction: CFA 3.086 h
Bouguer correction: CB 0.419 r1 h
r1 2.6 g cm3 is the density of the material above the reference level
C am is the anomaly produced by the buried sphere at a point at height h and a horizontal
distance x from its centre:

am

4
G pR3 rsph  r2 h d
GM h d
3


3=2
3=2
2
2
x h d
x2 h d2

32:1

where d is the depth to the centre from the reference level; and rsph and r2 are the densities
of the sphere and of the medium where it is located, respectively. In our case: d 200 m,
rsph 3.5 g cm3, and r2 2.5 g cm3.
For point A, x 0, we obtain
C FA 1234 gu

am

C B 436 gu

4
G pR3 rsph  r2
6 gu
3

h d 2
h d 2
GM

The value of gravity is gA 9 799 208 gu.


At point B:
CFA 617 gu
CB 218 gu
The anomaly produced by the sphere is calculated by Equation (32.1), substituting
x xB 400 m and h hB 200 m
C am 5 gu
We obtain gB 9 799 606 gu.
At point C:
The free-air and Bouguer corrections are null, because the point is at the reference level.
The anomaly due to the sphere, by substitution in Equation (32.1), x xC 800 m, and
h hC 0, is
C am 1 gu
The value of gravity is: gC 9800 001 gu.
(b) The Bouguer anomaly is given by
DgB g CFA CB g

61

Gravity anomalies. Isostasy

By substitution of the values for each point we obtain that the anomalies correspond to
those produced by the sphere:
gAB 6 gu
gBB 5 gu
gCB 1 gu
33. For a series of points in a line and at zero height which are affected by the
gravitational attraction exerted by a buried sphere of density contrast 1.5 g cm3, the
anomaly versus horizontal distance curve has a maximum of 4.526 mGal and a point
of inexion at 250 m from the maximum. Calculate:
(a) The depth, anomalous mass, and radius of the sphere.
(b) The horizontal distance to the centre of the sphere of the point at which the
anomaly is half the maximum.
(a) We know that the inection point of the curve of the anomaly produced by a
sphere buried at depth d corresponds to the horizontal distance d/2. Then
d
xinf ) d 2xinf 2  250 500 m
2
The maximum value of the anomaly at x 0 is
GM
4:526 mGal 45:26 gu
gmax
d2
and solving for DM
DM

45:26  106 m s2  5002 m2


1:6964  1011 kg
6:67  1011 m3 s2 kg1

From this value we calculate the radius of the sphere:


1=3
4
3M 1=3
3  1:6964  1011 kg
M pR3 r ) R

300 m
3
4pr
4  3:14  1:5  103 kg m3
(b) In order that g 12 gmax with z 0 we write
GMd
1 GM
) x1=2 383 m

3=2
2 d2
x21=2 d 2

34. At a point at height 2000 m, the measured value of gravity is 9.794 815 m s2.
The reference value at sea level is 9.8 m s2. The crust is 10 km thick and of density
2 g cm3, and the mantle density is 3 g cm3. Calculate:
(a) The free-air, Bouguer, and isostatic anomalies. Use the Pratt hypothesis with a
cylinder of radius 10 km and a 40 km depth of compensation.
(b) If beneath this point there is a spherical anomalous mass of GDM 160 m3 s2 at
a 2000 m depth, what should the compensatory cylinders density be for the
compensation to be total?

62

Gravity

Fig. 34a

(a) The free-air anomaly is


gFA g  g C FA 9 794 815  9 800 000 3:086  2000 987 gu
The Bouguer anomaly is
gB g  g C FA  C B gFA  0:4191rh
987  0:4191  2  2000 689 gu
To calculate the isostatic anomaly (Fig. 34a) we begin with the calculation of the isostatic
correction assuming Pratts hypothesis and using only a vertical cylinder of radius 10 km
under the point and the compensation level at 40 km. In this way, the correction consists
of the gravitational attraction of a cylinder of radius a and height b at a point at distance c
from the base of the cylinder, which is given by

q p
I
C 2pGr b a2 c  b2  a2 c2
34:1
where Dr is the contrast of densities, which according to Pratts hypothesis is given by
r

hr0
Dh

34:2

where r0 is the density for a block at sea level, which in our case is formed by a crust
of density 2 g cm3 and thickness 10 km over a mantle of density 3 g cm3 and thickness
30 km. For the whole 40 km we use a mean value of density
r0

2  10 3  30
2:75 g cm3
40

63

Gravity anomalies. Isostasy

P
h

rc = 2 g cm3

10 km

D = 40 km

rM = 3 g cm3

Fig. 34b

By substitution in (34.2) we obtain


r

2  2:75
0:13 g cm3
42

The isostatic correction (34.1) is


C I 2  3:1416  6:67  1011  0:13  103 b40
382 gu

p p
100 4  100 1764c  103

Finally the isostatic anomaly is given by


gI g  g C FA C B C I g B C I 689 382 307 gu
(b) If under the point considered there is an anomalous spherical mass (Fig. 34b) at
depth d 2 km, the anomaly it produces is
g am

GM
160

40 gu
d2
20002

The total anomaly now is the Bouguer anomaly plus the anomaly due to the sphere:
g 689  40 729 gu
If the isostatic compensation is total (isostatic anomaly equal to zero), this anomaly must
be compensated by the cylinder. Thus, the necessary contrast of densities Dr to do this can
be calculated using expression (34.1):
729  106 2  3:1416  6:67  1011
p p

 40 100 4  100 1764  103  r

64

Gravity

so
r 0:25 g cm3
As the mean value (crustmantle) of the density is 2.75 g cm3, the density of the cylinder
must now be
r0  r r 0:25 2:75  r ) r 2:50 g cm3
35. At a point on the Earth at height 1000 m, the observed value of gravity is 979 700
mGal. The value at sea level is 980 000 mGal.
(a) Calculate the free-air and Bouguer anomalies.
(b) According to the Airy hypothesis, which is the state of compensation of that height?
(c) What should the depth of the root be for the compensation to be total?
To calculate the compensation, use cylinders of radius 40 km, crustal thickness
H 30 km, crust density 2.7 g cm3, and mantle density 3.3 g cm3.
(a) The free-air anomaly is
g FA g  g C FA g  g 3:086h 86 gu
The Bouguer anomaly is
gB g  g C FA  C B gFA  2p G r h 1046 gu
(b) To calculate the isostatic anomaly according to the Airy hypothesis we rst need to
obtain the value of the root given by the equation
t

rc
h 4500 m
rM  rc

The isostatic correction is given by



q p
C I 2pGr b a2 c  b2  a2 c2
Substituting the values
Dr rM rC 600 kg m3
b t 4500 m
c h H t 35 500 m
a 40 km
we obtain:
CI 409 gu
the isostatic anomaly is:
gI g  g C FA C B C I g B C I 637 gu
The negative value of the anomaly indicates that the zone is overcompensated.

65

Gravity anomalies. Isostasy

(c) If we want the compensation to be total, the value of the isostatic correction
must be
DgI DgB CI 0 > CI DgB 1046 gu
Since the isostatic correction under the Airy hypothesis is

q q
C 2pGr t a2 h H 2  a2 h H t2
I

substituting and solving for t, we obtain


t 13 068 m
For a total isostatic compensation the value of the root (13 068 m) must be much larger
than that corresponding to the 1000 m height, which is only 4500 m.
36. Gravity measurements are made at two points A and B of altitude 1000 m and
 1000 m above the reference level, respectively, 2 km apart along a W-E prole at
latitude 38.80 N. Below a point C located in the direction AB and 1 km from A
is buried a sphere of radius 1 km and centre 3 km below the reference level, of density
r 1.76 g cm3. Calculate:
(a) The value of gravity at A and B.
(b) Using the Airy assumption and neglecting the sphere, calculate the root at A and B.
Crustal density rC 2.76 g cm3, mantle density rM 3.72 g cm3, a 10 km, and
H 30 km.
(a) The gravity observed at points A and B is given by
gA g  C FA C B C am
gB g C FA  C B C am
where g is the theoretical gravity, CFA the free-air correction, CB the Bouguer correction,
and C am the attraction due to the anomalous mass.
The theoretical gravity at the observation point at latitude 38.80 N is


g 9:780 32 1 0:005 3025sin2  0:000 0058sin2 2 9:800714 m s2

The free-air and Bouguer corrections are:

C FA 3:806h 3:806  1000 3806 gu


C B 0:419rC h 0:419  2:76  1000 1156 gu
The attraction due to the spherical anomalous mass (Fig. 36) is given by
C am 

GMz d
3=2
x2 z d2

66

Gravity

XA

XB

h
B

r c = 2.76 g / cm3

Fig. 36

For points A and B, by substitution of the values


4
M pR3 r
3
R 1000 m;
r 1000 kg m3
zA h 1000 m
zB h 1000 m
xA xB 1000 m
d 3000 m
we nd
CAam 16 gu
CBam 50 gu
Then the values of gravity at both points are
gA 9800 627:9  3806 1156:4  16 9:798 048 m s2
gB 9800 627:9 3806  1156:4  50 9:803 314 m s2
(b) To calculate the value of the root under A and B according to the Airy hypothesis we
use the equation
t

rC
h
rM  rC

67

Gravity anomalies. Isostasy

where rC and rM are the crust and mantle densities. Then we nd


2:76
 1000 2875 m
3:72  2:76
2:76
 1000 2875 m
tB
3:72  2:76
tA

37. At a point at latitude 43 N, the observed value of gravity is 9800 317 gu, and the
free-air anomaly is 1000 gu.
(a) Calculate the Bouguer anomaly. Take rC 2.67 g cm3.
(b) If the isostatic compensation is due to a cylinder of radius 10 km which is beneath
the point of measurement, what percentage of the Bouguer anomaly is compensated by the classical models of Airy and Pratt?
(c) According to the Pratt hypothesis, what density should the cylinder have for the
compensation to be total?
(a) First we calculate the normal gravity at latitude 43 N:


g 978:0320 1 0:005 3025sin2  0:000 0058sin2 2 Gal
g 9804 385 gu

The height of the point is determined from the free-air anomaly,


g FA g  g 3:086h 9 800 317  9 804 385 3:086h 1000 gu
and solving for h,
h 1642 m
From this value we calculate the Bouguer anomaly
gB g  g 1:967h 838 gu
(b) To apply the isostatic compensation using the Airy hypothesis we rst calculate the
root corresponding to the height h 1642 m:
t 4:45h 7307 m
The isostatic correction is determined using Equation (34.1) of Problem 34:

q p

C I 2pGr b a2 c  b2  a2 c2

37:1

where
a 10 km; b t 7307 m; c t 30 000 h 38 949 m
r rM  rC 3:27  2:67 0:6 g cm3
which results in C I 70 gu. This represents 8% of the observed Bouguer anomaly.

68

Gravity

If we use the Pratt hypothesis, the contrast of densities corresponding to h 1642 m is


given by
hr0
0:043 g cm3
r
Dh
where we have used r0 rC 2.67 g cm3 as the density of the crust. We now substitute
in Equation (37.1), b = D = 100 km, c = D h = 101.642 km and the obtained value of
Dr 0.043 g cm3, and obtain
C I 148 gu
We have to determine again the Bouguer anomaly using the density according to the Pratt
hypothesis
Dr0
2:63 g cm3
Dh
gB g AL  2pGrh 810 gu
r

The isostatic correction corresponds now to 18% of the Bouguer anomaly.


(c) If the compensation is total the isostatic correction must be equal to the Bouguer
anomaly with changed sign:
C I g B
Using the Pratt hypothesis in order to calculate the density r of the cylinder under the
point, we have to take into account that this density must also be the density used in the
determination of the Bouguer anomaly. Then we write
C I gB gFA 2p G r h

q p

2pGr0  r b a2 c  b2  a2 c2 gFA 2pGrh


and putting
N
and solving for r, we obtain


q p

b a2 c  b 2  a2 c 2

gFA 2pGr0 N
2:46 g cm3
2pGh N

where we have used the values r0 2.67 g cm3 and DgFA 1000 gu.
38. At a point on the Earths surface, a measurement of gravity gave a value of
9795 462 gu. The point is 2000 m above sea level. At sea level the crust is 20 km thick
and of density rC 2 g cm3. The density of the mantle is rM 4 g cm3.
(a) Calculate the free-air and Bouguer anomalies.

69

Gravity anomalies. Isostasy

(b) Calculate the isostatic anomaly according to the Airy and Pratt assumptions.
Use cylinders of 10 km radius and compensation depth of 60 km.
(c) Beneath the point, there is an anomalous spherical mass of GDM 1200 m3 s2.
How deep is it?
Take g 9.8 m s2.
(a) The free-air anomaly is given by
g FA g  g 3:086h 9 795 462  9 800 000 3:086h 1634 gu
For the Bouguer anomaly we rst calculate the Bouguer correction
C B 0:419rh 0:419  2  2000 1676 gu
Then we obtain
g B g  g C FA  C B gFA  C B 42 gu
(b) To calculate the isostatic anomaly according to the Airy hypothesis we determine
rst the value of the root corresponding to the height 2000 m:
t

rc
2
2000 2000 m
h
rM  rc
42

The isostatic correction, using a single cylinder under the point, is given by

q p

C I 2pGr b a2 c  b2  a2 c2
where (Fig. 38a)

Fig. 38a

38:1

70

Gravity

a 10 km;

b t 2 km;

c H t h 20 2 2 24 km

Calling A the term inside the brackets in Equation (38.1)


p p
A 2 100 484  100 576 0:166 km
The isostatic correction is, then, given by

C I 2  3:1416  6:67  108 cm3 =gs2  2 g=cm3  166  102 cm 139 gu


Finally, the isostatic anomaly using the Airy hypothesis is
g I g  g C FA  C B C I 97 gu
According to the Pratt hypothesis, the regional density is given by
r

D
60
r0
3:33 3:22 g cm3
Dh
60 2

where D is the compensation depth (in this problem 60 km) and for r0 (Fig. 38b) we have
used the mean value of the density of the crust (2 g cm3) and of the mantle (4 g cm3)
along the compensation depth
1
2
r0 2 4 3:33 g cm3
3
3
The contrast of densities is
r 3:33  3:22 0:11 g cm3
For the isostatic correction, using the Pratt hypothesis, the term A is now
q p
A 60 102 62  602  102 622 7:4 km

and the correction

P
x
h

20 km

rC = 2 g cm3

40 km
rM = 4 g cm3

Fig. 38b

71

Gravity anomalies. Isostasy

C I 2  3:1416  6:67  1011 m3 =kg1 s2  0:11  103 kg=m3  7:4  103 m
341 gu
Since according to the Pratt hypothesis, the density of the compensating cylinder extends
to the surface of the height 2000 m, we have to calculate again the Bouguer anomaly using
this density (3.33 g cm3). We nd for the Bouguer and isostatic anomalies the values
CB 2 p G r h 2699 gu
DgB DgFA CB 1065 gu
DgI DgB CI 724 gu
(c) If we assume that the isostatic anomaly is produced by a spherical anomalous mass
buried under the point at a depth d under sea level its gravitational effect is given by
gmax

GM
h d2

gI 724 gu

Solving for d we obtain


h d 3517 m ) d 1517 m
39. At a point P at height 2000 m above sea level, a measurement is made of gravity.
The crust at sea level, where gravity is 9.8 m s2, is 20 km thick and of density
3 g cm3, and the density of the mantle is 4 g cm3. Below the point P, at 2000 m depth
under sea level, is an anomalous spherical mass of GDM 1200 m3 s2.
(a) Neglecting the isostatic compensation, what would be the value of gravity at the
point P?
(b) With isostatic compensation, what now is the value of gravity at that point?
Use the Airy and Pratt assumptions for the isostatic compensation (Pratt
depth of compensation, 100 km) with single cylinders of 20 km radius under the
point.
(a) Without isostatic compensation, the gravity observed at point P is equal to the sum of
the normal gravity plus the free-air and Bouguer corrections and the effect of the
anomalous mass. Remember that the free-air correction has negative sign:
gP g  C FA C B C am

39:1

where,
g 9800 000 gu
C FA 3:086  h 3:086  2000 6172 gu
C B 0:419rh 0:419  3  2000 2514 gu
The anomaly due to the anomalous mass is given by
C am gmax

GM
h d 2

39:2

72

Gravity

By substitution of the values,


C am

GM
h d2

75 gu

The gravity observed a P is, then, given by


gP 9 800 000  6172 2514 75 9 796 417 gu
(b) If there is isostatic compensation, according to the Airy hypothesis, we determine rst
the depth of the root, using the density of the crust rC and of the mantle rM:
t

rC
3
h
 2000 6000 m
rM  rC
43

The isostatic correction is


q p

C 2pGr b a2 c  b2  a2 c2
I

39:3

where a 20 km, b 6 km, c 2 20 6 28 km, Dr 1 g cm3, so



q p

C I 2 3:14 6:67  1011 1103  6 202 28  62  202 282  103


553 gu

Then the observed gravity at point P is


gP 9796 417  553 9795 864 gu
According to the Pratt hypothesis we rst determine the contrast of densities
r

hr0
Dh

where D is the level of compensation (100 km) and r0 is the mean density for the crust and
mantle down to depth 100 km:
r0

20  3 80  4
3:8 g cm3
100

Substituting we nd
r

2  3:8
0:074 g cm3
100 2

The isostatic correction is


C I 2  3:14  6:67  1011  0:074  103

q p

 100 202 102  1002  202 1022  103


501 gu

73

Gravity anomalies. Isostasy

where we have used in Equation (39.3) the values


a 20 km;

b 100 km;

c 100 2 102 km

The value of gravity at point P is now


gP 9796 417  501 9795 916 gu
which is larger by 52 gu than using the Airy hypothesis
40. A point P is at altitude 1000 m above sea level. Beneath this point is a sphere
of 1 km radius and GDM 650 m3 s2, with its centre 4 km vertically below the
point P. Given that the density of the sphere is twice that of the crust and 3/2 that of
the mantle, calculate:
(a) The density of the sphere, crust, and mantle.
(b) The value of gravity that would be observed at P for the isostatic compensation to
be total including the sphere.
(c) The radius of the sphere for the root to be null. Comment on the result.
Use the Airy hypothesis for the isostatic compensation with H 30 km, 20 km radius
of the cylinder, and theoretical gravity g 980 Gal.
(a) If we know GDM we can calculate the contrast of densities between the anomalous
mass and the crust
4
3GM
3  650
GM prR3 G ) r

3
4pR3 G 4  3:1416  109  6:67  1011
2:326 g cm3
Since the density of the sphere rsph is double that of the crust rc, the densities of the crust
and mantle are
rsph 2rC ) r rsph  rC 2rC  rC rC 2:326 g cm3
rsph 4:652 g cm3
2
rM rsph 3:101 g cm3
3
(b) If isostatic compensation is total we have
g I 0 g gP C FA C B C I C am

40:1

where g is the normal gravity, gP the observed gravity at point P, CFA the free-air correction,
CB the Bouguer correction, CI the isostatic correction and C am the gravitational effect of
the anomalous mass.
The free-air and Bouguer corrections are given by
C FA 3:806h 3:806  1000 3086 gu
C B 0:419r h 0:419  2:326  1000 975 gu

74

Gravity

P
x
h

d
a

rC

rsph

rM

Fig. 40

The effect of the spherical mass is


C am

GM
h d

650
2

4  106

41 gu

We calculate the isostatic correction using the Airy hypothesis and taking into account the
presence of the spherical anomalous mass. Thus, according to Fig. 40, the equilibrium
between the gravity at P and at sea level far from P is given by


4
pa2 rC H pa2 rM t pa2 rC h H t pR3 rsph  rC
3

and solving for t:


4 
a2 hrC R3 rsph  rC
3
t
3011 m
a2 rM  rC
As in previous problems we calculate the isostatic correction using a cylinder under point P

q p
C I 2pGr b a2 c  b2  a2 c2

75

Gravity anomalies. Isostasy

where a 20 km, b t 3011 m, c h H t 34 011 m and Dr rM rC


0.775 g cm3, resulting in
C I 145 gu
If compensation is total, the isostatic anomaly must be null. This implies that the Bouguer
anomaly is equal, with opposite sign, to the isostatic correction
gI 0 gB  C I ) gB 145gu
But the Bouguer correction can be obtained from Equation (40.1):
g I 0 g gP C FA C B C I C am
Solving for CB we obtain
C B 975 gu
From the denition of the Bouguer anomaly we can nd the value gP of gravity at P:
gB g P  g C FA  C B C am
145 g P  9800 000 3086  975 41 ) g P 9797 703 gu
(c) Since the density of the sphere is greater than the density of the crust, there is an excess
of gravity at P with respect to other points at sea level far from P, which must be
compensated by a root of crustal material inside the mantle with negative gravitational
inuence. In this situation the root can never be null.
41. At 10 km beneath sea level vertically under a point P of height 2000 m there exists
an anomalous spherical mass GDM 104 m3 s2. At sea level, gravity is 9800 000 gu
and the crustal thickness 20 km. The density of the crust is 2 g cm3, and of the
mantle 4 g cm3. Using the Airy assumption for the isostatic compensation with a
cylinder of 10 km radius, calculate for that point:
(a) The observed gravity.
(b) The free-air, Bouguer, and isostatic anomalies.
(a) For point P the Bouguer correction is
C B 0:419rh 0:419  2  h 0:838h 1676 gu
The gravity at point P, if there is no isostatic compensation and other effects, can be
obtained from the normal gravity and the free-air and Bouguer corrections
gP g  C FA C B 9800 000  3:086  2000 0:838  2000
9795504 gu

41:1

Since there is an anomalous mass under point P we have to add its gravitational contribution to the gravity at P. For a spherical mass at depth h d under P the gravitational
attraction is

76

Gravity

C am

GM
2

h d

104 m3 s2
104 m s2 100 gu
108 m2

We calculate the root corresponding to the isostatic compensation, assuming the Airy
hypothesis, and taking into account the presence of the anomalous mass in the same way as
in Problem 40:


4
pa2 rC H pa2 rM t pa2 rC h H t pR3 rsph  rC
3
pa2 rC h H t Ma
t

pa2 rC h Ma
2239 m
pa2 rM  rC

The isostatic correction using a cylinder is given by



q p

C I 2pGr b a2 c  b2  a2 c2

41:2

where a 10 km, b t 2.239 km, c 20 2 2 24.239 km, Dr 4  2 2 g cm3,


resulting in
C I 154 gu
The gravity at point P is the value obtained in (41.1) plus the contribution of the anomalous
mass and minus the isostatic correction:
g P 9795 504 100  154 9795 450 gu
(b) The free-air anomaly is equal to this observed value minus the normal gravity plus
the free-air correction:
gFA gP  g C AL
Substituting the values we obtain
gFA 9795 450  9800 000 3:086  2000 1622 gu

The Bouguer anomaly is given by


g B gP  g C AL  C B 54 gu
Finally the isostatic anomaly is the Bouguer anomaly plus the isostatic correction:
gI 54 154 100 gu
This value corresponds to the gravitational contribution of the anomalous mass.
42. At a point P of height 2000 m above sea level the measured value of gravity is
979.5717 Gal. Beneath P is a sphere centred at a depth of 12 km below sea level, 1 g cm3
density, and radius 5 km. Assuming the Airy hypothesis (H 30 km, rC 2.5 g cm3),
rM 3.0 g cm3), calculate the isostatic anomaly at the point in gu and mGal.

77

Gravity anomalies. Isostasy

For the compensation, assume cylinders of the same radius as the sphere. Normal
gravity g 9.8 m s2.
We rst calculate the root t, assuming the Airy hypothesis, corresponding to the height
2000 m of point P. If the situation is of total isostatic equilibrium, we have to introduce the
effect produced by the sphere in the determination of the root (Fig. 40):

4 
p rsph  rC a3 pa2 rC h pa2 rC H pa2 rC t pa2 HrC pa2 rM t
3
so


4 
rC h a rsph  rC
3
t
10 000 m
r M  rC

42:1

The negative value of t (anti-root) is due to the decit of mass produced by the presence of
the sphere (rsph < rC) under point P.
The isostatic correction, as in previous problems, is calculated taking a cylinder under the point:

q p

C I 2pGrM  rC b a2 c  b2  a2 c2

where b t 10 000 m, a 5000, and c H h 32 000 m.


We obtain CI 36 gu.
The isostatic anomaly at P is equal to the observed gravity minus the normal gravity and
the free-air, Bouguer, and isostatic corrections, and the attraction of the spherical mass:
gI g P  g C FA  C B  C I  C am

42:2

The effect of the anomalous mass is given by




4
G pa3 rsph  rC
267 gu
C am 3
h d 2
By substitution in (42.2)
g I 9 795 717  9 800 000 3:086  2000  0:419  2:5  2000 267  36
25 gu
43. A point A on the Earths surface is at an altitude of 2100 m above sea level.
Calculate:
(a) The value of gravity at A if the isostatic anomaly is 2.5 mGal. Assume the Airy
hypothesis (rC 2.6 g cm3, rM 3.3 g cm3, H 30 km).
(b) If the previous value had been measured with a Worden gravimeter of constant
0.301 82 mGal/division giving a reading of 630.6, calculate the value of gravity at
another point B at which the device reads 510.1 (both readings corrected for drift).
(c) At what depth is the centre of a sphere of density 4 g cm3 and radius 5 km which
is buried in the crust, given that the anomaly created at a point A, 12 km from the

78

Gravity

centre of the sphere, not in the same vertical, is 321 gu. Also calculate the
horizontal distance from the centre to point A.
Take, for compensation, cylinders of 10 km radius. g 9.8 m s2
(a) The isostatic anomaly is given by
g I gA  g 3:086h  0:419rC h C I

43:1

To calculate the isostatic correction CI, assuming the Airy hypothesis, we must rst
calculate the root t that corresponds to the height h
t

rC h
7800 m
rM  rC

As in other problems the isostatic correction is calculated using



q p

C I 2pGrM  rC b a2 c  b2  a2 c2
and substituting the values
b t 7800 m;
a 10 km;

c h H t 39900 m;
r rM  rC 700 kg m3

we obtain
C I 84 gu
Solving for gA in Equation (43.1) we obtain
gA gI g  3:086h 0:419rC h  C I 9795 748 gu
(b) For a Worden gravimeter the increment in gravity between two points (Dg) is proportional to the increment in the values given by the instrument (DL) corrected by the
instrumental variations
g KL
gB  gA KLB  LA ) gB gA KLB  LA
gA  364 gu 9795 384 gu
where K is the constant of the gravimeter.
(c) The anomaly produced by a sphere buried at depth d under sea level at a point at height
h and at a horizontal distance x from the centre of the sphere is given by

C am



4
G pa3 rsph  rC h d
3
3=2
x2 h d2

43:2

79

Gravity anomalies. Isostasy

r
d
rC
a
x

x
rsph

Fig. 43

For point A (Fig. 43) if r is the distance from the centre of the sphere to the point A,
x2 (h d)2 r2 and solving for d in Equation (43.2) gives
d

C am r3

h
4
G pa3 rsph  rC
3

Substituting the values r 12 km, a 5 km, rsph rC 1400 kg m3, h 2100 m,


C am 321  106 m s2, we obtain:
d 9245 m
The horizontal distance is:
x

q
r2  h d2 3910 m

44. In a gravity survey, two points A and B on the Earths surface gave the values 159
and 80 mGal for the free-air anomaly, and 51 and 25 mGal for the Bouguer
anomaly, respectively. Given that B is at an altitude 1000 m lower than A, and that the
density of the mantle is 25% greater than that of the crust, calculate:
(a) The value of gravity at A and B, and the densities of the crust and mantle.
(b) The isostatic anomaly according to the hypotheses of Airy (H 30 km) and Pratt
(D 100 km, r0 the value determined in the previous part) at point A. Take, for
compensation, cylinders of 10 km radius. g 980 Gal.
(a) The free-air anomaly at point A is given by
gAFA gA  g CAFA gA  g 3:086hA

44:1

80

Gravity

The Bouguer anomaly is


gAB gA  g CAFA  CAB gAFA  0:419rC hA
Changing values from mGal to gu, we write for the Bouguer anomalies at points A and B
gAB 510 1590  0:419rC hA ) 0:419rC hA 2100 gu
gBB 250 800  0:419rC hB ) 0:419rC hB 1050 gu

44:2

Dividing both equations, we nd


hA
2
hB
Knowing that the difference in height between A and B is 1000 m, we obtain for the
heights of both points,
hB hA  1000 ) hA hB 1000 2hB ) hB 1000 m ) hA 2000 m
The density of the crust can be obtained from Equation (44.2):
0:419rC hB 105  105 m s2 0:419rC  1000 ) rC 2:505 g cm3
The density of the mantle is 25% more than that of the crust, so
rM rC 1 0:25 1:2  2:505 3:131g cm3
The gravity at A and B is obtained using Equation (44.1):
gA gAFA  3:086hA g 1590  3:086  2000 9800 000 9 795 418 gu
gB 800  3:086  1000 9800 000 9797 714 gu
(b) For the isostatic anomaly at point A, according to the Airy hypothesis, we rst
calculate the value of the root t corresponding to its height:
t

rC h A
hA
8000 m

1:25rC  rC 0:25

For the isostatic correction we use, as in other problems, a cylinder



q p

I
C 2pGrM  rC b a2 c  b2  a2 c2
Substituting the values
b t 8000 m;
a 10 km;

c h H t 40 000 m;
r rM  rC 626 kg m3

we obtain
C I 77 gu

81

Gravity anomalies. Isostasy

The isostatic anomaly is


gAI gAB C I 510 77 433 gu
If we use the Pratt hypothesis, we rst calculate the density corresponding to the material
under point A:
r

Dr0
100  2:505

2:456 g cm3
Dh
100 2

and the contrast of density


r r0  r 2:505  2:456 0:049 g cm3
The isostatic correction is determined using a cylinder,

q p

C I 2pGrM  rC b a2 c  b2  a2 c2
where Dr 0.049 g cm3, b D 100 km, a 10 km, c D h 102 km.
Then, we obtain
C I 158 gu
The isostatic anomaly will be the Bouguer anomaly plus the isostatic correction
gAI gAB C I 510 158 352 gu
In both cases the anomaly is negative, but using the Airy model the value is greater than
using the Pratt model.
45. At a point P on the Earths surface, the observed value of gravity is 9.795
636 m s2, and the Bouguer anomaly is 26 mGal. Assuming the Airy hypothesis
(rC 2.7 g cm3, rM 3.3 g cm3, H 30 km), calculate:
(a) The height of the point.
(b) The isostatic anomaly.
(c) The value of gravity that would be observed at the point if beneath it were a
sphere at a depth of 10 km below sea level, with a density of 2.5 g cm3 and a
radius of 5 km, such that the compensation was total.
Compensation with cylinders of 5 km radius; g 9.8 m s2.
(a) We calculate the height of point P from the Bouguer anomaly:
gB g P  g 3:086h  0:419rC h ) h

gB  gP g
3:086  0:419rC

so
h

260  9 795 636 9 800 000


2099:8 m 2100 m
3:086  0:419  2:7

82

Gravity

(b) To calculate the isostatic anomaly, using the Airy hypothesis, we rst calculate the
value of the root t corresponding to the height of the point:
t

rC
2:7
h
 2100 9450 m
rM  rC
3:3  2:7

Using a cylinder, the isostatic correction is given by



q p

C I 2pGr b a2 c  b2  a2 c2

45:1

where we substitute the values


a 5 km;
b t 9450 m;
c h H t 2100 30 000 9450 41 550 m;
r 3:3  2:7 0:6 g cm3
and obtain
C I 22 gu
The isostatic anomaly is then
gI gB C I 260 22 238 gu
(c) If the compensation is total then the isostatic anomaly must be zero. But now we have
to include the gravitational effect Cam produced by the presence of the anomalous mass
of the sphere.
g I 0 gP  g 3:086h  0:419rC h  C am C I
Solving for gP:
gP g  3:086h 0:419rC h C am  C I

45:2

where the effect of the sphere is given by




4
G pa3 rsph  rC
C am 3
48 gu
h d 2
We calculate the isostatic correction according to the Airy hypothesis. First we calculate
the value of root t, but now we add the effect of the sphere on point P (Fig. 45):


4
rC pa2 h rC pa2 H pa3 rsph  rC rC pa2 t rC pa2 H rM pa2 t
3
Solving for t, we obtain


4 
rC h a rsph  rC
3
7228 m
t
rM  rC

83

Gravity anomalies. Isostasy

P
h

d
x a
rsph

rC

rM

Fig. 45

We substitute this value of t in Equation (45.1) together with the other values
a 5 km;

b 7 7228 m;

c h H t 39 328 m;

r 3:3  2:7 0:5 kg m3

and obtain
C I 18 gu
By substitution in (45.2) we nd the value of the gravity at P under the given conditions:
gP 9800000  3:086  2100 0:419  2:7  2100  48  18 9795830 gu
46. Consider a point on the surface of the Earth in an overcompensated region
at which the values of the free-air and the Bouguer anomalies are 1300 gu and
1200 gu, respectively.
(a) Is this a mountainous or an oceanic zone? Give reasons.
(b) Calculate the altitude and the value of gravity at the point given that the density
of the crust is 2.72 g cm3.
(c) If the isostatic anomaly is -1062 gu calculate, according to the Airy hypothesis
(rC 2.72 g cm3, rM 3.30 g cm3, H 30 km), the value of the root
responsible for this anomaly. Compare it with the value that it would have if
the region were in isostatic equilibrium.

84

Gravity
Compensation with cylinders of 10 km radius; g 9.8 m s2.
(a) Since the free-air anomaly is positive and the Bouguer anomaly is negative, this
indicates that this is a mountainous region.
(b) From the free-air and Bouguer anomalies we can easily calculate the height of the point:
g FA g P  g 3:086h
g B gP  g 3:086h  0:419rC h

46:1

and solving for h,


h

gFA  g B
2193 m
0:419rC

The observed gravity at P can be obtained from either of the two equations (46.1):
gP g  3:086h gFA 9794 532 gu
(c) The isostatic correction is found from the known Bouguer and isostatic anomalies:
g I gB C I ) C I 1062 1200 138 gu
The isostatic correction, using the Airy hypothesis, is given, as in previous problems, as a
function of the root t, by

q p

C I 138 2pGr b a2 c  b2  a2 c2
where we substitute
r 3:3  2:72 0:58 g cm3 ;
a 10 km;
b t;
c h H t 2:193 30 t km
and obtain for t,
t 19 984 m
If the region is in equilibrium the root due to the height h would be
t

rC h
2:72  2193
10 284 m

rM  rC
3:3  2:72

Since we have already obtained a larger value (t 19 984 m), this indicates that the region
is overcompensated.
47. In an oceanic region, gravity is measured at a point on the surface of the sea,
obtaining a value of 979.7950 Gal. Calculate, using the Airy hypothesis (H 30 km,
rC 2.9 g cm3, rM 3.2 g cm3, rW 1.04 g cm3):
(a) The isostatic anomaly if the thickness of the crust is 8.4 km.

85

Gravity anomalies. Isostasy

(b) The thickness that the water layer would have to have if 15 km vertically below
the point there was centred a sphere of 10 km radius such that the anti-root is null
and the compensation total. Also calculate the density of the sphere.
For the compensation, take cylinders of 10 km radius. g 9.8 m s2.
(a) We are in an oceanic region, therefore in the calculation of the root for the isostatic
compensation according to the Airy hypothesis we have to consider the layer of water
of density rW. The value of the root is now given by
t0

rC  rW 0 2:9  1:04 0
h
h 6:2h0
3:2  2:9
rM  r C

47:1

According to Fig. 47a, we have the following relation: H  h0  t 0 e ) h0 H  t 0  e,


where e is the thickness of the crust at the oceanic region, H the thickness of
the normal (sea level) continental crust, h0 the thickness of the water layer, and t 0 the
negative root.
Substituting the values of t 0 from Equation (47.1) we obtain for h0
h0 30  6:2h0  8:4 ) h0 3 km 3000 m
t 0 6:2  3000 18 600 m
From the value of the root we calculate the isostatic correction using a cylinder of 10 km
radius,

q p

I
C 2pGr b a2 c  b2  a2 c2

where Dr 3.2  2.9 0.3 g cm3 300 kg m3, b t0 , c H 30 000 m, and so


C I 269 gu

To calculate the isostatic anomaly we rst have to apply the Bouguer correction which in
this case consists of two terms: the rst to eliminate the attraction of the water layer
P
h

rW

rC

Fig. 47a

rM

86

Gravity
(2pGrWh0 ) and the second to replace this layer by one of density equal to the crustal
density (2pGrCh0 ). Since the point is at sea level the free-air correction is null:
DgI g  g C B  C I
C B 0:419rC  rW h0 2338 gu
Dg I 19 gu
(b) If the anti-root is null and there is total compensation, then we have
t0 0
g I 0 g  g  2pGrW h0 2pGrC h0  C am

47:2

Since the isostatic anomaly must be null, then the anomalous spherical mass and the water
layer must compensate each other. The attraction of the anomalous mass is Cam GDM/d2
where d is the depth of its centre below sea level. Then we can write
g  g 2pGrC  rW h0 

GM
0
d2

where the mass of the sphere is


4
M pa3 rsph
3
If the point P is totally isostatically compensated and the anti-root is null, then (Fig. 47b)
4
pa2 h0 rW pa2 H  h0 rC pa3 rsph  rC pa2 HrC
3
Solving for h0 gives
h0



4a rC  rsph

47:3

3rW  rC

P
h

rW
d
H

Fig. 47b

x a
rsph

rC

87

Gravity anomalies. Isostasy

Substituting this value in (47.2) we obtain


rsph

gg

rC 3372 kg m3
4
a3
pG 2a  2
3
d

and putting this value in (47.3), h 0 3369 m.


48. At a point on the Earths surface, 500 m below sea level, a gravity value is
measured of 980.0991 Gal. If the region is in isostatic equilibrium calculate, using
the Airy hypothesis (H 30 km, rC 2.7 g cm3, rM 3.2 g cm3):
(a) The thickness of the crust.
(b) The isostatic anomaly in gu, with reasons for the sign of each correction. Take
compensating cylinders of 5 km radius; g 9.8 m s2.
(a) We calculate rst the root, according to the Airy hypothesis which corresponds to
the depth of the point, applying the condition of isostatic equilibrium (Fig. 48)
rC H rC H  h0  t 0 rM t 0
so
t0

rC
h0 2700 m
rM  rC

The thickness of the crust e at that point is


e H  h0  t 0 30 000  2700  500 26 800 m
(b) The isostatic anomaly is given by
g I gP  g  3:086h0 0:419rC h0  C I
The free-air correction (3.086h0 ) is negative because the point is below sea level. For the
same reason the Bouguer correction (0.419 rC h0 ) is positive. The isostatic correction,
using the Airy hypothesis, is calculated as in previous problems using a cylinder under the
point of 5 km radius and density contrast D r 3.2  2.7 0.5 g cm3. The value of the
anti-root t 0 has already been calculated, so

q p

I
C 2pGr b a2 c  b2  a2 c2
h

rC

t
rM

Fig. 48

88

Gravity
Substituting c H h0 and b t0 , we obtain
C I 9 gu
Then the isostatic anomaly is DgI 5 gu.
49. In an oceanic region where the density of the crust is 2.90 g cm3 and that of the
mantle 3.27 g cm3, the value of gravity measured at a point P on the sea oor at
depth 4000 m is 9.806 341 m s2.
Calculate, according to the Airy hypothesis:
(a) The thickness of the crust.
(b) The isostatic anomaly in gravimetric units.
Data: rw 1.04 g cm3, H 30 km, g 9.8 m s2. Take, for compensation, cylinders
of 10 km radius.
(a) First we calculate the value of the root according to the Airy hypothesis
t0

rC  rW 0
h 20 108 m
rM  rC

The thickness of the crust under the point is found by (Fig. 49)
e H  h0  t 0 30000  40000  2018 5892 m
(b) Because the point is located at the bottom of the sea, to reduce the observed value of
gravity to the surface of the geoid (sea level) we eliminate rst the attraction of the
water layer. Then we apply the free-air and the Bouguer corrections, to take into
account the attraction of a layer of crustal material which replaces the water. Finally we
apply the isostatic correction:
g I gP  g 0:419rW h0  3:086h0 0:419rC h0  C I
The isostatic correction is calculated using a cylinder of 10 km radius,

q p

C I 2pGr b a2 c  b2  a2 c2
h

rW

rC

Fig. 49

rM

89

Gravity anomalies. Isostasy


where b t0 20 108 m, c H  h0 26 000 m, Dr 3.27  2.9 0.37 g cm3
370 kg m3
By substitution we obtain
C I 598 gu
The isostatic anomaly is:
gI 9 806 341  9 800 000  3:086  4000 0:419  1:04 2:9  4000  598
4 gu
50. At a point with coordinates 42.78 N, 0.5 E and height 1572 m, the observed value
of gravity is 980.0317 Gal.
(a) Calculate the free-air and Bouguer anomalies.
(b) If cylinders of 10 km radius beneath that point are used for the isostatic compensation, calculate the gravimetric attraction of the mass defect corresponding to
the altitude of the point according to the Airy and Pratt hypotheses. Take, for
the crust, H 30 km, rC 2.67 g cm3, for the mantle, rM 3.27 g cm3, and
D 100 km for the Pratt level of compensation.
(c) How deep should the root of the Airy model be for the compensation to be
total?
(a) We calculate rst the normal gravity at the point where gravity has been observed
using the expression


g 9:780 32 1 0:005 3025 sin2 9:804 243 m s2

where is the latitude.


The free-air anomaly is given by

g FA g P  g C FA 9800 317  9804 243 3:086  1572 925 gu


and the Bouguer anomaly by
gB gP  g C FA C B 9800 317  9804 243 1:967  1572 834 gu
(b) If we approximate the isostatic compensation by means of a cylinder of radius a
under the point, we use the expression

q p

C I 2pGr b a2 c  b2  a2 c2
50:1

where Dr is the density contrast, b the height of the cylinder, and c the distance from the
base of the cylinder to the observation point.
Airy: We calculate rst the root given by the equation
t 4:45  h 4:45  1572 6995 km
For the isostatic correction we substitute in (50.1) the values

90

Gravity

a 10 km;

b t 6995 km;

c t H h 38 567 km;

r rM  rC 3:27  2:67 0:6 g cm3


and obtain
C I 2  3:1416  6:67  1011 m3 s2 kg1  0:5  103 kg m3  A
where:
A

6:995

q p

102 38:567  6:9952  102 38:5672  103 270 m

so
C I 68 gu
Pratt: The contrast of densities is now given by
r

h
1575
r
 2:67 0:04 g cm3
D h 0 100 000 1572

and substituting in Equation (50.1) with the values


a 10 km,
b 100 km,
c D + h 101 572 m, we have
C I 135 gu
(c) If the isostatic compensation is total (isostatic anomaly null) the isostatic correction, according to the Airy hypothesis, coincides with the value of the Bouguer
anomaly (834 gu):

q q

g B 2pGr t a2  H h2  a2 t H h2
so
t

q q
gB
a2  H h2  a2 t H h2
2pGr

In this expression we solve for the value of the root t:


q
q
gB
a2 t H h2 t a2  H h2 
tN
2pGr
a2 t H h2 t 2 N 2 2tN

a2 t 2 H h2 2tH h t 2 N 2 2tN
so
t

N 2  a2  H h2
2H h  N

91

Gravity anomalies. Isostasy

By substitution of the values of N, a, H, and h we obtain


t 58 875 m
Because this root has a negative value greater than the thickness of the crust, total
compensation is not possible.
51. Calculate the free-air anomaly observed on a mountain of height 2000 m which
is fully compensated by a root of depth t 10 km. The compensation is by a
cylinder of radius 20 km, the density of the crust is 2.67g cm3, and that of the
mantle is 3.27g cm3.
The free-air anomaly is given by
gFA g  g 3:086h
Since the point is isostatically compensated, we calculate the isostatic correction using a
cylinder as in previous problems:

q p

C I 2pGr b a2 c  b2  a2 c2

where we substitute
b t 10 km,
c h H t 42 km,
and obtain

a 20 km,

Dr 600 kg m3

CI 306 gu
Since the point is totally compensated the isostatic anomaly must be zero:
gI g  g C FA  C B C I 0
The free-air anomaly can now be written as
gFA g  g C FA C B  C I

51:1

We can calculate the Bouguer correction:


C B 1:119h 1:119  2000 2238 gu
and substituting in (51.1) we obtain, for the free-air anomaly,
gFA 2238  301:5 1932 gu
52. Calculate for a point P at height 2100 m, latitude 40 N, and observed gravity
979.7166 Gal the rened Bouguer anomaly in gravimetric units. Consider a surplus
mass compartment of 3000 m, mean height, 3520 m inner radius, 5240 m outer
radius, with n 16. The density of the crust is 2.5 g cm3.
The rened Bouguer anomaly is obtained using, besides the free-air and Bouguer corrections, the correction for the topography or topographic or terrain correction (T):
gB g  g C FA  C B T

92

Gravity

a2
P
Px

a1

a1

a2
hm

h
rC

Fig. 52

The normal gravity at latitude 40 N is given by




g 9:780 32 1 0:0053025 sin2 9:801 747 m s2

the free-air correction by

C FA 3:086h 6481 gu
and the Bouguer correction by
C B 0:419rC h 2200 gu
The topographic correction is introduced in order to correct for the topographic masses not
included in the Bouguer correction, that is, in this case those above the height h (Fig. 52).
Remember that the Bouguer correction corrects for an innite layer or plate of thickness h
and doesnt consider the additional masses above h or the lack of masses below h. The
topographic correction is always positive because the masses above height h produce on
point P an attraction of negative sign which must be added and the lack of mass under h
must also be taken into account with a positive sign, since it has been subtracted in the
Bouguer correction.
To calculate the attraction of the mass above height h we use the attraction of
concentric cylinders (in our case two) with axis passing through point P and with
height equal to the difference between the height h of the point P and the height of
the mass of the topography hm above h. The cylinder is divided into n sectors with
radius a1 and a2 to approximate the topography (Fig. 52). Then the topographic
correction is given by
q q q q

2pGrC
T
a22 c  b2  a22 c2  a21 c  b2 a21 c2
n
In our case we substitute the values
a2 5240 m;

a1 3520 m;

b hm  h 900 m;

c 0;

n 16

93

Gravity anomalies. Isostasy

and obtain
T 2 gu
The rened Bouguer anomaly is
gB g  g C FA  C B T 298 gu
53. Calculate the topographic correction for a terrestrial compartment of inner radius
a1 5240 m, outer radius a2 8440 m, n 20, mean height 120 m, with 2000 m being
the height of the point P. Take r 2.65 g cm3.
In this problem we consider the topographic correction for the case of the lack of mass in
the topography at heights below that of the point P. Since in the Bouguer correction we
have subtracted an innite layer of thickness h, we have to correct for the places where the
mass was not present (Fig. 53).
The topographic or terrain correction T in this case is calculated in the same way as in
the previous problem. Thus we take n sectors of cylinders with axis at point P and height
equal to the difference between h and hm (Fig. 52b). The correction is then given by
q q q q

2pGrC
a22 c  b2  a22 c2  a21 c  b2 a21 c2
T
n
where we substitute a2 8440 m, a1 5240 m, b c h hm, n 20 to obtain
T 0:67 mGal
54. Calculate the topographic correction for an oceanic sector or compartment
of inner radius a1 5240 m, outer radius a2 8440 m, n 20, mean depth
525 m, with 600 m being the height of the point P. Take rC 2.67 g cm3, rW
1.03 g cm3.
In this problem we have to correct for the lack of mass in the oceanic area near the point P,
between the sea level and height h (column 1 in Fig. 54). Also we have to take into account
the attraction produced by the water layer between sea level and the bottom of the sea
(column 2 in Fig. 54).
a2
a1

hm

Fig. 53

rC

94

Gravity

a2
a1

rw

Fig. 54

To calculate the necessary topographic correction we proceed as in Problems 52 and 53,


using cylindrical sectors:
q q q q

2pGr
T
a22 c  b2  a22 c2  a21 c  b2 a21 c2
n
where
a2 8440 m;

a1 5240 m;

n 20

For the correction corresponding to the attraction of column 1, between height h and sea
level, we substitute the values:
b h c;

r rC

and obtain
T1 0:07 mGal
For the correction of the attraction of column 2 between sea level and the bottom of the sea
we use the difference between the densities of the crust and of water:
b p 525 m
c p hs 1125 m
c  b hs
r rC  rW
T2 0:11 mGal
The total topographic correction is
T T1 T2 0:18 mGal

95

Tides

Tides
55. Two spherical planets A and B of radii 2a and a and masses 3m and m are
separated by a distance (centre to centre) of 6a. The only forces acting are gravitational, and the system formed by the two planets rotates in the equatorial plane.
(a) Calculate the value of the components of the acceleration of the tides at the Pole of
each planet directly and using the tidal potential. On which planet are they greater?
(b) If each planet spins on its axis with the same angular velocity as the system, what,
for each planet, is the ratio between the centrifugal force and the maximum of the
tidal force at the equator? On which planet is this ratio greater?
(a) From Fig. 55a we can deduce that at the Pole of planet A, the radial component of
the acceleration of the tides produced by planet B is
grT 

Gm
cos a
q2

where q is the distance from the Pole of planet A to the centre of planet B and a the angle
formed by q and the radius at the Pole of planet A.
By substitution of the required values we obtain
q p
q 6a2 2a2 40a
2a
cos a p ) a 71:6
40a
Gm
Gm
grT 
cos71:6 0:008 2
2
40a
a

The tangential component gyT is given by (Fig. 55a)


Gm Gm
Gm Gm
2 sin a

sin71:6
36a2
q
36a2 40a2
Gm
Gm
2 0:028 0:024 0:004 2
a
a

gyT 

2a

q
b

6a
3m

Fig. 55a

96

Gravity

If we use the tidal potential,



GMr2 
3cos2 #  1
3
2R

where R is the distance between the centres of planets A and B (Fig. 55a), and # is the
angle the position vector r forms with the distance vector R (in this case it is equal to the
colatitude, # y)

GMr2 
3cos2 y  1
c
3
2R
The radial and tangential components of the acceleration are given by



@c
@ Gmr2 
Gmr 

3cos2 y  1
3cos2 y  1
3
3
@r @r 26a
216a
2 

1 @c
1 @ Gmr
Gmr

3cos2 y  1
3cos y sin y
gyT
r @y
r @y 26a3
216a3

grT

For planet A, at the Pole, r 2a and y 90, so


grT 

Gm
Gm
0:009 2
2
108a
a

gyT 0
For planet B, we proceed in a similar manner:
q02 36a2 a2 37a2
a
cos a p ) a 80:5
37a

Therefore,

G3m
Gm
cos80:5 0:013 2
37a2
a
G3m G3m
Gm
T

sin80:5 0:003 2
gy 
36a2 37a2
a

grT 

Using the tidal potential, we obtain the acceleration components


c

G3mr2 
3

3cos2 y  1

26a

@c Gmr 
3cos2 y  1

grT
3
@r 72a
1 @c
9 Gmr
gyT 

cos y sin y
r @y 216 a3
Substituting at the Pole of planet B, r a and y 90, we have
Gm
grT 0:014 2
a
T
gy 0

97

Tides

2a
(3/2)a

(9/2)a

6a

3m

Fig. 55b

(b) First we calculate the centre of gravity of the system formed by the two planets
measured from the centre of planet A (Fig. 55b):
x

3m  0 m  6a 3
a
3m m
2

The rotation radius for planet A is 3/2a and for planet B


3
9
6a  a a
2
2
In the rotating system the centrifugal force equals the force of gravitational attraction,
which at the equator (y 0) is
fg

Gm3m
6a2

9
fC o2 r mo2 a
2

From this expression we obtain the value of the angular velocity o of the rotation system:
3Gm2
9
Gm
mo2 a ) o2
2
36a
2
54a3
Since the angular velocity of the spin of each planet is equal to that of the system, the spin
centrifugal force at the equator of planet A is
fC o 2 r

Gm
Gm
2a
3
54a
27a2

The tidal force is


fT grT

2Gmr 2Gm2a
3 3
R3
6 a

98

Gravity

and their ratio


Gm
fC 27a2

2
Gm
fT
54a2
If we repeat these calculations for planet B, we obtain
Gm
Gm
a
3
54a
54a2
2G3ma
Gm
fT 3 3
6a
36a2

fC o 2 r

and the ratio


Gm
fC 54a2
0:666

Gm
fT
36a2
The ratio is larger for planet A, as expected owing to its larger radius.
56. Two planets of mass M and radius a are separated by a centre-to-centre distance
of 8a. The planets spin on their own axes with an angular velocity such that the value
of the centrifugal force at the equator is equal to the maximum of the tidal force (the
equatorial plane is the plane in which the system formed by the two planets rotates
around an axis normal to that plane).
(a) Calculate the value of the components of the vector g as multiples of GM/a2 for a
point of l' 60 and w 45 (with l 0, being the meridian in front of the other
planet) including all the forces that act.
(b) What is the relationship between the angular velocity of each planet and that of
the system?
(a) The tidal potential is given by
c


GMr2 
3cos2 #  1
3
2R

where R is the centre-to-centre distance between the planets, and # the angle formed by the
vector r to a point and R (Fig. 56a). From this potential we calculate the radial component
of the tidal force:



@c
@ GMr2 
GMr 
2

3cos
#

1
frT
3 3cos2 #  1
3
@r @r 2R
R
At the equator of one planet # 0, r a, and R 8a, so
frT

GMa
8a

GM
256a2

99

Tides

w
gr
g

g
a

x
R
M

Fig. 56a

The spin centrifugal force is


fC o 2 a
Equating these two expressions we nd the value of the spin angular velocity,
r
GM
1 GM
T
2
)o
fC fr ) o a
256a2
16
a3
At a point on the surface of one of the planets the total potential is the sum of the
gravitational potential V, the spin potential F, and the tidal potential c:
U V Fc


GM 1 2 2 2
GMr2 
o r cos
3cos2 #  1
3
r
2
2R

For a point P at latitude and longitude l (Fig. 56b)

cos # cos cos l


and the potential U is
U


GM 1 2 2 2
GMr2 
o r cos
3cos2 cos2 l  1
3
r
2
2R

The components of gravity including the three effects are,



@U
GM
GMr 
 2 o2 r cos2 3 3 cos2 cos2 l  1
@r
r
R
1 @U
GMr
2
o r cos sin 3 3 cos2 l cos sin
g 
r @
R
1 @U
1 GMr2
gl

6 cos2 cos l sin l
r cos @l
r cos 2R3
GMr
 3 3 cos cos l sin l
R
gr

100

Gravity

j
j
a

Fig. 56b

At the required point,


r a;

45 ;

l 60 ) cos # cos 45 cos 60 0:35 ) # 69:3


so

GM
GM
1 GMa
11
GM
3
gr  2
a
 1 0:9993 2
3
3
a
256a
2 8a
24
a
GM
1 1
GMa 1 1 1
GM
g
a p p
3 p p 0:0027 2
256a3
a
2 2 8a3 4 2 2
p p
GMa
GM
21 3
3
gl 
0:0018 2
a
8a3 2 2 2
(b) To obtain the angular velocity of the rotation of the system (O) we take into account
that the centrifugal force due to the rotation of the system at the equatorial plane is
equal to the gravitational attraction between the two planets:
M
2 p G

MM
8a2

where p is the distance from the centre of one planet to the centre of gravity of the system.
Then we nd
M
2 4a G
and nally O o.

MM
8a2

)
2

GM
GM

4  64a3 256a3

101

Tides

57. Consider two planets of equal mass m and radius a separated by a centre-to-centre
distance of 8a. Only gravitational forces act.
(a) Calculate the tidal force at the equator on one of the planets directly, using the
formula of the tidal potential (do so at l 0, i.e. the point in front of the other
planet). Express the result in mGal given that Gm/a2 980 000 mGal.
(b) Compare and comment on the reason for the difference between the results of the
direct calculation and using the tidal potential.
(c) What relationship must there be between the angular velocities of the planets
spin and of the systems rotation for the centrifugal force due to the planets spin
to be equal to the tidal force at the equator and l 0?
(a) The exact calculation of the tidal force at a point located at the equatorial plane in
front of the other planet is (Fig. 57)

Gm
Gm
Gm 1
1
fT


2
46 875 gu
a
49 64
7a2 8a2
Using the tidal potential
c

Gmr2 

2 R

3cos2 #  1

where R is the centre-to-centre distance between the planets (8a), r the radius to the point
where the tide is evaluated (a), and # the angle between r and R (at the equator in front of
the other planet # 0), the radial component of the tidal force can be derived from the
potential. At the equator this is the total tidal force
!

@c
@ Gmr2 
2Gma
2
T

3cos #  1

2 38 281 gu
fr
@r @r 28a3
2512a3
(b) The difference between the value obtained by the exact calculation and by using
the tidal potential is 8594 gu, that is, 18%. This is explained because the tidal

a
x

x
8a
m

Fig. 57

102

Gravity

potential is a rst-order approximation corresponding to terms of the order of


(r/R)2. For a relatively large value of r/R (1/8) this approximation is not very good.
(c) Now we make the spin centrifugal force equal to the tidal force for a point at the
equator of one of the planets which is given by exact calculation

Gm 1
1
fT 2

a
49 64
If the spin angular velocity is op, the spin centrifugal force at the equator is given by
fC o2p a
Equating these two expressions we obtain

Gm 1
1
Gm 1
1

) o2p 3

fC fT ) o2p a 2
a
49 64
a
49 64

57:1

The centrifugal force due the rotation of the system with angular velocity os is equal to the
gravitational attraction between the two planets:
mo2s 4a

Gm
Gm2
) o2s
2
64a
256a3

57:2

From (57.1) and (57.2) we obtain the relation between the two angular velocities:

Gm 1
1

o2p
op
a3 49 64
1:22 )

1:10
Gm
o2s
os
256a3
58. Two planets of equal mass m and radius a are separated by a distance R. The spin
angular velocity of each planet is such that the centrifugal force at the equator is
equal to the maximum of the tidal force. If the sum of the two forces at the equator
cancels the gravitational force, what is the distance R?
The tidal potential is given by
c


Gmr2 1 
3cos2 #  1
3
R 2

where R is the centre-to-centre distance between the planets, r the radius to the point where
the tide is evaluated, and # the angle between r and R. The maximum value is at a point at
the equator in front of the other planet, # 0 and r a. Then
fT

 2Gma
@c Gmr 
3 3cos2 #  1
@r
R
R3

58:1

The spin centrifugal force for a point at the equator is


fC o 2 a

Equating (58.1) and (58.2) we obtain the spin angular velocity,

58:2

103

Tides

2Gma
2Gm
o2 a ) o2 3
3
R
R
If, at the point considered, the sum of the spin centrifugal force and the tidal force cancel
the gravitational force of the planet, then
F fC fT 0
where F  Gm/a2.
The value of R must be

p
4Gma Gm
3
2 ) R 4a
3
R
a

59. Two spherical planets A and B of radii 2a and a and masses 5M and M spin on
their axes with equal angular velocities. They are separated by a centre-to-centre
distance of 8a, and form a system that rotates in the equatorial plane of both planets
with an angular velocity that is equal to that of the spin angular velocity of each one.
(a) Determine the total potential for points on planet A.
(b) Determine the expression for the three components of the total gravity, including
the tide, for a point on the surface of planet A at longitude 0.
(c) If the Love number h on planet A is 0.5, determine the height of the terrestrial tide
as a multiple of a at the equator, at local noon with respect to planet B, in the case
that the systems rotational angular velocity is the same as that of the spin of the
two planets about their axes.
(a) We calculate the centre of gravity of the system, measured from the centre of planet
A (Fig. 59a):
X

5M  0 M  8a 4
a
5M M
3

A
B

2a
(4/3)a

x
8a
5M

Fig. 59a

104

Gravity

j
2a

Fig. 59b

The total potential U at a point on the surface of planet A at latitude is given by the sum
of the gravitational potential V, plus the spin potential , plus the tidal potential c
produced by planet B:
U V Fc
U


5GM 1 2 2 2
GMr2 
o r cos
3cos2 #  1
3
r
2
8a 2

59:1

where according to Fig. 59b

cos # cos cos t


where t is the local time of planet B with respect to planet A (hour-angle), at a point of l 0,
the geographical longitude at planet A. For t 0, the planet B is in front of the point, so
U


5GM 1 2 2 2
GMr2 
o r cos
3cos2 cos2 t  1
3
r
2
28a

(b) The components of gravity are


gr


@U
5GM
GMr 
 2 o2 rcos2
3cos2 cos2 t  1
3
@r
r
8a

gy 
gl

1 @U
GMr
o2 r cos sin
3 cos sin cos2 t
r @
8a3

1 @U
GMr

3 cos cos t sin t


r cos @t
8a3

105

Tides

At a point on the surface of planet A, r 2a,



5GM
GM 2a 
o2 2acos2
3cos2 cos2 t  1
3
2
4a
8a
GM
2
gy o2 2a cos sin 3 2 3 cos sin cos2 t
8 a
GM 2
gl 3 2 3 cos cos t sin t
8 a
gr 

(c) At the equator 0, at 12 h with respect to B, t 180, h 1/2, and o O. The


height of the equilibrium terrestrial tide is given by
Bh

c
g

At the equator of planet A the tidal potential (59.1) is


c

2GM 2a2
28a3

GM
128a

If we approximate g by gr
5GM
GM 2a
5GM
GM
o2 2a
2

o2 2a
3
2
2
4a2
4a
128a
8a
159GM

2o2 a
128a2

gr 

and the height of the equilibrium tide is


GM
1
128a

B
2 159MG
2

2ao
128a2

59:2

We know that the angular velocity of the rotation of the system is equal to the spin angular
velocity of both planets, so the spin angular velocity is given by
4
G5M 2
3GM
) o2
5M o2 a
2
3
256a2
8a
By substitution in (59.2)
GM
1
a
128a


B
3GM
2 159MG
312
2a
128a2
256a2
60. The Earth is formed by a sphere of radius a and density r, and a core of radius a/2
and density 2r, in the northern hemisphere, centred on the axis of rotation and

106

Gravity
tangent to the equatorial plane. The Moon has mass M/4 (where M (4/3)pra3), is at
a distance (centre-to-centre) of 4a, and orbits in the equatorial plane. Determine:
(a) The total potential and the components of gravity including the tidal forces.
(b) The total deviation of the vertical from the radial at lunar noon, and the deviation
due to the tide at the same hour for latitude 45 N, with m 1/8.
(a) The total potential U is equal to the gravitational potential of the planet V1 with
uniform density plus that of the core V2 using the differential mass, the spin
potential F, and the tidal potential c. The gravitational potentials are given by
(Fig. 60):
GM
r
GM 0
V2
q
V1

where the differential mass of the core is


4
a3 M
GM
M 0 p2r  r ) V2
3
8
8q
8
and q is the distance to the centre of the core. Its inverse can be approximated by


 a 2 1 

1 1
a
2
1 cos y
3cos y  1
q r
2r
2r 2

The spin potential is

1
F o2 r2 sin2 y
2
and the total potential is
"
!
#
 r 3 m

GM
1
a
a2 1 
2
2
1
1 cos y
3cos y  1
sin y c
U
r
8
2r
a 2
2r2 2
w

2r
a/2

q r

j
4a
t 180
r

Fig. 60

107

Tides
where m o2 a3/GM.
The tidal potential c due to the Moon is given by
c
According to Fig. 60


GML r2 
3cos2 #  1
2R3

60:1

cos2 # cos2 cos2 t  180


By substitution in (60.1), since R (the centre-to-centre distance between the Earth and the
Moon) is 4a, ML M/4, and 90  y, we obtain
c
The potential U is
U GM


GMr2 
3cos2 cos2 t  180  1
3
512a


5
a
a2 

sin

3 sin2  1
8r 16r2
64r3


r2 m 2
r2 
2
2
3 cos
3 cos cos t  1
a 2
512a3

The components of gravity are found by taking the derivatives of U with respect to r and :


5
2a
3a2 
sin 
3 sin2  1
gr GM  2 
3
4
8r
16r
64r



r
2r
2
2
3
cos

cos
3 m cos2
t

1
a
512a3
1 @U
r @

GM
a
a2


cos 
6 cos sin
2
r
16r
64r3

r2 m
r2
2
6
sin

cos

cos
t
3 2 sin cos
a 2
512a3

gy 

By substituting r a, 45, m 1/8, and at 12 h lunar time, t 180, we obtain


gr 1:128

GM
a2

gy 0:023

GM
a2

(b) The deviation of the vertical with respect to the radial direction at 12 h lunar time is
given by
tan i

gy
0:020 ) i 1:2
gr

108

Gravity

The part of the deviation due to the lunar tide is given by


tan i0

gyM
1 1 @c 1 GMr2

6 sin cos cos2 t

gr r @ gr r512a3
gr

and by substitution of the same values


i0 0:3
The greater part of the deviation is due to the core.
61. Two spherical planets, planet A of radius 2a and mass 3m and planet B of radius a
and mass m, are separated by a centre-to-centre distance of 6a. The system rotates in
the equatorial plane and each planet spins on its axis with the same angular velocity.
What, for each planet, is the ratio between the force of gravity and the maximum of
the tidal force at the equator in front of the other planet?
The centre of gravity of the system measured from the centre of planet A is (Fig. 61)
0  3m 6am 3
a
4m
2
The angular velocity of the system is given by
3mGm

3
Gm
3mo2 a ) o2
3
2
54a
6a
2

The tidal force (radial component) can be calculated from the tidal potential
c



GMr2 
@c GMr 
3 3cos2 #  1
3cos2 #  1 ) frT
3
@r
R
2R

where R is the centre-to-centre distance between the planets. For a point on the equator in
front of the other planet, # 0, and
frT

2GMr

6a3
2Gm2a
Gm
Planet A : frT
0:018 2
216a3
a
2G3ma
Gm
T
0:028 2
Planet B : fr
216a3
a
Gravity without tides is the sum of the gravitational and centrifugal forces:
g

Gm
 o2 r
r2

G3m
3Gm Gm
Gm
 o2 2a

2a 0:71 2
4a2
4a2
54a3
a
Gm
Gm Gm
Gm
2
a 0:98 2
Planet B : g 2  o a 2 
a
a
54a3
a

Planet A : g

109

Tides

2a
(3/2)a

(9/2)a

x
6a

3m

Fig. 61

The ratios between the gravity and tidal forces are:


Planet A :

g
0:71
39:44

T
fr
0:018

Planet B :

g
0:98

35:00
T
fr
0:028

62. The Earth is of radius a and density r, with a core of radius a/2 and density 3r
on the axis of rotation in the southern hemisphere tangent to the equatorial
plane. The Moon has mass M/2 and its centre is at 4a from the centre of the Earth
(M 4/3pra3).
(a) Write down the total potential.
(b) What is the value of the angular velocity of the Earth if at the point 30 N, 30 E at
06:00 lunar time the radial component of gravity is equal to GM/a2?
(c) In this case, what is the ratio between the angular velocity of the Earths rotation
and that of the system?
(a) As in previous problems the total potential U is given by
U V1 V2 F c
The differential mass of the core is:
M1

4p
a3 M
3r  r
3
8
4

110

Gravity

90 q
4a

t
a/2 x
3r

Fig. 62

Using the approximation for 1/q, where q is the distance from a point on the surface of the
Earth to the centre of the core, the gravitational potential due to the core is



GM GM 1
a
a2 
V2

 2 cos y 3 3cos2 y  1
4q
4 r 2r
8r
The potential due to the spin of the Earth is given by

1
GM m
F o2 r2 sin2 y 3 r2 sin2 y
2
a
2
where
m

o2 a3
GM

and the tidal potential due to the Moon is


c

 GMr2 

GM r2 1 
2
3cos
#

1

3sin2 ycos2 t  1
3
3
2 64a 2
256a

where t is the hour-angle of the Moon (Fig. 62)


Then, the total potential U is


5
a
a2 
 2 cos y
3 cos2 y  1
U GM
3
4r 8r
32r

3

r m
r2 
2
2
3 sin2 y
3
sin
y
cos
t

1
a 2
256a3
(b) The radial component of gravity is given by
gr

@U
@r



5
2a
3a2 
GM  2 3 cos y 
3 cos2 y  1
4r
8r
32r4


rm sin2 y
2r 
2
2

3 sin y cos t  1

a3
256a3

111

Tides

Substituting y 60 and at 6 h, t 90, we have


gr

GM
a2

141
3
1
GM
m 

2
128
4 128
a

Solving for m we obtain m 2.81, and the spin angular velocity is


r
2:81GM
rad s1
o
a3
(c) The centre of gravity r0 of the EarthMoon system, measured from the centre of the
Earth, is given by
M
4a
M
r

M
r
8a
T 1
L 2
r0
2

5
1
MT ML
7
M
4 2
where the mass of the Earth MT includes that of the core,
MT M

M
M
and ML
4
2

and r1 0 and r2 4a.


To calculate the angular velocity of the system, O, we put the centripetal force at the
Moon equal to the gravitational force between the Earth and the Moon:
ML r2  r0
2 G

MT ML
r2

Substituting and solving for O, we obtain


5 2

M
M 2
8
8
G
O 4  a;
16a2
2
7

r
GM 7
so O
a3 256

Then, the ratio between the angular velocities of the spin of the Earth and of the system is:
r
2:81GM
o
a3
r 10:14

7GM
256a3
63. Consider two planets of equal mass m and radius a separated by a centre-tocentre distance of 8a. The planets revolve around their centre of mass and
spin around their own axes. Their spin angular velocity is such that the value
of the centrifugal force is equal to the maximum tidal force of the two at the
equator.

112

Gravity

(a) Calculate, for a point on the equator of one of the planets and longitude l 90, the
value of the vector g including all the forces acting at that point (l 0 corresponds
to the point on the line joining the two centres of the planets) at t 0.
(b) What is the deviation of the vertical from the radial at the point w 45, l 0?
(a) If o is the spin angular velocity of the two planets, the centrifugal force at the
equator y 90 , r a, only has radial component:
frC o2 a

63:1

The radial component of the tidal force can be obtained from the tidal potential c which, in
the rst-order approximation, is given by (Fig. 63)
c


Gmr2 
3cos2 #  1
3
2R

where, if is the latitude and l the longitude,

cos # cos cost  l


On the equator f 0, so
c


Gmr2 
3cos2 t  l  1
2R3

The radial component of the tidal force is






@c
@ Gmr2 
Gmr 
2
frT

3cos
t

l

1
3 3cos2 t  l  1
3
@r @r 2R
R

8a
x
m

Fig. 63

90

x
m

113

Tides

The maximum value is for t l and putting R 8a and r a, we obtain


frT

Gma
8a

Gm
256a2

63:2

Then, as the centrifugal force is equal to the tidal force, we put (63.1) equal to (63.2) and
solve for o:
o2 a

Gm
Gm
) o2
256a2
256a3

63:3

We know that m O2 4a Gm2/R2, so solving for O,

Gm
256a3

and then using (63.3), we obtain o/O 1.


The total potential U is the sum of the gravitational, spin, and tidal potentials, which for
t 0, is given by
U V Fc


Gm 1 2 2 2
Gmr2 
o r sin y
3sin2 ycos2 l  1
3
r
2
28a

The components of gravity including the tidal forces are


gr
gy
gl


@U
Gm
Gmr 
 2 o2 rsin2 y
3sin2 ycos2 l  1
3
@r
r
8a
1 @U
Gmr
o2 r sin y cos y
3 sin y cos ycos2 l
r @y
8a3

1 @U
Gmr

3 sin y cos l sin l
r sin y @l
8a3

By substitution of r a, l 90, and y 90 we have


gr 

Gm
Gm
o2 a 
a2
512a2

gy 0
gl 0
(b) By substitution of r a, l 0, and y 45 in (63.4) we obtain

Gm
1
Gm 3
gr  2 o2 a

1
a
2 512a2 2
Gm
Gm 1
Gm 1
1021 Gm


 2
2
2
a
256a 2 512a 2
1024 a2
1
Gm
3
Gm
1
Gm
3
5 Gm

gy o 2 a
2
2
2
2 512a 2 256a 2 512a 2 1024 a2
gl 0

63:4

114

Gravity

The deviation of the vertical with respect to the radial direction is


tan i

gy
0:28
gr

64. Two spherical planets of radii 2a and a and masses 8M and M separated by a
centre-to-centre distance of 4a spin on their own axes and rotate in the equatorial
plane with the same angular velocity.
(a) Determine all the forces acting at a point on the smaller planet at geocentric
coordinates w 60 N, l 0 (00:00 h local time corresponds to passage of the
other planet through the zero meridian).
(b) For this same point, calculate the astronomical latitude and the tidal deviation of
the vertical.
(a) First we determine the centre of gravity of the system, putting the origin at the
centre of the small planet (Fig. 64):
x

0  M 4a  8M 32a

9M
9

Because the spin angular velocity of each planet is equal to the angular velocity of the
system (o O), we can write, for the small planet, putting the gravitational attraction of
the two planets equal to the centripetal force:
G8MM
4a

M o2

32a
9

and solving for o,


o2

9GM
64a3

q r

a
x

x
4a

Fig. 64

8M

2a

115

Tides

On the small planet the gravitational force is


gr 

GM
r2

gy 0
and the force due to its spin is
fr o2 rsin2 y
fy o2 r cos y sin y
For the point under consideration, r a, y 30, we obtain
Gm 9 GM 1
247GM


2
2
2
a
64
p a 4
p256a

3 GM 9 3
9 GM

2
a
4
64 a3
a 256

grGC 
gyGC

To add the tidal force we use the tidal potential in the rst-order approximation,
c


G8Mr2 
3cos2 #  1
3
2R

where cos # sin y cos (t  l).


The tidal force for the point considered, r a, y 30, t l 0, and R 4a, is
given by
@c
GM 1
 2
@r
a 32
p
1
@c
GM
3 3
T
2
fy
r @y
a 32

frT

The total force acting at the point is the sum of the three forces, gravitational, centrifugal,
and tidal:
255GM
2
256a
p
GM 12 3
2
a 256

grtotal 
gytotal

(b) The astronomical latitude is given by a i, where i is the deviation of the


vertical without considering the tide:
p
g total 9 3
0:06 ) i 3:6 ) a 60 3:6 63:6
tan i ytotal
247
gr
The maximum deviation of the vertical due to the tide at the point considered is
i0 given by

116

Gravity

fyT
grtotal

tan i0

p
3 3
32 0:163 ) i0 9:3
255
256

Gravity observations
65. Determine the values of gravity at the following series of points belonging to a
gravimetric survey with a Worden gravimeter, specifying the drift correction for each
of them.
Station

Time

Reading

A (base)
B
C
D
A

08:30
09:21
11:34
13:20
14:20

562.5
400.7
437.9
360.1
568.8

The gravity at the base is 980.139 82 Gal, and the gravimeter


constant is 0.301 81 mGal/ru (ru: reading unit).
The instrument drift is given by
d

LAe  LAb
tAe  tAb

where LAb and LAe are the readings at the base A at the beginning and end of the
measurements taken at times tAb and tAe, respectively. By substitution we obtain
d

568:8  562:5
1:08 ru=hour
14:33  8:50

The corrected reading for station j is given by


Ljc Lj d (tj tAb)
where Lj is the reading taken at time tj.
For a Worden gravimeter the increment in gravity between two points (D g)
is proportional to the increment in the readings corrected by the instrument drift
(D Lc ):
D g K D Lc
where K is the instrument constant.
Thus, from the readings we obtain the following results.

117

Gravity observations

Station

Corrected reading

D g (mGal)

g (mGal)

A (base)
B
C
D
A

562.5
399.8
434.6
354.9
562.5

49.10
10.50
24.05
62.66

980
980
980
980
980

139.82
090.72
101.22
077.17
139.83

66. Point A is at a geopotential level of 97.437 43 gpu. Point B is at a difference


of 15.213 m in height relative to A, and has a value of gravity of 9.712 611 m/s2.
Calculate:
(a) The value of gravity at point A, if the difference in readings of a Worden gravimeter
between B and A is 17.8 ru, and the gravimeter constant is 0.308 21 mGal/ru.
(b) The geopotential number, dynamic height, and Helmert height of the point B given
that the normal gravity at a point of latitude 45 on the ellipsoid is 980 629.40 mGal.
(a) Using a Worden gravimeter the increment of gravity between points A and B is given by
DgAB KDLBA 5:5 mGal
where K is the instrument constant and DL is the difference between the readings at points A and B.
The gravity at point A is
gA gB DgAB 971 266:6 mGal
(b) The geopotential number at B can be calculated from the value at A in the form
CB CA

g g 
A
B
hBA 82:661 59 gpu
2

where gravity is given in Gal and increments in height in km, because the geopotential
units are, 1 gpu 1 kGal m 1 Gal km.
The dynamic height is given by
HDB

CB
84:294 m
g45

The Helmert height can be calculated from the dynamic height by


H

C
g 0:0424H

where C is given in gpu, g in Gal, and H in km. Solving for H, we obtain


p
g  g2 4  0:0424C
H
2  0:0424
Taking the positive solution because point B is above the geoid (CB > 0) we obtain
HB 85:107 m

118

Gravity

67. In a geometric survey with measurements of gravity using a Worden gravimeter,


the following values were obtained:
Station

Gravimeter reading (ru)

Time

A (base)
B
C
A

1520.23
1759.15
1583.11
1521.30

8h
9h
9h
9h

50
15
35
50

m
m
m
m

Gravity (gu)

Height difference (m)

9 793 626.8
9 794 363.9
9 793 820.7

30.410
301.863

Calculate the gravimeter readings corrected for drift, and the gravimeter constant
The instrument drift is given by
d

LAe  LAb
tAe  tAb

where LAb and LAe are the readings at the base A at the beginning and end of the
measurements taken at times tAb and tAe, respectively. By substitution we obtain,
d 1:07 ru=hour
A reading corrected at station j is given by
Lcj Lj  dtj  tAb
where Lj is the reading at time tj.
For a Worden gravimeter the increment in gravity between two points (D g) is proportional to the increment in the readings corrected by the instrument drift (D Lc):
D g K DLc
where K is the instrument constant. Thus, K can be calculated in the form
K

g
Lc

From each pair of observations we obtain a value of K. Finally we take the arithmetic mean
(Km) from all the values obtained. The results are given in the following table.
Station

Corrected reading (ru)

Gravity (mGal)

K (mGal/ru)

A (base)
B
C
A

1520.23
1758.70
1582.31
1520.23

979 362.68
979 436.39
979 382.07

0.3091
0.3079
Km 0.3085

68. The following table is obtained from observations with a LacosteRomberg


gravimeter:

119

Gravity observations

Station

Gravimeter reading

Time

A
B
C
A

3614.351
3650.242
3610.633
3614.414

10:10
10:25
10:37
11:02

The gravimeter scale factor is 1.000 65, and the equivalence between reading units
and the relative value of gravity in mGal is given by

Reading

Value in mGal

Interval factor

3600
3700

3846.02
3953.15

1.071 25
1.071 40

Given that the value of gravity at point A is 9.794 6312 m s2, calculate the values at
B and C.
First we correct the readings by the instrument drift:
d

LAe  LAb
tAe  tAb

where LAe and LAb are the readings at the base A at the end and the beginning of the survey
at times tAe and tAb. Then
d 0:0727 ru=hour
The corrected reading at each station j is given by
Lcj Lj  dtj  tAb
where Lj is the reading at time tj. The corrected readings are:
LcA 3614:351
LcB 3650:224
LcC 3610:600
These readings are converted into relative gravity values Rj using the conversion table.
The reading at station A is
LcA 3600 14:351
and the relative gravity value is
RA (3846.02 (14.351  1.07125)) 1.00065 3863.90 mGal
For stations B and C,

120

Gravity

RB (3846.02 (50.224  1.07125)) 1.00065 3902.36 mGal


RC (3846.02 (10.600  1.07125)) 1.00065 3859.88 mGal
To convert the relative values into absolute values we need to know both values at one
station, in our case in station A:
gA 9.794 6312 m s2 979 463.12 mGal
gB gA RA RB 979 501.58 mGal
gC gA RA RC 979 459.10 mGal

Geomagnetism

Main eld
69. Assume that the geomagnetic eld of the Earth is a geocentric dipole with a North
Pole at 80 N, 45 E and a magnetic moment 8  1022 A m. Calculate for a point with
geographical coordinates 45 N, 30 W the components NS, EW, and Z of the Earths
magnetic eld, the declination and inclination, and the geomagnetic longitude.
Earths radius: 6370 km and the constant C 107 H m1 (this value is used in all
problems).
We calculate rst the geomagnetic latitude and longitude (f  , l  ) from the geographical
coordinates (f, l) of the point and the geographical coordinates of the geomagnetic North
Pole (fB, lB) by the equation
sin  sin B sin cos B cos cosl  lB
sin l

sinl  lB cos
cos 

Substituting the values


fB 80 N
lB 45 E
f 45 N
l 30 W 330
we obtain
f 46:70
l 84:82
In the geocentric magnetic dipole model, the vertical (Z  ) and horizontal (H  ) components
of the magnetic eld can be obtained from
Z  2B0 sin f
H  B0 cos f
Cm
B0 3
a
121

69:1

122

Geomagnetism

In these equations B0 is the geomagnetic constant, m the magnetic moment of the dipole,
a the Earths radius, and the constant C 107 H m1.
In this case we are given that
m 8  1022 A m2
a 6370 km 6.37  106 m
By substitution in Equations (69.1) we obtain:
B0 30 951 nT
Z  45 051 nT
H  21 227 nT
The geomagnetic declination is given by
 cos fB sinl  lB
cos f


D 14:16

sin D

The NS (X  ) and EW (Y  ) components are


X  H  cos D 20 582 nT
Y  H  sin D 5193 nT
Finally, the geomagnetic inclination or dip (I  ) at that point is given by
tan I  2 tan f ) I  64:77
70. Assume that the geomagnetic eld is produced by a geocentric dipole of magnetic
moment 8  1022 Am, with North Pole at 80 N, 70 W, and that the Earths radius is
6370 km. Calculate for a point with geographical coordinates 60 N, 110 E:
(a) Its geomagnetic coordinates, the components of the Earths magnetic eld (X  ,
Y , Z  , H  ), the total eld, the declination, and the inclination.
(b) The equation of the line of force passing through it.
(a) For this point the difference in longitude from the Geomagnetic North Pole
(GMNP) is 180 (Fig. 70), so both are on the same great circle. Then, the
geomagnetic coordinates are obtained from
f f  90  fB 50
l 180
The expressions for the geomagnetic vertical and horizontal components and for the total
geomagnetic eld are
Z  2B0 sin f
H  B0 cos f
p
F  H 2 Z 2
B0

Cm
a3

Main eld

123

GMNP

GNP
P
x
f

fB

l=0

l
l

Fig. 70

The values of the constants are


m 8  1022 A m2
a 6370 km 6.37 106 m
C 107 H m1
Substituting in the above equations we obtain
B0 30 951 nT
Z  47 420 nT
H  19 895 nT
F  51 424 nT
For this point the geomagnetic declination D  0. So the NS (X  ) and EW (Y  )
components are
D 0
X  H  cos D H  19 895 nT
Y  H  sin D 0
The geomagnetic inclination (I  ) is given by
tan I  2 tan f ) I  67:23
(b) The equation of the line of force passing through a point with geomagnetic
co-latitude y is
r r0 sin2 y
In this equation r0 is the distance from the Earths centre to a point on the line of force with
y 90. The distance r0 is different for each line of force.

124

Geomagnetism
For the point with geomagnetic latitude f  50 located in the Earths surface (r a),
y 90 f  40, so
a
r0 2 15 417km
sin y
71. Assume that the geomagnetic eld is produced by a geocentric dipole of magnetic
moment 7.5  1022A m, with North Pole at 75 N, 65 W, and that the Earths radius
is 6372 km. Calculate:
(a) The NS and EW components for a point on the Earths surface at which the
inclination is 67 and the geomagnetic longitude is 120 .
(b) The geographical coordinates of that point.
(c) The geomagnetic coordinates, eld components, declination, and inclination of the
point on the geographical equator of zero geomagnetic longitude.
(a) The geomagnetic latitude f  is obtained from
tan I  2 tan f ) f 49:7
The horizontal component, H  , can be calculated from the geomagnetic constant, B0, and
the geomagnetic latitude:
Cm
B0 3 28 989 nT
a
H  B0 cos f 18 761 nT
To obtain the NS (X  ) and EW (Y  ) components it is necessary to calculate the declination (D  )
from the spherical triangle with vertices at the Geographical North Pole (GNP), Geomagnetic
North Pole (GMNP), and the point P (Fig. 71a). But we need to calculate the geographic latitude
rstly by solving the spherical triangle. Applying the cosine law to the angle (90 f):
cos90  f cos90  fB cos90  f
sin90  fB sin90  f cos180  l
sin f sin fB sin f  cos fB cos f cos l

71:1

By substitution of the values,


f 55:1
To obtain the geomagnetic declination we apply the sine law in the spherical triangle of
Fig. 71a:
sin D
sin180  l
cos fB sin l


)
sin
D


sin90  fB
cos f
sin90  f

71:2

and substituting the values we nd


D 23:1
It is important to note that we have added a minus sign in the last equation in order for the
declination be positive toward the east.

Main eld

125

GNP
90 fB
l lB
180 l

90 f

GMNP
q = 90 f

Fig. 71a

The NS (X  ) and EW (Y  ) components are


X  H  cos D 17 262 nT
Y  H  sin D 7 349 nT
(b) The calculated geographical latitude is
f 55:1
The geographical longitude is obtained by applying the cosine law to the spherical triangle
of Fig. 71a:
cos90  f cos90  fB cos90  f
sin90  fB sin90  f cosl  lB

71:3

cos y sin f sin fB sin f cos fB cos f cosl  lB


so
cosl  lB

sin f  sin fB sin f


cos fB cos f

71:4

Substituting the values, gives


l  lB 101:6
To take the inverse cosine in the correct quadrant we bear in mind that l  < 0 implies
that the point is to the west of the Geomagnetic North Pole, that is l lB < 0. So we
obtain
l 166:6 W
(c) If this point is on the geographical equator (f 90 ) and has zero geomagnetic
longitude (l  0), it is on the same geographical meridian as the Geomagnetic
North Pole. Then from Fig. 71b

126

Geomagnetism

GMNP

GNP

fB

P
f
x

Fig. 71b

f 90  fB 15:0
l 0 
Z  2B0 sin f 15 006 nT
H  B0 cos f 28 001 nT
D 0
X  H
Y  0 nT
tan I  2 tan f ) I  28:2
72. Assume that the geomagnetic eld is that of a dipole with North Pole at 75 N, 0 E.
What is the conjugate point of that of geographical coordinates 30 N, 30 E?
First, we calculate the geomagnetic coordinates (f  , l  ) (Problem 71; Fig. 71a):
sin f sin fB sin f cos fB cos f cosl  lB
sin l

72:1

sinl  lB cos f
cos f

The values of the geographical coordinates are


fB 75
f 30

lB 0
l 30

Main eld

127

GMNP

GNP

fB

Px
f

f
P1
x

Fig. 72

By substitution in Equations (72.1) we obtain:


f 42:6
l 36:0
A magnetic conjugate point is a point on the Earths surface that is located on the same line
of force and in the opposite hemisphere (Fig. 72, P and P1). Then, its geomagnetic
coordinates (f1 , l1 ) are
f1 f 42:6
l1 l 36:0
To calculate the geographical coordinates for this point (f1, l1) we use the spherical
triangle of Fig. 71a. We calculate the geographical latitude applying the cosine law:
cos90  f1 cos90  fB cos90  f1
sin90  fB sin90  f1 cos180  l1
sin f1 sin fB sin f1  cos fB cos f1 cos l1
f1 53:9
To calculate the geographical longitude we apply the cosine law again:
cos90  f1 cos90  fB cos90  f1
sin90  fB sin90  f1 cosl1  lB

128

Geomagnetism

cosl1  lB

sin f1  sin fB sin f


cos fB cos f

l1  lB 47:3
and taking the solution in the correct quadrant
l1 > 0 ) l1  lB > 0
l1 47:3
73. Assume the centred dipole approximation, with the coordinates of the Geomagnetic North Pole being 65 N, 0 E, and the magnetic moment of the dipole 8  1022
A m. Calculate, for a point on the Earths surface at geographical coordinates 30 N,
30 E:
(a) The geographical coordinates of the conjugate point.
(b) The declination, inclination, and vertical and horizontal components of the eld
at both points. Compare and contrast the results.
Earths radius: 6370 km.
(a) First, we calculate the geomagnetic coordinates (f  , l  ) for point P with geographical coordinates f 30 N, l 30 E using the equations (Problem 71; Fig. 71a)
sin f sin fB sin f cos fB cos f cosl  lB
sin l

73:1

sinl  lB cos f
cos f

The geographical coordinates of the Geomagnetic North Pole are


fB 65 ;

lB 0 

Substituting in Equations (73.1) results in


f 50:4
l 42:7
The geomagnetic coordinates (f1 , l1 ) of the magnetic conjugate point P1 satisfy
(Problem 72):
f1 f 50:4
l1 l 42:7
To calculate the geographical coordinates for this point (f1, l1) we use the spherical
triangle of Fig. 71a. We calculate the geographical latitude applying the cosine law
cos90  f1 cos90  fB cos90  f1
sin90  fB sin90  f1 cos180  l1
sin f1 sin fB sin f1  cos fB cos f1 cos l1
f1 63:7

Main eld

129

To calculate the geographical longitude we apply the cosine law again and, solving for l1,
sin f1  sin fB sin f1
cos fB cos f1

l1  lB 77:1
l1 > 0 ) l1  lB > 0
cosl  lB

l1 77:1
(b) First we calculate the geomagnetic constant B0
B0

Cm
8  1022

30 951 nT
a3
63793  109

We calculate the declination D  , inclination I  , and vertical Z  and horizontal H 


components of the eld at both points. The results are shown in the table, where we notice
that except for the declinations, which are very different, all other values are equal for both
points except in sign.
P

P1

 cos fB sinl  lB
sin D
cos f
D  19.33
tan I  2 tan f  ) I  67.5
Z  2B0 sinf  47 662 nT
H  B0 cosf  19 750 nT

 cos fB sinl1  lB
sin D1
cos f1
D1 40:2
tan I1 2 tan f1 ) I1 I1 67:5
Z1 2B0 sin f1 Z  47 662 nT
H1 B0 cos f1 H  19 750 nT

74. Assume the centred dipole approximation, with the coordinates of the Geomagnetic North Pole 78.5 N, 70.0 W, and the magnetic dipole moment being 8.25  1022
A m. Calculate, for a point on the surface with coordinates 60.0 S, 170.0 W:
(a) Its geomagnetic coordinates, declination, inclination, and vertical and horizontal
components of the eld.
(b) The potential at that point.
(c) The declination and inclination at the point diametrically opposite to it.
Earths radius: 6370 km.
(a) We calculate the geomagnetic coordinates (f  , l  ) using the equations (Problem 71,
Fig. 71a)
sin f cos y sin fB sin f cos fB cos f cosl  lB
sin l

74:1

sinl  lB cos f
cos f

The values of the geographical coordinates are


fB 78:5
f 60:0

lB 70:0
l 170:0

130

Geomagnetism

Substitution in Equations (74.1) gives


f 60:0
l 80:0
The geomagnetic declination is given by (Problem 71, Fig. 71a)
 cos fB sinl  lB
cos f


D 23:1
sin D

The inclination (I  ) at that point is given by


tan I  2 tan f ) I  73:9
The vertical (Z  ) and horizontal (H  ) components of the magnetic eld can be obtained
from
Z  2B0 sin f
H  B0 cos f
Cm
B0 3
a
Substituting the values given we obtain:
B0 31 918 nT
Z  2B0 sin f 55 279 nT
H  B0 cos f 15 963 nT
(b) The potential at a point on the Earths surface (r a) at geomagnetic latitude f  is
given by
F

Cm cos y Cm sin f

176 T m
a2
a2

(c) We can observe in Fig. 74 that at the point diametrically opposite the geographical
and geomagnetic coordinates are
f1 f 60:0 N
l1 l 180 10:0 E
f1 f 60:0
l1 l 180 100:0
The geomagnetic declination at that point satises (Fig. 71a)
 cos fB sinl1  lB  cos fB sinl 180  lB

cos f1
cos f
cos fB sinl  lB

 sin D
cos f

sin D1

Main eld

131

GNP

GMNP

fB

P1
f1

f
l
P

Fig. 74

Then, D1 D .


The geomagnetic inclination is given by
tan I1 2 tan f1 2 tan f ) I1 I 
So, we can notice that the two points P and P1 (Fig. 74) that are diametrically opposite
are not magnetic conjugate points because the geomagnetic longitudes are different
by 180.
75. Consider a point P on the Earths surface at coordinates 30 S, 10 W at which
the NS component of the geomagnetic eld is 27 050 nT and the EW component is
5036 nT, with the geomagnetic inclination being negative. Assuming the centred
dipole hypothesis with magnetic moment 7.8  1022 A m, calculate:
(a) The geographical coordinates of the Geomagnetic North Pole.
(b) The geomagnetic coordinates of Ps conjugate point.
(a) We calculate rst the geomagnetic constant B0:
B0

Cm
30 177 nT
a3

The geomagnetic declination D  is obtained from the NS (X  ) and EW (Y  ) components


of the geomagnetic eld:
tan D

Y  5036

) D 10:5
X  27 050

132

Geomagnetism
The geomagnetic latitude f  is calculated from the horizontal component H  :
p q
H  X 2 Y 2 270502 50362 27 515 nT
H  B0 cos f ) f

H
24:3
B0

We then have two solutions for the geomagnetic latitude. To choose the correct one we bear
in mind that a negative value of the geomagnetic inclination implies a negative value of the
geomagnetic latitude:
tan I  2 tan f
I  < 0 ) f < 0
f 24:3
With these results we calculate the geographical coordinates of the Geomagnetic North
Pole (fB, lB) using the spherical triangle in Fig. 71a. Applying the cosine rule for the angle
90 fB:
cos90  fB cos90  f cos90  f sin90  f sin90  f cos D
sin fB sin f sin f cos f cos f cos D
fB 79:0
To calculate the longitude lB of the Geomagnetic North Pole, we apply the cosine law for
the angle 90 f  :
cos90  f cos90  fB cos90  f sin90  fB sin90  f cosl  lB
sin f sin fB sin f cos fB cos f cosl  lB
cosl  lB

sin f  sin fB sin f


) l  lB 61:0
cos fB cos f

To choose the correct sign for the longitude we notice that the declination is negative and
then the point must be to the east of the Geomagnetic North Pole:
D  < 0 ) l  lB > 0
lB 61  10 71:0 W
(b) The geomagnetic coordinates (f1 , l1 ) of Ps conjugate point verify that
f1 f 24:3
l1 l
We calculate the geomagnetic longitude l  by
sinl  lB cos f
cos f

l 56:2 l1
sin l

Main eld

133

76. At a point P on the Earths surface with coordinates 45 N, 30 W, the value of the
total geomagnetic eld is 49 801 nT, the horizontal component is 21 227 nT, and the
EW component is 5171 nT, with the magnetic inclination being positive. Calculate:
(a) The geographical coordinates of the Geomagnetic North Pole.
(b) The value of the geomagnetic potential at P.
(c) The distance from the Earths centre to the point at which the line of force passing
through P intersects the geomagnetic equator.
Earths radius: 6370 km.
(a) We calculate rst the geomagnetic inclination, latitude, and declination by
cos I 

H
) I  64:8
F

tan I  2 tan f ) f 46:7


sin D

Y
) D 14:1
H

With these results we calculate the geographical coordinates of the Geomagnetic North
Pole solving the spherical triangle (Fig. 71a) in the same way as in Problem 71:
sin fB sin f sin f cos f cos f cos D
fB 80:0
sin f sin fB sin f cos fB cos f cosl  lB
sin f  sin fB sin f
cos fB cos f
l  lB 75:2
D > 0 ) l  lB < 0

cosl  lB

lB 45:2 E
(b) The geomagnetic potential at point P on the Earths surface (r a 6370 km) is given by
F

Cm cos y
B0 a sin f
a2

76:1

We calculate the geomagnetic constant B0 from the horizontal component H  :


H  B0 cos f ) B0

21 227
30 951 nT
cos46:7

Substituting in the potential equation (76.1) we obtain


F 143 T m
(c) The equation of the line of force passing through a point with geomagnetic colatitude y is
r r0 sin2 y

134

Geomagnetism

In this equation r0 is the distance from the Earths centre to the point at which the line of
force passing through P intersects the geomagnetic equator. Substituting r a 6370 km
gives
r0

a
a

 13 543 km
2
sin y cos2 f

77. Assume the centred dipole approximation, with the coordinates of the Geomagnetic North Pole being 75 N, 65 W, and the magnetic moment of the dipole
7.5  1022 A m2. For a point on the Earths surface at which the inclination is 67 and
the geomagnetic longitude is 120 , calculate:
(a) The NS and EW components.
(b) Its geographical coordinates.
The Earths radius: 6372 km.
(a) We calculate rst the geomagnetic constant B0, latitude f  , and horizontal H 
component:
B0

Cm
28 989 nT
a3

tan I  2 tan f ) f 49:7


H  B0 cos f 18 761 nT
To calculate the NS (X  ) and EW (Y  ) components it is necessary to obtain rst the
geographic latitude (f  ) and the geomagnetic declination (D  ). Applying the cosine rule
for the angle 90 f (Fig. 71a),
sin f sin fB sin f  cos fB cos f cos l
f 55:1
The geomagnetic declination is given by
sin D

 cos fB sin l
cos f

D 23:1
From this value we obtain the NS and EW components:
X  H cos D 17 262 nT
Y  H sin D 7349 nT
(b) The geographic latitude was already obtained,
f 55:1

Main eld

135

To calculate the geographical longitude we apply the cosine law for the angle 90 f 
(Fig. 71a):
sin f sin fB sin f cos fB cos f cosl  lB
cosl  lB

sin f  sin fB sin f


cos fB cos f

l  lB 101:7
To choose the correct solution we notice that the declination is positive and then the point
must be to the west of the Geomagnetic North Pole
D  > 0 ) l  lB < 0
l 101:7  65 166:7 W
78. Assume a spherical Earth of radius 6370 km, with magnetic eld produced by a
centred dipole whose northern magnetic pole is at 70 N, 60 W. Given that for a point
on the surface with coordinates 50 S, 80 W the horizontal component is 24 890 nT,
calculate:
(a) The magnetic dipole moment.
(b) The geographical coordinates of the conjugate point.
(a) We calculate rst the geomagnetic latitude by (Fig. 71a)
sin f sin fB sin f cos fB cos f cosl  lB
f 30:9
To obtain the magnetic dipole moment m we need the geomagnetic constant B0, which is
related with the horizontal component H by
B0

H
29 007 nT
cos f

B0

Cm
B0 a3
7:5  1022 A m2
)m
3
a
C

(b) Let us obtain rst the geomagnetic longitude by (Problem 71, Fig. 71a)
sin l

sinl  lB cos f
cos f

l 14:8
The geomagnetic coordinates of the conjugate point (f1 , l1 ) are (Fig. 78)
f1 f 30:9
l1 l 14:8

136

Geomagnetism

GNP

GMNP

fB

P1

f1

Fig. 78

Solving again the spherical triangle (Fig. 71a) we calculate the geographical coordinates
(f1, l1):
sin f1 sin fB sin f1   cos fB cos f1  cos l1 
f1 11:4
cosl  lB

sin f1   sin fB sin f1


cos fB cos f1

l1  lB 13:0
l < 0 ) l1  lB < 0
l1 73:0
79. If the Earths geomagnetic eld is produced by a centred dipole, tilted 15 away
from the axis of rotation, of magnetic moment 7.6  1022 A m2, and the Geomagnetic
North Pole is at longitude 65 W, calculate:
(a) The geomagnetic constant in nT.
(b) The geographical coordinates of a point on the Earths surface at which the
declination is D  14 15.50 and the inclination is I  65 23.50 . Discuss the
possible solutions.
(c) The geographical and geomagnetic longitude of the agonic line.
Assume a spherical Earth of radius 6370 km.

Main eld

137

(a) We calculate the geomagnetic constant from the magnetic moment m, the Earths
radius a, and the constant C107 H m1, by
B0

Cm
29 403 nT
a3

(b) Let us obtain rst the geomagnetic latitude from the inclination I  by
tan I  2 tan f ) f 47:5
If the dipole is tilted 15 away from the axis of rotation the latitude of the Geomagnetic
North Pole will be
fB 90  15 75:0
With these results we calculate the geographical latitude f solving the spherical triangle
(Fig. 71a). Applying the cosine rule,
cos90  fB cos90  f cos90  f sin90  f sin90  f cos D
sin fB sin f sin f cos f cos f cos D

79:1

To obtain the geographical latitude from this equation we can carry out a change of
variables, introducing two new variables (m, N ) such that
sin f m cos N
cos f cos D m sin N

79:2

From these equations we can calculate P and N:


cos f cos D cos D

) N 41:6
sin f
tan f
cos N
P
1:02
sin f

tan N

Substituting Equations (79.2) in Equations (79.1) we obtain


sin fB P cos N sin f P cos N cos f P sinf N
sin fB sin fB cos N

sinf N
m
sin f

f N 78:4 ) f 36:8
But another solution is also possible:
f N 180  78:4 258:4 ) f 60:0
The two solutions are correct and we dont have any additional information to choose one
or the other.
(c) The agonic line is the line where the declination is zero and this implies that the
point is on the great circle that contains the Geographic North Pole and the
Geomagnetic North Pole. So the geomagnetic longitude l  is zero or 180:
l 0 ) l lB 65 W
l 180 ) l lB 180 115 E

138

Geomagnetism

80. The Earths magnetic eld is produced by two dipoles of equal moment
(M Cm 9.43  109 nT m3) and polarity, forming angles of 30 and 45 with the
axis of rotation, and contained in the plane corresponding to the 0 meridian. Find
the potential of the total eld and the coordinates of the resulting magnetic North
pole, taking the Earths radius to be 6000 km.
The total potential at a point is the sum of the potentials of the two dipoles. If M is the
magnetic moment (M Cm), r is the distance from the dipoles centre, and y1 and y2
are the geomagnetic co-latitude relative to each dipole (Fig. 80), the total potential F is
given by
F F1 F2

M cos y1 M cos y2 M cos y1 cos y2




r2
r2
r2

80:1

We calculate the angles y1 and y2 (Fig. 71a) by


cos y1 sin fB1 sin f cos fB1 cos f cosl  lB1
cos y2 sin fB2 sin f cos fB2 cos f cosl  lB2
The geographical coordinates of the two Geomagnetic North Poles are given by
fB1 90  30 60 ; lB1 0
fB2 90  45 45 ; lB2 180

45

GNP
30
GMNP1

GMNP2

q2
q1

Fig. 80

80:2

Main eld

139

Substituting these values in Equations (80.2):


cos y1 sin fB1 sin f cos fB1 cos f cos l
cos y2 sin fB2 sin f  cos fB2 cos f cos l
Adding the two equations gives
cos y1 cos y2 sin fB1 sin fB2 sin f cos fB1  cos fB2 cos f cos l
and substituting in the equation of the potential (80.1)
M sin fB1 sin fB2 sin f cos fB1  cos fB2 cos f cos l
r2
p
p p



M
3 2 sin f 1  2 cos f cos l
F
2r2
F

If we call # the geographic co-latitude, # 90 f, then


p
p p


M
3 2 cos # 1  2 sin # cos l
F
2r2
The resulting magnetic North Pole, the point where the inclination I 90, due to the
combined effect of the two dipoles is given by
tan I

Z
H

and therefore at the magnetic Pole, H 0.


We derive the component H by taking the gradient of the potential F
p
 p p


M  3 2 sin # 1  2 cos # cos l
1 @F
X B#

r @#
2r3
p

1 @F M 1  2 sin l
Y Bl

r sin # @l
2r3
p
2
2
H X Y

Since the magnetic North Pole is contained in the plane corresponding to the 0 geographical meridian, then its longitude is either 0 or 180.
If the longitude is 0

i
p
M h p p
l 0 ) H 3  3 2 sin # 1  2 cos # 0
2r
# 8 172
But this result doesnt correspond to the north hemisphere. Then we must take the
geographical longitude 180:

i
p
M h p p
l 180 ) H 3  3 2 sin #  1  2 cos # 0
2r
# 8 ) fB 82

140

Geomagnetism

This is the correct result and the coordinates of the magnetic North Pole are
fB 82 ; lB 180
81. The Earths magnetic eld is produced by one dipole in the direction of the axis
of rotation (negative pole in the northern hemisphere) and another with the same
moment in the equatorial plane which rotates with differential angular velocity v
with respect to the points on the surface of the Earth (consider that the Earth doesnt
rotate). Its negative pole passes through the 45 E meridian at time t 0 and
completes a rotation with respect to that point in 24 hours. Consider a point of
geographical coordinates 45 N, 45 E.
(a) Calculate the magnetic eld components (Br, Bu, Bl) at that point.
(b) Illustrate graphically how each of them varies with local time.
(a) The total potential at a point on the surface of the Earth is the sum of the potentials
of the two dipoles (Problem 80, Equation 80.1):
F F1 F2

M cos y1 M cos y2 M cos y1 cos y2




r2
r2
r2

Dipole 1 is in the direction of the axis of rotation and so the geomagnetic co-latitude of the
point with respect to this dipole (Fig. 81a) is equal to the geographical co-latitude,
y1 90  f
cos y1 sin f

GMNP1
GNP

q1

P
q2
f
GMNP2

Fig. 81a

Main eld

141

Dipole 2 is on the equatorial plane (fB2 0) and rotates with respect to the points of
the surface. Owing to this rotation its geographical longitude lB2 changes with time t in
the form
lB2 ot 45
where o is the angular velocity, o 360 /T, T being the rotation period of 24 h.
The co-latitude y2 is
cos y2 sin fB2 sin f cos fB2 cos f cosl  lB2
Substituting the geographical coordinates of the negative geomagnetic equatorial Pole
(fB2, lB2):
cos y2 cos cosl  ot  45
Substituting in the equation for the potential
F F1 F2

M sin f cos f cosl  ot  45 


r2

If we consider the geographical co-latitude # 90  f, the potential is given by


F

M cos # sin # cosl  ot  45 


r2

We obtain the magnetic eld components (Br , By, Bl) at the point (#, l) by taking the
gradient in spherical coordinates of the potential F:
Br 

@F 2M cos # sin # cosl  ot  45 

@r
r3

B# 

1 @F M  sin # cos # cosl  ot  45 

r@#
r3

Bl 

1 @F
M sinl  ot  45

r sin # @l
r3

Substituting the values # 45 , l 45 , B0 M/a3, gives


2B0
Br  p 1 cos ot
2

B0
B#  p 1  cos ot
2
Bl B0 sin ot

(b) The variation of each component with local time is shown in Fig. 81b

142

Geomagnetism

Bl

Br,q,l
B0

Br

1
B

10

15

20

t (h)

Fig. 81b

Magnetic anomalies
82. Calculate the magnetic anomaly created by a magnetic dipole buried at depth d,
arbitrarily oriented, at an angle to the vertical of a. The negative pole is upwards.
Consider a point P with coordinates (x, z), where x is measured along the horizontal from
the projection of the centre of the dipole and z is the vertical from the reference level (the
Earths surface). The position vector r forms an angle b to the vertical (Fig. 82). The
anomalous magnetic potential created by the dipole for this point is
F

Cm cosa b Cmsin b cos a cos b sin a

r2
r2

where (Fig. 82)


zd
cos b q
x2 z d2
x
sin b q
x2 z d2

82:1

143

Magnetic anomalies

x
z

d
b

+
Z

Fig. 82

Substituting in Equation (82.1) we obtain


F

Cmz d cos a  x sin a


h
i3=2
x2 z d2

To calculate the magnetic anomaly DB:


B rF
The vertical component of the magnetic eld anomaly, taking the z-coordinate positive
downward, is
h

i
2
2
Cm
x

d
cos
a

3z

d
cos
a

x
sin
a

@F

Z
h
i5=2
@z
x2 z d2
For points on the Earths surface (z 0)
Z

Cmx2  2d 2 cos a 3dx sin a


x2 d 2 5=2

82:2

The component of the magnetic anomaly in an arbitrary horizontal direction x for the
Earths surface points (z 0) is given by
X 

@F Cm2x2  d 2 sin a  3dx cos a

@x
x2 d 2 5=2

82:3

144

Geomagnetism

83. Calculate the magnetic anomaly produced at a point with geographical


coordinates 38 N, 30 W by a horizontal dipole buried at a depth of 10 m with
Cm 5  105 T m3 which is in the vertical plane of geographical east, and with the
negative pole to the west. Also calculate the total values of the eld in the NS, EW, and
vertical directions, and total eld F, as well as the variations in the magnetic declination and inclination due to the existence of the dipole. Consider the Earths magnetic
eld to be produced by a centred dipole with North Pole at 72 N, 30 W, and with
B0 32 000 nT.
We calculate the horizontal and vertical components of the magnetic anomaly from
Equations (82.2) and (82.3), taking a 90 because the dipole is horizontal and x 0
because the dipoles centre is beneath the point (Fig. 83a). We call DX and DY the
horizontal components in the NS and EW directions, respectively. Because the dipole is
on the vertical eastwest plane, the northsouth component DX 0,
Z 0
Y

Cm2x2  d 2 sin a  3dx cos a


x2

d 2 5=2

Cm
d3

Substituting
Cm 5  105 T m3
d 10 m
we obtain
Y 50 nT
jBj Y 50 nT
X 0
P

a
+

Fig. 83a

145

Magnetic anomalies

GMNP

GNP

fB

P
f
f

Fig. 83b

To calculate the components of the magnetic anomaly in the direction of the Earths
magnetic eld, F, and their horizontal component, H, we need to determine the magnetic
declination and inclination at the point:
H X cos D Y sin D
F H cos I  Z sin I 

83:1

Because the point has the same longitude as the Geomagnetic North Pole (Fig. 83b),
f 90  fB  f 56:0
D 0
The inclination is given by
tan I  2 tan f ) I  71:4
Substituting these values in Equations (83.1) we obtain
H 0
F 0
The total value of the eld in the NS, EW, and vertical directions, and total eld F are
XT X  X
YT Y  Y
ZT Z  Z
FT F  F

146

Geomagnetism
The vertical Z  and horizontal H  components of the geomagnetic eld are given by
Z  2B0 sin f 53 058 nT
H  B0 cos f 17 894 nT
The NS (X  ) and horizontal EW (Y  ) components of the geomagnetic eld and its
magnitude F  are given by
X  H  cos D H  17 894 nT
Y  H  sin D 0
p
F  H 2 Z 2 55 994 nT

We nally obtain that the total eld components are

XT 17 894 nT
YT 50 nT
ZT 53 058 nT
FT 55 994 nT
The variations in magnetic declination and inclination due to the presence of the buried
dipole are
YT
) D0 0:02
tan D0
XT
D  D0 0:02
tan I 0

ZT
) I 0
71:4 I 
HT

84. Buried at a point with magnetic latitude 30 N and a depth of 50 m is a horizontal
magnetic dipole with Cm 107 nT m3 with the positive pole to the geographical
north.
(a) Calculate DF if B0 30 000 nT and the declination at that point is 15 . Find the
ratio DF /F.
(b) How far from the dipoles centre along the northsouth line will the dipole eld
strength be in the same direction as that of the Earth (take D  0  ).
(a) The component of the magnetic anomaly DF in the direction of the Earths
magnetic eld (the total eld anomaly) is given by
F H cos I  Z sin I 

84:1

We rst calculate the components in the geographical directions of the magnetic anomaly
produced by the buried dipole using Equations (82.2) and (82.3) of Problem 82, substituting a 90 because the dipole is horizontal, and x 0 because the dipoles centre is
beneath the point. In this problem DY 0 because the dipole is on the geographical north
south vertical plane (Fig. 84). Then

147

Magnetic anomalies

a
+

Fig. 84

X 

Cm
d3

Y 0
Z 0
Substituting the values
Cm 107 nT m3
d 50 m
we obtain
X 80 nT
Substituting D  15, the component of the magnetic anomaly in the direction of the
horizontal component H of the Earths magnetic eld is
H X cos D 77 nT
At a point of magnetic latitude f  30 the magnetic inclination is
tan I  2 tan f ) I  49:1
Substituting in Equation (84.1), the total eld anomaly is
F 50 nT
To calculate the geomagnetic eld F  we rst obtain the components H  and Z  :
Z  2B0 sin f 30 000 nT
H  B0 cos f 25 981 nT
p
F  H 2 Z 2 39 686 nT

148

Geomagnetism

The ratio of the total eld anomaly and the Earths total magnetic eld is
F
1:26  103
F
(b) If the dipole eld strength is in the same direction as that of the Earth then the
inclination I 0 due to the dipole is equal to that of the Earths eld I  , where
Z
H
Z
tan I  
H
tan I 0

Assuming D  0, then
H E
If we substitute in Equations (82.2) and (82.3) of Problem 82, the angle a 90 because
the dipole is horizontal, we obtain
3Cmdx
Z
5=2
2
x d2

Cm2x2  d 2
X
5=2
x2 d 2

We have changed the sign of the vertical component because the negative pole is toward
the south.
Applying the condition, tan I 0 tan I  , we obtain
Z
Z
Z


H X H
3Cmdx

5=2
2
x d2
3dx
Z
 2

Cm 2x  d 2 2x2  d 2 H 

5=2
x2 d 2

2Z  x2  3dH  x  Z  d 2 0
Substituting the values
d 50 m
Z  30 000 nT
H  25 981 nT
and solving the equation, we obtain
x1 80 m
x2 15 m

We have two solutions: a point 80 m to the north from the surface projection of the dipoles
centre and another 15 m to the south.

149

Magnetic anomalies
85. Located at a point with geocentric geographical coordinates 45 N, 30 W, at a
depth of 100 m, is a dipole of magnetic moment Cm 1 T m3, tilted 45 from the
horizontal to true north, with the negative pole to the north and downwards. At this
point on the surface, the following magnetic eld values were observed (in nT):
F 55 101; H 12 413; DF 1268; DH 547.
Determine:
(a) At the indicated point, the main eld components X  , Y , Z  .
(b) At the indicated point, the deviation of the compass with respect to geomagnetic
north.
(c) The geocentric geographical coordinates of the North Pole of the Earths dipole.
Precision 1 nT.
(a) We calculate rst the magnetic anomaly produced by the dipole, applying Equations (82.2) and (82.3) of Problem 82, substituting a 225 and x 0. The
horizontal component is in the NS direction (DX) (Fig. 85a)
2Cm cos a
1414 nT
d3
Cm sin a
X
707 nT
d3
Y 0
Z

85:1

To calculate the declination we use the equation


H X cos D ) cos D

H
X

D 39:3

d
+
a

45

Fig. 85a

150

Geomagnetism

To obtain the Earths main eld we eliminate the buried dipole contribution from the
observed values:
F  F  F 56 369 nT
H  H  H 11 866 nT
q
Z  F  2  H  2 55 106 nT

and the X  and Y  components

X  H  cos D 9182 nT
Y  H  sin D 7516 nT
(b) The observed declination is given by
tan D0

Y
Y  Y

) D0 37:2
X X X

The deviation of the compass due to the buried dipole with respect to geomagnetic north is
D0  D 2:1
(c) We calculate rst the geomagnetic latitude of the point from the vertical and
horizontal components:
Z  2B0 sin f
H  B0 cos f
Z
2H 

f 66:7

tan f

With this value, the declination D  and the geographical coordinates of the point (f, l), we
can solve the spherical triangle (Fig. 85b) and obtain the geographical coordinates of the
Geomagnetic North Pole:
GNP
90 fB
l lB
90 f

180 l
GMNP
q = 90 f

Fig. 85b

151

Magnetic anomalies

cos90  fB cos90  f cos90  f sin90  f sin90  f cos D


sin fB sin f sin f cos f cos f cos D
fB 60:0
cos90  f cos90  fB cos90  f sin90  fB sin90  f cosl  lB
sin f sin fB sin f cos fB cos f cosl  lB
sin f  sin fB sin f
cos fB cos f
l  lB 30:0

cosl  lB

The correct solution is the negative one because a positive value of the declination implies
that the point is to the west of the Geomagnetic North Pole:
D  > 0 ) l  lB < 0
lB 0
86. Located at a point with geographical coordinates 45 N, 30 W, at a depth of 100 m,
is a dipole of magnetic moment Cm 1 T m3, inclined 45 to the vertical towards the
south, with the positive pole upwards, and in the geographical northsouth vertical
plane. The Earths dipole has its north pole at 60 N, 0 E and B0 30 000 nT.
Calculate:
(a) The values of Z, H, F at the given point.
(b) Where does the compass point to at that same point?
(a) We calculate rst the geomagnetic latitude corresponding to the point by
sin f sin fB sin f cos fB cos f cosl  lB
f 66:7
From this value we obtain the geomagnetic components Z  , H  and the total main eld F  :
Z  2B0 sin f 55 107 nT
H  B0 cos f 11 866 nT
q
F  H  2 Z  2 56 370 nT

The geomagnetic declination D  is given by

 cos fB sinl  lB
cos f
D 39:2
sin D

and the geomagnetic inclination I  by


tan I  2 tan f ) I  77:8

152

Geomagnetism

The magnetic anomaly created by the dipole buried at depth d is given by Equations (85.1)
of Problem 85. Substituting Cm 1 T m3, d 100 m, and a 45, we obtain
2Cm cos a
1414 nT
d3
Cm sin a
X
707 nT
d3
Z

Y 0
The eld anomalies DH and DF are given by
H X cos D 548 nT
F H cos I  Z sin I  1266 nT
Finally, the observed values are
Z Z  Z 53 693 nT
F F  F 55 104 nT
H H  H 12 414 nT
(b) To calculate in what direction the compass points we need the value of the observed
declination D including the effects of the geomagnetic eld and the buried dipole:
tan D0

Y
Y  Y

X X X

Y  H  sin D 7500 nT
X  H  cos D 9195 nT
D0 37:1
87. Located at a point on the Earth with geographical coordinates 45 N, 30 E, at a
depth of 100 m, is a dipole of magnetic moment Cm 107 nT m3, tilted 45 to the
vertical towards the south, with the positive pole downwards, and contained in the
plane of true north. The Earths eld is produced by a centred dipole tilted 30 from
the axis of rotation in the plane of the 0 meridian, with B0 30 000 nT. Calculate
the total values of F, Z, and H observed at the point of the surface above the centre of the
buried dipole.
We rst calculate the geographical coordinates of the Geomagnetic North Pole and the
geomagnetic latitude
fB 90  30 60
lB 0
sin f sin fB sin f cos fB cos f cosl  lB
f 66:7

153

Magnetic anomalies
The geomagnetic eld components Z  , H  and the total main eld F  are given by
Z  2B0 sin f 55 107 nT
H  B0 cos f 11 866 nT
q

F H  2 Z  2 56 370 nT

We calculate the geomagnetic declination D  by

 cos fB sinl  lB
cos f
D 39:2

sin D

The inclination I  is given by


tan I  2 tan f ) I  77:8
We obtain the magnetic anomaly produced by the buried dipole applying Equations (82.2)
and (82.3) of Problem 82, substituting a 45 and x 0. The horizontal component is in
the NS direction (DX):
2Cm cos a
14 nT
d3
Cm sin a
7 nT
X
d3
Y 0
Z

The eld anomalies DH and DF are given by


H X cos D 5 nT
F H cos I  Z sin I  13 nT
Finally the observed values are
Z Z  Z 55 121 nT
F F  F 56 383 nT
H H  H 11 861 nT
88. Buried at a point with geographical latitude 20 N and the same longitude as the
geomagnetic pole, at a depth of 200 m, is a sphere of 50 m radius of material with
magnetic susceptibility 0.01. The Earths eld is produced by a centred dipole tilted
10 from the axis of rotation and magnetic moment M 1030 g cm3 (Earths radius:
6000 km). Calculate:
(a) The anomaly produced by induced magnetization in the sphere at a point on the
Earths surface above the centre of the sphere. Give the vertical and horizontal
components in units of nT.
(b) The total anomaly for a point on the Earths surface 100 m south of the above
point.

154

Geomagnetism

GNP

GMNP
q = 90 f

10

P
20

Fig. 88a

(a) We rst calculate the geomagnetic co-latitude (y) and latitude (f  ) of the point,
knowing that it is in the same meridian as the Geomagnetic North Pole (Fig. 88a):
y 90  f 90  10  20 60
f 30
The geomagnetic eld is given by
M
4630 nT
a3
Z  2B0 sin f 4630 nT
B0

The inclination is given by

H  B0 cos f 4009 nT
q
F  H  2 Z  2 6124 nT
tan I  2 tan f ) I  49

The magnetic anomaly created by a sphere is the same as the anomaly created by a
magnetic dipole oriented in the same direction as the geomagnetic eld, that is, tilted
90 I  to the vertical and with the negative pole upwards (Fig. 88b). So we use Equations
(82.2) and (82.3) taking a 41 and x 0, but we change the sign of Equation (82.3)
because the negative pole is toward the north. The horizontal component is DX DH
because the dipole is in the magnetic north-vertical plane and DY 0:
2Cm cos a
d3
Cm sin a
H
d3
Z

155

Magnetic anomalies

49

GN
MN

41

Fig. 88b

To calculate Cm we use the magnetic susceptibility w and the volume V of the sphere:
Cm wF  V 3:2  107 nT m3
Substituting this value in the equation of the components of the magnetic anomaly we
obtain
Z

2Cm cos a 2  3:2  107  cos 41

6 nT
8  106
d3

X

Cm sin a 3:2  107  sin 41

3 nT
d3
8  106

(b) The anomaly created by the sphere at a point at a distance x 100 m to the south of
the above point is given by


Cm x2  2d 2 cos a 3dx sin a
0:8 nT

5=2
x2 d 2


Cm 2x2  d 2 sin a  3dx cos a
3:3 nT
H

5=2
x2 d 2
Z

The total magnetic eld anomaly is therefore

F H cos I  Z sin I  3:4 nT

156

Geomagnetism
89. Buried at a point on the Earth at magnetic latitude 45 N, at a depth d, is a vertical
dipole of magnetic moment M (Cm) with negative pole upwards. If M/d3 10B0 (B0 is
the geomagnetic constant of the main eld) calculate how far along the magnetic
meridian the direction of the buried dipoles eld will coincide with that of the Earth
(the terrestrial dipole eld).
First we calculate the geomagnetic eld components and the inclination by
p
Z  2B0 sin f 2B0 nT
p
2
B0 nT
H  B0 cos f
2
Z
tan I   2 ) I  63:4
H
The components of the magnetic eld created by the dipole are given by Equations (82.2)
and (82.3) of Problem 82, putting a 0:


Cm x2  2d 2
Z 
5=2
x2 d 2
Cm3dx
H 
5=2
2
x d2

If the buried dipoles eld coincides with that of the Earth the magnetic inclinations due to
both have to be equal and so
Z
tan I  2
H

p
x2 6xd  2d 2 0 ) x d 3  28

Of the two solutions, x 2.3d and x 8.3d, only the positive corresponds to the equal
direction of the two elds.

External magnetic eld


90. The Earths magnetic eld is produced by two dipoles of equal moment and
polarity that are at an angle of 60 to each other, with the bisector being the axis of
rotation. The dipoles are contained in the plane of the 0 geographical meridian.
(a) Calculate the potential on points of the Earths surface as a function of geographical coordinates f and l.
(b) At what points on the surface are the magnetic poles located?
(c) What form would the external eld have in order to annul the internal eld at the
magnetic equator?

External magnetic eld

157

30
GMNP2

GNP

GMNP1
q2

q1

Fig. 90

(a) The total potential at a point is the sum of the potentials of the two dipoles (see
Equation 80.1 of Problem 80):
F F1 F2

M cos y1 M cos y2 Mcos y1 cos y2




r2
r2
r2

We calculate the geomagnetic co-latitudes by Equation (71.3) of Problem 71:


cos y1 sin fB1 sin f cos fB1 cos f cosl  lB1
cos y2 sin fB2 sin f cos fB2 cos f cosl  lB2

90:1

The geographical coordinates of the North Pole of each dipole are given by (Fig. 90)
fB1 fB2 90  30 60
lB1 0
lB2 180
Substituting these values in Equation (90.1):
p
3
1
sin f cos f cos l
cos y1
2
2
p
3
1
sin f  cos f cos l
cos y2
2
2
Adding the two equations:
cos y1 cos y2

p
3 sin f

158

Geomagnetism

Substituting in the equation of the potential:


p
 3M sin f
F
r2
We note that this p
is the same potential as that produced by only a centred dipole with
magnetic moment 3M oriented in the direction of the rotation axis, because the geomagnetic co-latitude is 90 f.
(b) Bearing in mind the last results, we calculate the inclination by
tan I 2 cot90  f 2 tan f

90:2

The magnetic poles are the points on the surface of the Earth where the value of the
inclination is equal to  90:

f 90
I  90 )
f 90
Therefore the magnetic poles coincide with the geographical poles.
(c) We derive the main eld by taking the gradient of the potential F. The components are
p
@F 2 3M sin f

@r
r3
p
1 @F  3M cos f

Bf
r @f
r3
1 @F
Bl 
0
r cos f @l
Br 

At the magnetic equator the inclination is null (I 0) and according to Equation (90.2) the
latitude is null too (f 0). Substituting in the last equations:
Br 0

p
 3M
Bf
r3
Bl 0
The external magnetic eld to annul out the internal eld is therefore
p

3M
Be 0; 3 ; 0
r
91. The Earths magnetic eld is formed by a centred dipole with northern geomagnetic pole at 60 N, 0 E and B0 32 000 nT and a uniform external eld from the Sun
of 10 000 nT parallel to the equatorial plane.
(a) For a point at coordinates 60 N, 60 W, calculate the components X, Y, Z of the
total eld, and the values of D and I.
(b) How do D and Z of the total eld vary throughout the day with local time t?

External magnetic eld

159

(a) We calculate the geomagnetic main eld components Z  , H  obtaining the


geomagnetic latitude given by (71.3):
sin f sin fB sin f cos fB cos f cosl  lB
f 61
Z  2B0 sin f 55 976 nT
H  B0 cos f 15 514 nT
To calculate the NS (X  ) and EW (Y  ) components we need the geomagnetic declination
D  (71.2):
 cos fB sinl  lB
cos f


D 63
X  H  cos D 7043 nT
sin D

Y  H  sin D 13 823 nT


The external eld is parallel to the equatorial plane and has a diurnal period (o 2p/24)
because it comes from the Sun. We assume that at local time t 0 the Sun is at the points
meridian. If we denote by N the modulus of the external eld (N 10 000 nT) and bearing
in mind Fig. 91a (representation of the plane parallel to the equator that contains the point)
we have at time t 6 0
Y Bel N sin ot
The radial and tangential components (Fig. 91b) are

(a)
P
N coswt
wt

Fig. 91a

B e

160

Geomagnetism

GNP

X = Bf

Ncost
P

z = Br

Fig. 91b

Z Br N cos ot cos f


X Bf N cos ot sin f
The total magnetic eld is the sum of the two contributions:
ZT 55 976 5000 cos ot nT
XT 7043 8660 cos ot nT
YT 13 823 10 000 sin ot nT
The declination D and inclination I are given by
tan D
tan I

YT 13 823 10 000 sin ot

XT
7043 8660 cos ot
ZT
ZT
p
HT
XT2 YT2

(b) To see how D and Z vary during the day with local time t we substitute several
values for t, obtaining the values in the table:
t (h)

Z (nT)

D ()

0
6
12
18

60976
55976
50976
55976

41
28
83
73

External magnetic eld

161

92. The Earths magnetic eld is formed by a centred dipole of moment m and
north pole 60 N, 0 E, and a uniform external eld of magnitude N B0/4 (B0 is
the geomagnetic constant of the internal eld) parallel to the axis of rotation.
Determine:
(a)
(b)
(c)
(d)

The total potential at any point.


The coordinates of the boreal magnetic pole.
The magnetic declination at the point 45 N, 45 E.
The angle along the meridian between that point and the magnetic equator.
(a) The total potential (F) is the sum of two contributions: the main (internal) eld
(Fi) and the external eld (Fe):
F Fi Fe

The main eld is formed by a centred dipole of moment m so the potential is given by
Fi

Cm cos y
r2

We calculate the geomagnetic co-latitude y 90 f  at a point with geographical


coordinates (f, l) by (71.3):
p
1
3
cos y sin fB sin f cos fB cos f cosl  lB
sin f cos f cos l
2
2

GNP

ze

Xe
f

Fig. 92a

Fe

162

Geomagnetism

The external eld is parallel to the axis of rotation; therefore its components are in the
vertical and NS directions (Fig. 92a). If we call # 90  f the geographical co-latitude,
the components are given by
Ber N cos # N sin f
Be# N sin # N cos f
Z e Ber N sin f
X e Be# N cos f
Bearing in mind that Be Fe the potential for the external eld is
Fe Nr cos # Nr sin f
Therefore the total potential is given by
p

Cm 23 sin f 12 cos f cos l
Nr sin f
F
r2
(b) At the magnetic boreal pole the inclination is I 90 and the horizontal eld
components H, X, Y are given by
Z
)H 0
H

p
X 0
2
2
H X Y )
Y 0
tan I

We obtain the components X and Y by taking the gradient of the potential


Be rFe
We bear in mind the relations
@F
@F

@y
@f
X By
and obtain that the components are given by
p

3
1
Cm
cos
f

sin
f
cos
l
2
2
1 @F

X 
 N cos f 0
r @f
r3
1 @F
Cm
 3 sin l 0 ) l 0
Y 
r cos f @l
2r
Substituting in the equation of component X (92.1) the values
l0
ra
Cm
a3
N B0 =4

B0

92:1

External magnetic eld

163

GNP

GMNP

fB
P

f
l
feq

Fig. 92b

we obtain the geographical latitude of the boreal magnetic pole


p
2 31
tan fBM
) fBM 51
2
Therefore the geographical coordinates of the magnetic boreal pole are 51 N, 0 . Notice
that this is different from the geomagnetic pole.
(c) The declination is given by
Y
tan D
X

B0
sin l
2

3
1
B0
cos f  sin f cos l  cos f
2
2
4
 sin l

p
1
cos f  sin f cos l
3
2
B0

Substituting the geographical coordinates of the point (45 N, 45 E) we obtain


D 62
(d) We call feq the angle between the geographical and magnetic equators at longitude
l 45 (Fig. 92b). Then the angle to calculate will be feq 45 .
To calculate feq we take into account that at the magnetic equator the vertical component is Z 0:
p

@F 2CM
1
3
3
Z Br
sin feq cos feq cos l N sin feq 0
@r
r
2
2
4 cos l
tan feq p
4 31

164

Geomagnetism

Substituting l 45 in the equation we obtain


feq 20
Then the angle we are asked for is 45 20 25.
93. The internal magnetic eld is formed by two orthogonal dipoles of equal moment
M, one of them in the direction of the axis of rotation and the other contained in the 0
meridian on the equator. There is also an external eld of constant intensity N B0/4
(B0 is the geomagnetic constant of the internal eld) and lines of force parallel to the
axis of rotation.
(a) Calculate the potential of the total eld for points on the surface.
(b) At which latitude is Z maximum on the 0 meridian?
(a) The potential of the total eld is the sum of the two potentials of the internal
dipoles and the potential of the external eld:
M cos y1 M cos y2

Fe
r2
r2
Mcos y1 cos y2

Fe
r2

F F1 F2 Fe

In this equation r is the distance from the dipoles centre (the Earths centre), y1 and y2 are
the co-latitudes relative to each dipole, and M Cm.
Dipole 1 is on the direction of the axis of rotation and so the geomagnetic co-latitude of
the point with respect to this dipole is
y1 90  f
cos y1 sin f
Dipole 2 is in the equatorial plane (fB2 0) and contained in the 0 meridian so the
geomagnetic co-latitude y2 is given by (71.3)
cos y2 sin fB2 sin f cos fB2 cos f cosl  lB2
fB2 0
lB2 0
cos y2 cos f cos l
The equation for the potential of the external eld is the same as that of Problem 92:
Fe Nr sin f
Therefore the potential of the total eld is given by
F

M sin f cos f cos l


Nr sin f
r2

External magnetic eld

165

(b) We obtain the component Z by taking the vertical component of the gradient of the
potential (B F)
Z Br

@F 2M sin f cos f cos l

N sin f
@r
r3

To calculate the maximum of Z on the 0 meridian we substitute l 0 and apply the


condition that the rst derivate with respect to the latitude is null:
@Z 2M
3 cos f  sin f N cos f 0
@f
r
At the Earths surface r a and we know that
B0

M
a3

Substituting this constant and solving the equation we determine the latitude at which the
Z component is maximum:
2B0 cos f  sin f

B0
cos f 0 ) f 48
4

94. The internal eld has its northern geomagnetic pole at the coordinates 60 N, 0 E,
and B0 30 000 nT. At a point with coordinates 30 N, 45 W, one observes an
increase of 7.7 in the value of the declination from 00:00 h to 09:00 h. There is
known to be an external eld parallel to the Earths axis of rotation in the direction
from N to S which is null at 00:00 h and maximum at 12:00 h local times. Calculate:
(a) The components of the internal and external elds.
(b) The difference in the inclination at 00:00 h and 09:00 h.
(c) The maximum value of the declination during the day.
(a) To calculate the geomagnetic main eld intensity components we obtain rst the
geomagnetic latitude (Equation 71.3):
sin f sin fB sin f cos fB cos f cosl  lB
f 47:7
The declination and inclination are given by (71.2):
 cos fB sinl  lB
) D 31:7
cos f
tan I  2 tan f ) I  65:5

sin D

So the geomagnetic main eld components are


Z  2B0 sin f 44 378 nT
H  B0 cos f 20 190 nT
X  H  cos D 17 178 nT
Y  H  sin D 10 609 nT

166

Geomagnetism

The external eld is parallel to the axis of rotation so its components are in the vertical and
NS directions. This eld is null at 00:00 h and maximum at 12:00 h local time (period T
24 h). Its components are given by
N
1  cos ot sin f
2
N
X e  1  cos ot cos f
2
Ye 0
Ze

He X e
o

2p 2p

T
24

(b) To calculate the difference in the inclination we obtain rst the value of N, bearing
in mind the time variation of the declination. The observed declination as a
function of time is given by
tan D

Y
Y Ye


X X Xe

Y
X 

For t 0 h:
tan D1

N
1  cos ot cos f
2

Y
tan D ) D1 D 31:7
X

Since we know the change in declination between 0 h and 9 h, we nd the declination at


9 h, D2:
D2  D1 7:7 ) D2 39:4
We know that at t 9 h
Y

tan D2
X 

N
1  cos ot cos f
2

0:82

Solving for N we obtain


2
Y

X 
5766 nT
N
1  cos ot cos f
tan D2
The magnetic inclination is given by
N
Z  1  cos ot sin f
Z
Z Ze
2

tan I 
N
H H He

H  1  cos ot cos f
2
At t 0 h:
I1 I  65:5
At t 9 h:
I2 71:2

External magnetic eld

167

The difference in the inclination is therefore


I2  I1 5:7
(c) The declination is given by
Y

tan D
X 

N
1  cos ot cos f
2

94:1

The maximum value is at 12 h because at that time the external eld has the maximum
value. Substituting ot p and the values obtained for X  , Y , N, and f:
Dmax 41:0
95. At a point on the Earth with coordinates 45 N, 45 E, measurements are made of
the magnetic eld components at 00:00 h and 12:00 h in nT with a 2 nT precision:
0h
X 20 732 Y 2500 Z 57 768
12 h X 24 267 Y 2500 Z 54 232
It is known that the modulus of the magnetic eld intensity has a harmonic diurnal
variation, and that the geomagnetic pole is on the zero meridian. Calculate:
(a) The moment and coordinates of the main eld dipole.
(b) Expressions for the potential and components of the external eld.
(Earths radius a 6400 km, m0 4p107 kg m s2 A2).
(a) To calculate the moment and coordinates of the main eld dipole we need to obtain the
geomagnetic main eld intensity components. The observed values are equal to the
sum of the geomagnetic main eld (X  , Y , Z  ) and the external eld (X e, Y e, Z e):
X X Xe
Y Y Ye
Z Z Ze
The geomagnetic main eld is constant but the external eld changes with time. So if we
denote by (X0, Y0, Z0) and (X12, Y12, Z12) the observed values at 0 h and at 12 h respectively
then the differences are due to the variations of the external eld:
e
 X0e
X12  X0 3535 nT X12
e
Y12  Y0 0 nT Y12
 Y0e
e
Z12  Z0 3536 nT Z12
 Z0e

We notice that the Y component doesnt vary, which implies that the component Y e is zero,
and so the external eld is parallel to the axis of rotation. We also notice that the NS
component increases in the time interval between 0 h and 12 h, while the vertical
component diminishes, which implies that the polarity of the external eld is inverted
with respect to that of the main eld.
The modulus of the magnetic eld intensity has a harmonic diurnal variation and
increases with time. Therefore the components of the external eld are (Fig. 95a)

168

Geomagnetism

Fe

GNP

Xe

Ze
P

Fig. 95a

X e N 1  cos ot cos f
Z e N 1  cos ot sin f
2p
2p
o

T
24h
We notice that at time t 0 the external eld is null and so
e
X12
3535 nT
e
Y12 0 nT
e
Z12
3536 nT

Therefore we calculate the main eld components by


e
X  X12  X12
20 732 nT
e
Y  Y12  Y12
2500 nT
e
57 768 nT
Z  Z12  Z12
q

H X  2 Y  2 20 882 nT

If we consider the centred magnetic dipole model, the vertical and horizontal eld
components are given by the equations
Z  2B0 sin f
H  B0 cos f

External magnetic eld

169

GNP
90 fB
l lB
90 f

180 l
GMNP
q = 90 f

Fig. 95b

From these equations we obtain the geomagnetic latitude (f  ) and the geomagnetic
constant B0:
tan f
B0

Z
) f 54:1
2H 

Z
35 657 nT
2 sin f

From B0 we calculate the magnetic moment:


B0

m0 m
) m B0 107 a3 9:3  1022 A m2
4p a3

The longitude of the Geomagnetic North Pole is lB 0 and we calculate the latitude fB
from the spherical triangle (Fig. 95b), but obtaining rst the declination from the X  and
Y  components:
tan D

Y
) D 6:9
X

Applying the cosine rule:


cos90  fB cos90  f cos90  f sin90  f sin90  f cos D
sin fB sin f sin f cos f cos f cos D
fB 79:9
(b) We obtain the radial and transverse components of the external eld from the
vertical and NS components:
@Fe
@r
1 @Fe
e
e
Bf X N 1  cos ot cos f
r @f
Ber Z e N 1  cos ot sin f

170

Geomagnetism

Therefore the potential of the external eld is given by


Fe Nr1  cos ot sin f
where
N

e
Z12
2500 nT
2 sin f

96. The Earths main magnetic eld is that of a centred dipole of moment M (M
Cm) in the direction of the axis of rotation, and the external eld is produced by
electric currents of intensity J circulating in a clockwise sense in a ring in the plane of
the ecliptic at a distance of 10 Earth radii around the Earth.
(a) Express the potential of the total eld and the components Br and B on the
Earths surface for l 0 .
(b) If the inclination of the ecliptic is 30 , and the external eld strength is N M /
4R3, what is the latitude of the northern magnetic pole?
(a) The potential of the total eld is the sum of the potentials of the dipole and of the
external eld:
FT F Fe

M cos y
Fe
r2

where r is the distance from the dipoles centre (Earths centre) and y is the geomagnetic
co-latitude. The dipole is in the direction of the axis of rotation and so the geomagnetic colatitude is equal to the geographic co-latitude:
y 90  f
cos y sin f
To calculate the potential of the external eld we know that it is produced by electric
currents of intensity J circulating in a clockwise sense at a distance of 10 Earth radii around
the Earth. These electric currents produce, at remote points, a magnetic dipolar eld whose
modulus is m0J/2R, J being the current intensity and R the radius of the circular currents;
the dipole is oriented in the direction of the ecliptics axes with the negative pole in the
southern hemisphere, because the currents are clockwise. So the potential of the external
eld is given by
Fe

Cme cos y2
r2

In this equation me Jp100a2, a is the Earths radius, and y2 is the angle between the axes
of the circular currents and the direction of the point from the negative pole (Fig. 96)
Therefore
y2 180  y e
cos y2  cosy e
where e is the angle between the axes of the circular currents and the axes of rotation of the
Earth.

External magnetic eld

171

GNP

q
e

q2

Fig. 96

The potential of the external eld is


Fe

Cme cosy e
r2

and the total potential is given by


FT 

M cos y Cme cosy e

r2
r2

We calculate the components of the main eld intensity by taking the gradient of the
potential:
@F 2M cos y

@r
r3
1 @F
M sin y
By 

r @y
r3
Br 

The magnitude of the external eld is given by


Be

m0 J
N
4 10a

The radial and tangential components are


Ber N cosy e
Bey N siny e

172

Geomagnetism

The total eld is the sum of the two contributions:


2M cos y
N cosy e
r3
M sin y
By 
N siny e
r3
Br 

(b) We know that the Earths dipole is in the direction of the axis of rotation and y is
the geographical co-latitude (Fig. 96). At the magnetic North Pole the total eld is
vertical and the tangential component is By 0:
By 

M sin y
N siny e 0
r3

The values of N and e are known:


M
M
) 3 4N
3
4r
r
e 30

From these values we calculate y, the geographical co-latitude of the magnetic North Pole,
By 4N sin y N siny 30 0
p
1
3
sin y cos y 0 ) y 9
 4 sin y
2
2
97. Two spherical planets of radius a and separated by a centre-to-centre distance of 4a
orbit around each other and spin in the equatorial plane. Each has a magnetic eld
produced by a centred dipole in the direction of the axis of rotation, with the positive
pole in the northern hemisphere and B0 10 000 nT. Determine the components X, Y, Z,
D, and I of the total magnetic eld at the North Pole of one of the planets (precision 1 nT).
The total magnetic eld in each planet is the sum of its main eld and the external eld
created by the other planet. To determine the main eld of either of the planets we need the
geomagnetic latitude, which is positive toward the negative pole, in this case, the South
Pole. Therefore the geomagnetic latitude and the components of the main eld at the North
Pole are
f 90
Z  2B0 sin f 20 000 nT
H  B0 cos f 0
X Y 0
The external eld at one of the planets is created by the main eld of the other planet and
corresponds to that of a magnetic dipole. Its components are (Fig. 97)
2Cm cos y
r3
Cm sin y
Bey
r3
Ber

External magnetic eld

173

B er
q 90

+
r

B e
q 90
a

a
4a

Fig. 97

At the North Pole r


(Fig. 97)

p
17a 4:12a and we calculate the geomagnetic co-latitude y by
a
1

4a 4
y 14 90 104
tany  90

Substituting these values in the equations for the radial and tangential components with
respect to the planet producing the external eld:
2Cm cos y Cm cos y B0 cos y

69 nT
r3
35a3
35
Cm sin y Cm sin y B0 sin y
Bey
139 nT

r3
70a3
70
Ber

From this value we calculate the vertical and horizontal components (Fig. 97):
Z e Ber cos180  y Bey cosy  90 Ber cos y Bey sin y 118 nT
H e Ber sin180  y Bey siny  90 Ber sin y  Bey cos y 33 nT
X e H e 33 nT
Ye 0
The components of the total eld are nally
Z Z  Z e 19882 nT
H H  H e 33 nT
X X  X e 33 nT
Y Y Ye 0
Y
tan D ) D 0
X
Z
tan I ) I 89:9
H

174

Geomagnetism

Main (internal), external, and anomalous magnetic elds


98. At a point with geographical coordinates 40 N, 45 E, measurements are made of
the magnetic eld components, obtaining the values (in nT):
At 06:00 h: X 19 204; Y 0; Z 38 195
At 12:00 h: X 11 544; Y 0; Z 44 623
Buried at a depth of 20 m below this point is a dipole of magnetic moment Cm 0.01
T m3, oriented in the NS plane at an angle of 45 with the vertical towards the south,
and the positive pole upwards. Given that the external eld at 12:00 h is twice that at
06:00 h, determine:
(a) The geomagnetic constant Bo and the coordinates of the northern geomagnetic pole.
(b) The magnitude and direction of the external eld. How does the magnitude of the
external eld vary with time?
(a) The observed magnetic eld is composed of three parts: the geomagnetic main
(internal) eld, the anomalous eld (magnetic anomaly) created by the buried
dipole, and the external eld.
We determine rst the magnetic anomaly produced by the buried dipole, applying Equations (82.2) and (82.3) in Problem 82, substituting a 225 and x 0. The horizontal
component is in the NS direction (DX), because the buried dipole is in the NS-vertical
plane (Fig. 98a):
2Cm cos a
1768 nT
d3
Cm sin a
X
884 nT
d3
Y 0
Z

+
45

Fig. 98a

175

Main (internal), external, and anomalous magnetic elds

The components of the total magnetic eld at 06:00 h are given by


X1 X  X X e 19 204 nT
Y1 Y  Y Y e 0
Z1 Z  Z Z e 38 195 nT
At 12:00 h given that the external eld is twice that at 06:00 h:
X2 X  X 2X e 11 544 nT
Y2 Y  Y 2Y e 0
Z2 Z  Z 2Z e 44 623 nT
If we subtract both sets of equations and obtain
X2  X1 X e 7660 nT
Y2  Y1 Y e 0
Z2  Z1 Z e 6428 nT
Now we can calculate the elements of the main eld
X  X1  X  X e 25 980 nT
Y  Y1  Y  Y e 0
Z  Z1  Z  Z e 33 535 nT
q
H  X  2 Y  2 X  25 980 nT
tan D

Y
) D 0
H

We calculate the geomagnetic latitude of the point f and the geomagnetic constant B0
from the vertical and horizontal geomagnetic main eld components by
Z  2B0 sin f
H  B0 cos f
Z
) f 32:8
tan f
2H 
Z
B0
30 953 nT
2 sin f
We obtain the coordinates of the Geomagnetic North Pole by (Fig. 98b)
D 0 ) lB 180 l 225 E 135 W
90  fB f  f ) fB 82:8
(b) The magnitude of the external eld at 06:00 h is
q
B6e Xe2 Ye2 Ze2 10 000 nT

176

Geomagnetism

GMNP

GNP

f
f

90 fB

Fig. 98b

At 12:00 h the magnitude is


6
B12
e 2Be 20 000 nT

The direction of the external eld is in the NS-vertical plane because the EW component is
null, forming with the horizontal an angle Ie (Fig. 98c). This direction is the same at 06:00
h and at 12:00 h. We calculate the angle Ie by
Ze
Xe
Ie 40

tan Ie

Xe

Ie
Ze
Be

Fig. 98c

Main (internal), external, and anomalous magnetic elds

177

Because at
06 : 00 h t p=2 ! N 10 000
12 : 00 ht p ! N 20 000
the variation of the magnitude of the external eld with time is given by
N 10 0001  cos t
99. Buried at a depth of 100 m at a point with geographical coordinates 45 N, 45 W is
a dipole anomaly of Cm 0.1 T m3, inclined 45 from the horizontal northwards in the
vertical plane with the negative pole downwards. Measurements gave the following
results (in nT):
09:00 h X 27 759; Y 0; Z 30 141
12:00 h X 28 052; Y 0; Z 30 141
Find:
(a) The coordinates of the magnetic dipoles North Pole.
(b) The value of B0.
(c) An expression for the variation Sq knowing that it is zero at 00:00 h and
maximum at 12:00 h.
(a) As in Problem 98, the observed eld is the result of three parts: the main (internal)
eld, the buried dipole eld, and the external eld. To calculate the coordinates of
the magnetic dipoles North Pole we need to obtain the components of the
geomagnetic main eld from the components of the total eld. With this aim we
begin by calculating the magnetic anomaly created by the buried dipole, applying
Equations (82.2) and (82.3), and substituting a 225 and x 0. The horizontal
component is in the NS direction (DX) given that the dipole is on the NS-vertical
plane (Fig. 99a).
2Cm cos a
141 nT
d3
Cm sin a
X
71 nT
d3
Y 0
Z

The total observed eld at 09:00 h is


X1 X  X X1e
Y1 Y  Y Y1e 0 ) Y  Y1e
Z1 Z  Z Z1e
The total eld at 12:00 h is
X2 X  X X2e
Y2 Y  Y Y2e
Z2 Z  Z Z2e

178

Geomagnetism

X P

N
Z
d

45

Fig. 99a

Subtracting both sets of equations we obtain


X2  X1 X2e  X1e 293 nT
Y2  Y1 Y2e  Y1e 0 ) Y2e Y1e Y 
Z2  Z1

Z2e

Z1e

0)

Z2e

99:1

Z1e

We assume that the time variation of the observations is due to the diurnal Sq variation
which is zero at 00:00 h and maximum at 12:00 h. Therefore the only possible values
for the components Ye and Ze are zero because these components have the same values at
09:00 h and at 12:00 h:
Y2e Y1e Y  0
Z2e Z1e 0
Then, the intensity of the external eld is given by
X e N 1  cos ot
o

2p
24

The NS components of this eld at 09:00 h (X1) and at 12:00 h (X2) are
p


3p
2
e
N 1
X1 N 1  cos
2
4
X2e N 1  cos p 2N

Subtracting the two values and using Equation (99.1) we obtain


p

p
e


2
2 X2
e
e
X2  X1 1 
293 nT ) X2e 2001 nT
N 1
2
2
2
X1e 1708 nT

179

Main (internal), external, and anomalous magnetic elds

GMNP

GNP

f
f

90 fB

Fig. 99b

The components of the geomagnetic main eld intensity are given by


X  X1  X  X1e 26 122 nT
Y  Y1  Y  Y1e 0
Z  Z1  Z  Z e 30 000 nT
q

H X  2 Y  2 X  26 122 nT
tan D

Y
) D 0
H

The geomagnetic latitude of the point f  is determined from the vertical and horizontal
geomagnetic main eld components:
Z  2B0 sin f
H  B0 cos f
tan f

Z
) f 29:9
2H 

We obtain the coordinates of the Geomagnetic North Pole by (Fig. 99b)


D 0 ) lB 180 l 135 E
90  fB f  f ) fB 74:9

180

Geomagnetism

(b) The geomagnetic constant B0 is given by


H  B0 cos f
H
B0
30 133 nT
cos f
(c) We have obtained above that
F e X e N 1  cos ot
To calculate N we take into account that the Sq variation is maximum at 12:00 h:
X2e 2N ) N 1000 nT
100. At a point with geographical coordinates 60 N, 45 E, a magnetic
observation results in the following values: FT 48 277 nT, DT 2.9 , IT 63.7 .
It is known that at 20 m below this point there is a horizontal magnetic dipole of
moment Cm 0.01 T m3, with the positive pole oriented in the direction N 60 E. There
is also an external eld parallel to the axis of rotation directed southwards of magnitude 1000 nT.
Determine the main (internal) eld components, the constant B0, and the geocentric
geographical coordinates of the northern and southern geomagnetic poles (precision
1 nT).
We rst calculate the components of the total eld intensity from FT, DT, and IT
(Fig. 100a) by
HT FT cos IT 21 390 nT
ZT FT sin IT 43 280 nT
XT HT cos DT 21 363 nT
YT HT sin DT 1082 nT

XT
DT

YT

HT
Y

ZT

IT

FT

Fig. 100a

181

Main (internal), external, and anomalous magnetic elds

P
X60

d
a
+

Fig. 100b

The observed eld is the result of three parts: the main eld (X  , Y , Z  ), the buried dipole
eld (DX, DY, DZ), and the external eld (Xe, Ye, Ze):
XT X  X X e
YT Y  Y Y e
ZT Z  Z Z e
We determine the magnetic anomaly created by the buried dipole from Equations (82.2)
and (82.3) substituting a 90 and x 0. If we call X60 the direction N 60 E the horizontal
component of this anomaly is DX60 (Fig. 100b):
2Cm cos a
0 nT
d3
Cm sin a
X60
1250 nT
d3
The NS and EW components will be given by (Fig. 100c)
Z

X X60 cos 60 625 nT


Y X60 sin 60 1083 nT
The external eld is parallel to the Earths axis of rotation in the southwards direction so its
components are in the vertical and NS direction and are given by (Fig. 100d)
Z e 1000 sin f 866 nT
X e 1000 cos f 500 nT
We calculate the main eld components from these values:
X  XT  X  X e 21 238 nT
Y  YT  Y  Y e 1082 nT
Z  ZT  Z  Z e 42 414 nT
q
H  X  2 Y  2 21 266 nT
Y
tan D  ) D 2:9
X

182

Geomagnetism

60

Y
E

X
X60

Fig. 100c
GNP

P
Ze
Xe

Fe

Fig. 100d

The geomagnetic latitude of the point f  and the geomagnetic constant B0 are found from
the vertical and horizontal geomagnetic main eld components by
Z  2B0 sin f
H  B0 cos f
Z
tan f
) f 44:9
2H 
Z
30 953 nT 30 044 nT
B0
2 sin f

Main (internal), external, and anomalous magnetic elds

183

GNP
90 fB
l lB
90 f

180 l
GMNP
q = 90 f

Fig. 100e

To calculate the geographical coordinates of the Geomagnetic North Pole we use the
corresponding spherical triangle (Fig. 100e). We obtain the latitude fB by applying the
cosine law for the angle 90 fB:
cos90  fB cos90  f cos90  f sin90  f sin90  f cos D
sin fB sin f sin f cos f cos f cos D
fB 75:0
We obtain the longitude lB by applying the cosine rule for the angle 90 f  :
cos90  f cos90  fB cos90  f sin90  fB sin90  f cosl  lB
sin f sin fB sin f cos fB cos f cosl  lB
sin f  sin fB sin f
cos fB cos f

l  lB 180:0
cosl  lB

lB 135:0 W
Therefore the coordinates of the Geomagnetic North Pole are
fB 75 N

lB 135 W

The coordinates of the Geomagnetic South Pole (the antipodal point) are:
fA fB 75 S

lA 180 lB 45 E

101. At a point with geographical coordinates 30 N, 30 E, the observed geomagnetic
eld components are (in nT): X 15 364, Y 7660, Z 48 980. The northern
geomagnetic pole is at 60 N, 0 E, and B0 30 000 nT. There is also a constant external
magnetic eld normal to the equatorial plane, with a southwards direction, of 1000 nT
intensity. Buried 10 m below the observation point is a magnetic dipole. Calculate:
(a) The magnetic anomalies DX, DY, DZ, DH, DF.

184

Geomagnetism
(b) The orientation and magnetic moment (Cm, in nT m3) of the buried dipole.
(a) The observed values are the sum of the geomagnetic main eld, the magnetic eld
due to the buried dipole, and the external eld:
X X  X X e
Y Y  Y Y e
Z Z  Z Z e
To obtain the magnetic anomalies from these equations we rst calculate the geomagnetic
latitude and the vertical and horizontal components of the geomagnetic main eld by (71.3):
sin f sin fB sin f cos fB cos f cosl  lB
f 54
Z  2B0 sin f 48 479 nT
H  B0 cos f 17 676 nT
The declination and inclination are given by
 cos fB sinl  lB
) D 25
cos f
tan I  2 tan f ) I  70

sin D

The NS and EW components are


X  H  cos D 16 007 nT
Y  H  sin D 7498 nT
The external eld is parallel to the axis of rotation directed southwards so its components
are in the vertical and NS direction and are given by (Fig. 101a)
Z e 1000 sin f 500 nT
X e 1000 cos f 866 nT
Ye 0
From these values we calculate the magnetic anomalies DX, DY, DZ, DH, DF:
X X  X   X e 223 nT
Y Y  Y   Y e 162 nT
Z Z  Z   Z e 1 nT
H X cos D Y sin D 271 nT
F H cos I  Z sin I  94 nT
(b) We call b the angle between the geographical north and the buried dipole directions. Then using Equations (82.2) and (82.3) we obtain
Z

2Cm cos a
d3

101:1

185

Main (internal), external, and anomalous magnetic elds

GNP

P
Ze
Xe

Fe

Fig. 101a

X

Cm sin a
cos b
d3

101:2

Y

Cm sin a
sin b
d3

101:3

To solve this system of three equations in three unknowns (Cm, a, b) we divide Equation
(101.3) by (101.2):
tan b

Y
) b 36 180 144
X

This value of the angle b implies that the dipole is oriented in the N 144 E direction.
To calculate the angle a between the buried dipole and the vertical we divide Equation
(101.2) by (101.1):
X
1
 tan a cos b
Z
2
2 X
) a
90
tan a 
cos b Z
Therefore the dipole is practically horizontal (Fig. 101b) and this explains the small value
of the vertical component DZ.
Finally we calculate the magnetic moment of the buried dipole from Equation (101.2):
Cm 

d3
X 2:8  105 nT m3
sin a cos b

186

Geomagnetism

P
N 144 E

d
a
+

Fig. 101b

102. At a point with geographical coordinates 30 N, 30 E, magnetic measurements


give the following values: F 52 355 nT, I 70.5 , and D -26.0 . The terrestrial
dipole has a north pole at 60 N, 0 E and B0 30 000 nT. There is also a constant
external eld of 1000 nT with lines of force contained in the plane of the 30 E meridian
at an angle of 60 with the equatorial plane and directed southwards.
(a) Calculate the anomalies DX, DY, DZ.
(b) If the anomalies are produced by a dipole buried at 10 m below the point, what is
its orientation and its magnetic moment, Cm? Neglect values less than 10 nT.
(a) We rst calculate the components of the total eld from F, D, and I by
H F cos I 17 476 nT
Z F sin I 49 352 nT
X H cos D 15 707 nT
Y H sin D 7661 nT
The magnetic anomalies are found by subtracting from the observed values the main and
external eld contributions:
X X  X   X e
Y Y  Y   Y e
Z Z  Z   Z e
First we determine the geomagnetic latitude and declination using (71.3) and (71.2):
sin f sin fB sin f cos fB cos f cosl  lB
f 53:9
sin D

 cos fB sinl  lB
) D 25:1
cos f

The vertical and horizontal components are given by


Z  2B0 sin f 48 479 nT
H  B0 cos f 17 676 nT
X  H  cos D 16 007 nT
Y  H  sin D 7498 nT

Main (internal), external, and anomalous magnetic elds

187

GNP

60
Xe

Ze
f

Fe

Fig. 102a

60
E
f
Xe
Ze

Fe

Fig. 102b

The external eld is on the plane containing the vertical and NS directions (Figs 102a
and 102b):
Z e 1000 cos60  f 866 nT
X e 1000 sin60  f 500 nT
Ye 0

188

Geomagnetism

Finally, the magnetic anomalies are given by


X X  X   X e 200 nT
Y Y  Y   Y e 163 nT
Z Z  Z   Z e 7 nT
(b) We call b the angle between the positive pole of the buried dipole and the geographical
north. Then applying Equations (82.2) and (82.3) we obtain
2Cm cos a
d3

102:1

X

Cm sin a
cos b
d3

102:2

Y

Cm sin a
sin b
d3

102:3

Z

We divide Equation (102.3) by Equation (102.2) and obtain


tan b

Y
) b 39:2 180 140:8
X

Therefore the dipole is oriented in the N 140.8 E direction.


To calculate the inclination of the dipole from the vertical we divide Equation (102.2) by
Equation (102.1):
X
1
 tan a cos b
Z
2
2 X
tan a 
) a 89:2
cos b Z
This value implies that the dipole is nearly horizontal.
Finally we calculate Cm from the total anomalous eld DB
p
X 2 Y 2 Z 2 258 nT
Cm
B 3 ) Cm 258 000 nT m3
d

B

103. The Earths magnetic eld is formed by a centred dipole with a geomagnetic pole
at 60 N, 0 E, and B0 30 000 nT, and an external eld of 10 000 nT parallel to the
equatorial plane and to the zero meridian.
(a) Calculate the components X, Y, Z of the total eld at a point P with geographical
coordinates 60 N, 30 W.
(b) If at 30 m in the direction of the compass needle from P there is a vertical dipole of
moment Cm 4000 nT m3 buried 40 m deep, what would be the anomaly DZ
produced at P?

189

Main (internal), external, and anomalous magnetic elds

(a) The components of the total eld are the sum of the geomagnetic main eld and
the external eld:
X X Xe
Y Y Ye
Z Z Ze
Let us rst calculate the geomagnetic latitude, declination, and inclination using (71.3)
and (71.2):
sin f sin fB sin f cos fB cos f cosl  lB
f 75:1
 cos fB sinl  lB
) D 76:5
cos f
tan I  2 tan f ) I  82:4

sin D

The vertical and horizontal components are given by


Z  2B0 sin f 57 983 nT
H  B0 cos f 7714 nT
X  H  cos D 1801 nT
Y  H  sin D 7501 nT
The external eld is parallel to the equatorial plane and to the zero meridian. If its
magnitude is N 10 000 nT, its components, from Fig. 103a (plane through the point
parallel to the equator) and Fig. 103b (plane through the geographical meridian of the
point), are given by
Z e Ber N cos l cos 4330 nT
X e Bef N cos l sin 7500 nT
Y e Bel N sin l 5000 nT
Therefore the components of the total eld are
Z 62 313 nT
X 9301 nT
Y 12 501 nT
(b) The buried vertical dipole is in the direction of the compass, that is, in the direction
of the magnetic north (Fig. 103c). We calculate the anomaly DZ produced at
P using Equations (82.2) and (82.3) substituting a 0 and x 30:
Z

Cmx2  2d 2 cos a 3dx sin a

x2 d 2 5=2
Cmx2  2d 2
Z
0:029 nT
x2 d 2 5=2

190

Geomagnetism

B e = Nsinl

Be

Ncos
l

Plane parallel
to the equator

Fig. 103a

GNP
B ef
Ncos

Ber
f

Fig. 103b

191

Main (internal), external, and anomalous magnetic elds

x
NM

Fig. 103c

104. The Geomagnetic North Pole is at 60 N, 150 W, with B0 30 000 nT, and there
is an external magnetic eld of intensity 3000 nT parallel to the axis of rotation
pointing away from the North Pole. Buried 10 m below a point with coordinates
30 N, 30 E there is a horizontal dipole with Cm 40 000 nT m3 and the negative
pole pointing in the direction N 45 E.
(a) What are the components X, Y, Z of the total eld?
(b) Calculate the total eld anomaly DF.
(c) What is the angle between the direction of the compass and geographic north?
(a) The components of the total eld are the sum of the geomagnetic main eld, the
magnetic eld due to the buried dipole, and the external eld:
X X  X X e
Y Y  Y Y e
Z Z  Z Z e
We determine rst the geomagnetic latitude, declination, and inclination using (71.3)
and (71.2):
sin f sin fB sin f cos fB cos f cosl  lB
f 0
sin D

 cos fB sinl  lB
) D  0
cos f

tan I  2 tan f ) I  0

192

Geomagnetism

X45

Fig. 104

The vertical and horizontal components are given by


Z  2B0 sin f 0 nT
H  B0 cos f 30 000 nT
X  H  cos D 30 000 nT
Y  H  sin D 0 nT
We calculate the magnetic anomaly created by the buried dipole from Equations (82.2) and
(82.3), substituting a 270 and x 0. The vertical component DZ and the horizontal
component DX45 of the anomaly in the direction N 45 E are (Fig. 104)
2Cm cos a
0 nT
d3
Cm sin a
40 nT

d3

Z
X45

The NS and EW components are given by


X 40 sin 45 28:3 nT
Y 40 cos 45 28:3 nT
The external eld is parallel to the axis of rotation directed southwards, so its components
are in the vertical and NS direction and are given by
Nr Z e N cos 60 1500 nT
N# X e N sin 60 2598 nT
The components of the observed total magnetic eld are
X 27 430 nT
Y 28 nT
Z 1500 nT

193

Main (internal), external, and anomalous magnetic elds

(b) The total eld anomaly DF is given by


F X 28 nT
(c) The direction of the compass is affected by the three elds. The angle D between
the direction of the compass and the geographic north is obtained from the
horizontal components of the total observed eld
tan D

Y
28

) D 0:06
X 27430

105. The geomagnetic eld is that of a dipole in the direction of the axis of rotation
and B0 30 000 nT. There is also a constant external eld of 2500 nT normal to the
equatorial plane in the direction of the South Pole.
(a) Calculate the value of the inclination observed at a point P with coordinates 45 N,
45 E given that, at 10 m below it, there is a vertical dipole with the negative pole
upwards and moment Cm 40 000 nT m3.
(b) For a point 20 m north of P, calculate the observed inclination and declination,
and the total eld anomaly DF.
(a) The components of the total observed eld are the sum of the geomagnetic main
eld, the magnetic eld due to the buried dipole, and the external eld:
X X  X X e
Y Y  Y Y e
Z Z  Z Z e
The magnetic dipole is oriented in the direction of the axis of rotation and therefore
f f 45
D 0
Then the components of the geomagnetic main eld are
Z  2B0 sin f 42 426 nT
H  B0 cos f 21 213 nT
X  H  cos D 21 213 nT
Y  H  sin D 0
The magnetic anomaly created by the dipole is obtained from Equations (82.2) and (82.3)
substituting a 0 and x 0 (Fig. 105a):
2Cm cos a
80 nT
d3
Cm sin a
X
0
d3
Y 0
Z

194

Geomagnetism

P
X

Fig. 105a
GNP

P
Ze
Xe
f

Fe

Fig. 105b

The external eld is parallel to the axis of rotation directed southwards so its components
are in the vertical and NS direction (Fig. 105b) and are given by
Z e 2500 sin f 1768 nT
X e 2500 cos f 1768 nT
Ye 0

Main (internal), external, and anomalous magnetic elds

195

Therefore the components of the observed eld are


X X  X X e 19 445 nT
Y Y  Y Y e 0
H X
Z Z  Z Z e 44 274 nT
From these values we calculate the observed inclination
tan I

Z
) I 66:3
H

(b) For a point Q located 20 m to the north of P we can assume that the main and
external elds have the same value as at point P and only the magnetic anomaly
created by the buried dipole is different. We calculate this anomaly from Equations
(82.2) and (82.3) substituting a 0 and x 20:
Z
X

Cmx2  2d 2 cos a 3dx sin a


x2 d 2 5=2
Cm2x2  d 2 sin a  3dx cos a
x2

d 2 5=2

Cmx2  2d 2
x2 d 2 5=2
Cm3dx

x2 d 2 5=2

1 nT

4 nT

Y 0
Therefore, the components of the observed eld at that point are
X X  X X e 19 441 nT
Y Y  Y Y e 0
H X
Z Z  Z Z e 44 193 nT
From these values we calculate the observed inclination and declination by the expressions
Z
) I 66:2
H
Y
tan D ) D 0
X

tan I

The total eld anomaly DF is given by


F X cos I Z sin I 4 cos 66  1 sin 66 3 nT
106. Consider a point with coordinates 30 N, 30 E under which is buried at a depth
of 100 m a horizontal dipole of moment m0 m/4p 1 T m3, with the positive pole in
the direction N 60 E. The terrestrial eld is formed by a centred dipole in the
direction of the axis of rotation and a constant external eld of 10 000 nT from the
Sun, B0 30 000 nT.
(a) Calculate F, D, and I at that point on December 21 at 12 noon.

196

Geomagnetism

(b) How do D and I vary throughout the year?


(a) The components of the total eld intensity are the sum of the geomagnetic main
eld, the magnetic eld due to the buried dipole, and the external eld:
X X  X X e
Y Y  Y Y e
Z Z  Z Z e
We rst calculate the geomagnetic latitude and the declination and inclination. The
magnetic dipole is oriented in the direction of the axis of rotation and therefore
f f 30
D  0
Then the components of the geomagnetic main eld are
Z  2B0 sin f 30 000 nT
H  B0 cos f 25 981 nT
X  H  cos D 25 981 nT
Y  H  sin D 0
We calculate the magnetic anomaly produced by the buried dipole from Equations (82.2)
and (82.3) substituting a 90 and x 0. We call X60 the direction N 60 E and DX60 the
horizontal component of the anomaly (Fig. 106a):
2Cm cos a
0 nT
d3
Cm sin a
1000 nT

d3

Z
X60

P
X 60

d
a
+

z
Fig. 106a

197

Main (internal), external, and anomalous magnetic elds

60

E
X

X60

Fig. 106b
GNP
N = Be
P
f +e
f
e

Sun

Fig. 106c

The NS and EW components are given by (Fig. 106b)


X X60 cos 60 500 nT
Y X60 sin 60 866 nT
The external eld has a diurnal period (o 2p / 24 h) and we know that the Sun is at the
meridian point at 12:00 h (solar time). This eld changes through the year because the Sun
moves on the ecliptic plane (apparent motion) which is tilted with respect to the equatorial
plane by an angle e 23. Therefore the solar declination (d), the angle from the Sun to the
celestial equator, changes through the year. On December 21 (winter solstice) this angle is
d e 23 (Fig. 106c). If we call N the magnitude of the external eld (N 10 000nT)
its components on December 21 at 12:00 are (Fig. 106d)
Z e Br N cosf e 6018 nT
X e Bf N sinf e 7986 nT
Ye 0

198

Geomagnetism

GNP
X = Bf

P
f+e
Z = Br

Fig. 106d

The components of the observed eld are


X X  X X e 33 467 nT
Y Y  Y Y e 866 nT
Z Z  Z Z e 36 018 nT
Therefore the total eld F and the declination D are
p
F X 2 Y 2 Z 2 49 166 nT
Y
tan D ) D 1:5
X
Z
sin I ) I 47:1
F
(b) Any other day at 12:00 h
X X  X X e 25 481 nT N sinf  d
Y Y  Y Y e 866 nT
Z Z  Z Z e 30 000 nT N cosf  d
107. The internal eld of the Earth corresponds to a centred dipole with the negative
pole in the northern hemisphere at coordinates 80 N, 130 W and B0 30 000 nT.
There is a uniform external eld from the Sun of 1000 nT. Buried at 500 m depth
under a point P with geocentric coordinates 40 N, 50 E there is a positive magnetic
pole of strength CP 0.5 T m. Calculate:
(a) The total eld components X, Y, and Z, and the magnetic and geomagnetic
declination at P on March 21 at 12:00 h.

199

Main (internal), external, and anomalous magnetic elds

(b) The same parameters for a point 200 m north of P, assuming that neither the
internal nor the external elds change (precision 1 nT).
(a) The components of the intensity of the total eld are the sum of the geomagnetic
main eld, the external eld, and the magnetic eld due to the buried pole.
To calculate the main eld we determine rst the geomagnetic latitude and the declination
by (71.3) and (71.2):
sin f sin fB sin f cos fB cos f cosl  lB
f 30
sin D

 cos fB sinl  lB
) D 0
cos f

From these values we obtain the vertical and horizontal components of the geomagnetic
main eld:
Z  2B0 sin f 30 000 nT
H  B0 cos f 25 981 nT
X  H  cos D 25 981 nT
Y  H  sin D 0
To calculate the external eld we notice that it comes from the Sun which on March 21
(spring equinox) is on the equatorial plane so that the external eld is parallel to this plane.
In addition this eld changes during the day as a function of local time t with a diurnal
period (o 2p/24 h). At t 12 h, the external eld is maximum given that at this time
the Sun is at the meridian point (Fig. 107a). Calling N its magnitude (N 1000 nT), the
components of the external eld are given by
GNP
Xe
P

N
Ze
f

Fig. 107a

200

Geomagnetism

Z e Br N cos f 766 nT


X e Bf N sin f 643 nT
Y e Bel 0
The magnetic eld anomaly created by the buried pole is derived from its potential DF,
given by
CP
F
r
Applying the gradient, we obtain
B F
but the only component is the vertical:
Z Br

@F
CP
 2
@r
r

Substituting r d 500 m in this equation we obtain


Z 2000 nT
Therefore the components of the total eld and the observed declination are given by
X X  X e X 26 624 nT
Y Y  Y e Y 0
Z Z  Z e Z 28 766 nT
tan D

Y
) D 0
X

(b) The radial component of the magnetic eld anomaly created by the buried pole for
a point 200 m north of P (x 200 m) is given by
Br 

@F CP
2
@r
r

From Fig. 107b the vertical and NS components are


CP d
CPd

r2 r x2 d 2 3=2
CP x
CPx
X Br sin a 2
2
r r x d 2 3=2
Z Br cos a

Y 0
Substituting the values given (d 500 m, x 200 m, CP 0.5 Tm2), we obtain
Z 1601 nT
X 640 nT
Y 0

201

Paleomagnetism

Fig. 107b

The components of the observed total eld and the declination are the sum of the three
contributions:
X X  X e X 27 264 nT
Y Y  Y e Y 0
Z Z  Z e Z 29 165 nT
tan D

Y
) D 0
X

Paleomagnetism
108. At a point with geographical coordinates 60 N, 60 W a 1 cm3 sample was taken
of a rock with remanent magnetism, age 10 000 years, specic susceptibility
0.01 cm3. The magnetization components of the rock were:
X 40, Y 30, Z 50 nT (N, E, nadir).
The current eld is B0 30 000 nT and the geomagnetic pole coincides with the
geographical pole. Calculate:
(a) The coordinates of the virtual geomagnetic pole which corresponds to the
sample.
(b) The magnetic moment of the terrestrial dipole 10 000 years ago.
(c) The secular variation of F, D, and I in nT and minutes per year assuming that the
variation since that time has been constant.

202

Geomagnetism

GNP
90 fB
l lB

90 f

180 l

VP

90 f

Fig. 108

(a) First we determine the declination D and the geomagnetic co-latitude y, corresponding to the virtual pole, from the magnetization components of the rock X, Y,
and Z:
Y
tan D ) D 36:9
X
p
H X 2 Y 2 50 nT
Z
tan I ) I 45
H
tan I 2 cot y ) y 63:4 ; fvirtual 26:6
Since at present the geomagnetic pole coincides with the geographical pole, the geographical latitude of the point coincides with the present geomagnetic latitude:
f fpresent 60 N
To determine the coordinates of the virtual Geomagnetic North Pole (VP), corresponding
to the magnetization of the rock, we solve the spherical triangle of Fig. 108 for B and lB
using the obtained values of y and D. The latitude fB applying the cosine rule is given by
cos90  fB cos y cos90  f sin y sin90  f cos D
sin fB cos y sin f sin y cos f cos D
fB 48:2
To obtain the longitude lB we again apply the cosine rule:
cos y sin90  f cos90  fB cos90  f
sin90  fB sin90  f cosl  lB
cos y sin fB sin f cos fB cos f cosl  lB

203

Paleomagnetism

cos y  sin fB sin f


cos fB cos f

l  lB 126:4
cosl  lB

To choose between the positive and negative solution we bear in mind that the declination
is negative and so the point is to the east of the virtual magnetic North Pole:
D < 0 ) l  lB > 0
lB 173:6 E
(b) To obtain the magnetic moment we rst calculate the constant B0. The susceptibility
w relates the magnetization and the magnetic eld. If we call F the magnitude of the
paleomagnetic eld and F0 the remanent magnetization, the relation between them is
F 0 wF

108:1

w 0:01
We calculate F0 from its components
p
F 0 X 2 Y 2 Z 2 71 nT
The eld F of the virtual pole is given by
p
F B0 1 3 cos2 y

Substituting in Equation (108.1) we obtain


p
F 0 wB0 1 3 cos2 y
F0
B0 p 5610 nT
w 1 3 cos2 y

From this value we calculate the magnetic moment of the virtual pole taking a 6370 km
for the Earths radius
B0

Cm
) m 1:45  1022 A m2
a3

(c) The magnetic eld, the declination, and inclination 10 000 years ago were
F0
7100 nT
w
D 36:9
F

I 45
At present the values of these parameters are
F p Ba0
D p 0

q
1 3 sin2 fpresent 54 083 nT

tan I p 2 tan f ) I p 73:9

204

Geomagnetism

The secular variation of F, D, and I, for this period of time, is


Fp  F
4:7 nT=yr
10 000
p
D D
0:220 =yr
10 000
Ip  I
0:170 =yr
10 000
109. The following table gives the demagnetization data for a sample that was
subjected to stepwise thermal demagnetization of its natural remanent magnetization
(NRM).
Demagnetization
temperature (C)

Declination (D, E)

Inclination (I, )

NRM Intensity (J, mA/m1)

20
100
200
300
400
500
600
650
700

32
36
38
39
41
41
41
41
300

33
22
12
4
5
5
5
5
55

0.056
0.056
0.057
0.058
0.058
0.050
0.016
0.009
0.000

Calculate the direction of each stable component identied by the demagnetization


curve.
First, construct a vector component diagram of the demagnetization data. Decompose
each observation into its north (X), east (Y), and vertical (Z) components:
H J cos I
X H cos D
Y H sin D
Z J sin I
In Fig. 109 plotting X versus Y gives the projection of the demagnetization vector
onto the horizontal plane, while plotting X versus Z gives the projection onto the vertical
plane.
A stable component of NRM is represented by collinear points on the vector component
diagrams, so that two stable components can be identied in the range 20300  C and in
the range 400700  C.
The declination of a stable component is determined by measuring or by calculating the
angle between the north axis and the trajectory of the stable component in the horizontal
plane.

205

Paleomagnetism

0.06

0.06

0.05

0.05
0.04

0.03

0.03

a
0.04

0.02
0.01

0.02
d

0.01

0.00
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035

0.00
0.01

0.00

0.01
Z

0.02

0.03

Fig. 109

For the 20300  C component:


X300  X20
0:0051
23:9
a tan1
tan1
0:0115
Y300  Y20
D 180 23:9 203:9

For the 400700  C component:


D b tan

1

Y400
X400

0:0379
tan
41:0
0:0436
1

Note: this value can be obtained directly from the declination of the observations between
400 and 650  C.
The apparent inclination, Iap, of a stable component is determined by measuring or by
calculating the angle between the north axis and the trajectory of the stable component in
the vertical plane. Iap is related to the true inclination, I, by:
tan I tan Iap j cos Dj
For the 20300  C component:
Iap g tan

1

Z20  Z300
X300  X20

tan

1

0:0265
0:0051

I tan1 tan79:1j cos203:9j 78:1

79:1

For the 400700  C component:


Iap d tan1

Z400
X400

tan1

0:0051
6:7
0:0436

I tan1 tan6:7j cos41j 5:0

Note: this value can be obtained directly from the inclination of the observations between
400 and 650  C.
Therefore the stable component isolated in the range 20300  C has D 203.9 and
I 78.1 and the stable component isolated in the range 400700  C has D 41.0 and
I 5.0 .

206

Geomagnetism

110. A palaeomagnetic study of a late Jurassic limestone outcrop near Alhama de


Granada (37 N, 4 W) in southern Spain yielded a well-dened primary remanent
magnetization whose directions are given in the table below. Calculate the mean
direction of the primary remanence of the seven samples. Compare this direction
with that dened by the reference late Jurassic palaeomagnetic pole for the stable
Iberian tectonic plate (252 E, 58 N). How much vertical axis rotation has the studied
outcrop suffered with respect to stable Iberia?

Declination (D, E)

Inclination (I, )

30
28
34
25
32
35
26

43
39
44
45
38
44
40

Use unit vector addition to calculate the mean direction of the primary remanence.
Calculate the direction cosines of each direction, the resultant total eld vector, F, and
then the mean direction using:
X cos I cos D
Y cos I sin D
Z sin I
v
!2
!2
u N !2
N
N
u X
X
X
Ft
Z 6:98542
Y
X
Xmean

N
P

i1

i1

i1

Xi

i1

F
N
P
Yi

0:6446;

Ymean i1 0:37186;
F
N
P
Zi
Zmean i1 0:66799
F

Ymean
Dmean tan1
30:0
Xmean
Imean sin1 Zmean 41:9

The mean direction of the primary remanence has a declination of 30.0 and an inclination
of 41.9 .

207

Paleomagnetism

Next, calculate the expected eld direction at the site using the reference palaeomagnetic
pole. The rst step is to determine y 90  f  , (Equation 71.3), from the pole (fp, lp) to
the site (fs, ls) using spherical triangles:
sin f sin p sin s cos p cos s cosls  lp ) f 24:1
The expected inclination can then be calculated using:
tan Iexp 2 tan f ) Iexp 41:8
The expected declination can be calculated by (71.2):
sin Dexp

 cos p sinls  lp
) Dexp 34:5 W 325:5
cos f

rotation about the vertical axis should give rise to a difference between the observed and
expected declinations, dened as positive for an observed declination clockwise from the
expected declination.
Therefore the outcrop has suffered 64.5 of clockwise rotation with respect to stable
Iberia.

Seismology

Elasticity
111. Determine the principal stresses and principal axes of the stress tensor:
0
1
2 1 1
@ 1 0 1 A
1
1 2

Find the invariants I1, I2, I3, the deviator tensor, its eigenvalues, and the invariants
J2 and J3.
To calculate the principal stresses (s1, s2, s3) and principal axes (n1i, n2i, n3i), we
calculate the eigenvalues and eigenvectors of the matrix. They are found through the
equation
tij  sdij ni 0

111:1

The eigenvalues are the roots of the cubic equation for s resulting from putting the
determinant of the matrix in (111.1) equal to zero:


2  s 1
1



1 0
1 s


1
1 2  s
) 2  ss2  s  1  1 s  2  s  2  s 0

s3  4s2 s 6 0

The three roots of the equation are the principal stresses


s1 1
s2 2
s3 3
From these values we obtain s0 13 s1 s2 s3 43.
The invariants of the matrix are the coefcients of the characteristic equation
s3  I1 s2 I2 s  I3 0

208

209

Elasticity

which, in terms of the roots of the equation, are


I1 4 s1 s2 s3
I2 1 s1 s2 s1 s3 s2 s3
I3 6 s1 s2 s3
The principal axes of stress are the eigenvectors ni associated with the three eigenvalues.
For s1 1
0
10 1
3 1 1
n1


@ 1 1 1 A@ n2 A 0 ) n1 ; n1 ; n1 1; 2; 1
1 2 3
1
1 3
n3
For s2 2

For s3 3

0 1
@ 1 2
1
1

10 1
1
n1


1 A@ n2 A 0 ) n21 ; n22 ; n23 1; 1; 1
0
n3

10 1
1
n1


1 A@ n2 A 0 ) n31 ; n32 ; n33 1; 0; 1
1
n3

1 1
@ 1 3
1
1

The deviatoric stress tensor is dened as

t0ij tij  s0 dij


where in our problem, s0 4/3.
The three components of the principal diagonal of the deviatoric tensor are
t011

2
3

t022 
t033

4
3

2
3

To calculate its eigenvalues we proceed as we did before:


0
1
2
1
1 C
B3 s
C
B
13
38
4
C
B
0
B 1   s
1 C 0 ) s3  s
B
C
3
27
3
A
@
2
s
1
1
3

Comparing with the characteristic equation

s3  J1 s2 J2 s  J3 0

210

Seismology

the invariants are


13
3
38
J3 
27

J2 

Solving the cubic equation we obtain s1 2.22, s2 0.33, s3 1.89.


112. Given the stress tensor
0

1
4

B
B
B
B
B
B 25
B
4
B
B
3
@ 1 2
2

5
2
4
1
4

3
1 2
2
2


3 1
1 2
C
C
2
C

3 C
2
1 C
C
2
2 C
C
C
A
3
2

calculate:
(a) The principal stresses.
(b) The angles formed by the greatest of these stresses with the axes 1, 2, 3.
(a) As in the previous problem to nd the principal stresses we calculate the eigenvalues of the stress matrix
1
0

3=2
1
5
1
B
C

C
B 4s
4
2
B
C

3=2 C
B
C
B
5
1
1
B 
C 0 ) s3  2s2  s 2 0
s
C
B
4
4
2
B
C

3=2
C
B
3=2
A
@ 1
1
3
s

2
2
2

Solving the cubic equation, its three roots are

s1 2
s2 1
s3 1
The largest is s1. The associated eigenvector corresponds to the axis of greatest stress
whose direction cosines are (n1, n2, n3). They are found by solving the equation
0

3=2 1
7
5
1
B
C


B
C0 1
4
4
C n1
B
2

B
C
5
7
1
C@ n2 A 0
B


3=2

B
C
4
4
2
C n3
B
C
B
3=2

3=2
A
@
1
1
1



2
2
2

211

Wave propagation. Potentials and displacements

Solving this equation with the condition that n21 n22 n23 1, we obtain,
n1 n2
1
n3 p
2

1
2

(b) From these values we obtain #, the angle with the vertical axis (x3) and , the angle
which forms its projection on the horizontal plane with x1:
9
1 >
>
n1 sin # cos
>
2 >
>
1=
n2 sin # sin 
) 315 ; # 45
2>
>
>
1
>
>
n3 cos # p
;
2
113. The stress tensor t ij in a continuous medium is
0
1
3x1 x2 5x22 0
@ 5x2
0 2x3 A
2
0
2x3 0
Determine the stress vector Tin acting at the point (2, 1, 3) through the plane
tangential to the cylindrical surface x22 1 x23 at that point.
First we calculate the value of the stress tensor at the given point:
0
1
6
5
0
p
p
0 2 3 A
tij 2; 1; 3 @ 5
p
0 2 3
0

A unit vector normal to the surface f x22 x23  4 0 at the given point is
0
1
@f
@f
@f
p


B @x1 @x2 @x3 C
grad f
C 0; 1 ; 3






B
ni
;
;
2 2
jgrad f j @ @f @f @f A
@x @x @x
1

Then, the stress vector acting at the point through that surface is given by
0
1
0
1 0

1 C
6
5
0 B
p
B
C
p
AB 2 C 5 ; 3; 3
0
2
3
Tin tij nj @ 5
B p C
p
2
@ 3A
0 2 3
0
2

Wave propagation. Potentials and displacements


114. The amplitudes of P- and S-waves of frequency 4/2p Hz are

4
4
4
p ; p ; p
uP
3 2 3 2 3
 p p p
S
u  3;  3; 2

212

Seismology
and their speeds of propagation are 6 km s1 and 4 km s1, respectively. Find the
scalar and vector potentials. Displacements are always given in m.
The displacements of P-waves can be deduced from a scalar potential function such that
uPi ri . The general form of the potential function for P-waves for harmonic motion is
A exp ika ni xi  at

114:1

where A is the amplitude, ni the direction cosines of the ray or propagation direction, a the
velocity of propagation, and ka the wavenumber. If uj is given in m and ka in km1, then A
is given in 103 m2. Taking the derivatives in (114.1) we obtain for the components of the
displacement
uPj

@
Aika nj exp ika nk xk  at
@xj

Their amplitude is
uPj Aka nj

114:2

and the wavenumber is


4
o 2p 2p 2
km1
ka
6
a
3
By substitution in (114.2), we obtain for the two horizontal components
4
uP1 Aka n1 p mm
3 2
4
uP2 Aka n2 p mm
3 2

Dividing these two expressions and writing the direction cosines of the ray in terms of the
incident angle i and azimuth az,
n1 sin i cos az
v2 sin i sin az
n3 cos i

114:3

we have
uP1
n1 sin i cos az
1
) a 45
P
n2 sin i sin az
u2
Using the uP3 and uP1 components,
p
uP3 Aka n3
cos i

3 ) i 30
uP1 Aka n1 sin i cos az

From the values of the direction cosines and the amplitude of the displacement we calculate
the amplitude of the potential A:
A

uP1
) A 4  103 m2
ka n1

213

Wave propagation. Potentials and displacements

Finally, the expression for the scalar potential of P-waves is


p
p
p

2
2
2
3
4 exp i
x1
x2
x3  6t
3
4
4
2
Displacements of the S-wave are obtained from a vector potential ci of null divergence,
whose general form is


ci Bi exp ikb nj xj  bt
where b is the velocity of propagation and kb the wavenumber. The displacements are
given by
uS r  c

114:4

The wavenumber is kb o/b 1 km1. According to (114.4) the relation between the
components of the displacement and of the amplitude of the potential is
p
p
p
2
3
S
 B2
 3 mm
u1 B3
4
2
p
p
p
3
2
uS2 B1
 B3
 3 mm
2
4
p
p
2
2 p
uS3 B2
 B1
2 mm
4
4
The potential must have null divergence,
p
p
p
3
2
2
r c
B3 0
B1
B2
2
4
4
From these equations we obtain, in units of 103 m2,
B1 2
B2 2
B3 0
The S-wave vector potential is
p
p
p

2
2
3
cj 2; 2; 0 exp i
x1
x2
x3  4t
4
4
2
Note: These units will be used for all problems but not explicitly given.
115. The components of the S-wave with respect to the axes (x1, x2, x3) are (6, 3.25, 3)
where x2 is the vertical axis, the azimuth is 60 , and the angle of incidence is 30 .
Determine the amplitude and direction cosines of the SV and SH components.
From the azimuth and incident angles we calculate the direction cosines (x2 is the
vertical axis)

214

Seismology

X2

SV

r
30
SVH

X1
60
SH

90
X3

Fig. 115a

1
n1 sin i cos az
4
p
3
n2 cos i
2
p
3
n3 sin i sin az
4
Since the SV and SH components are on a plane normal to the direction of the ray r
(Fig. 115a) unit vectors in the direction of SV (a1, a2, a3) and of SH (b1, 0, b3) must satisfy
the equations
p
p
a1 a2 3 a3 3

4
2
4
p
b1 b3 3
0
SH r 0 )
4
4

SV r 0 )

115:1

SH SV 0 ) a1 b1 a3 b3 0
The projections on the horizontal plane R of the ray r and SH are perpendicular
(Fig. 115b). Then SH forms an angle of 180 30 with the x1 axis. The direction
cosines of SH are
p
3
b1 
2
1
b3
2

215

Wave propagation. Potentials and displacements

SVH

30
X1

30
60

SH
X3

Fig. 115b

SV forms an angle of 60 with the vertical axis x2 (Fig. 115a). Then a2  sin i  12
From a2 using Equations (115.1) and a21 a22 a23 1, we calculate a1 and a2:
p
3
a1
4
3
a3
4
p
p p


7 5 7
7
116. Given the potential ci p ; p ; 6 exp 4i p1 x1 p1 x2 p x3  4t ,
5
3
5
5
15
calculate the polarization angle.
First we calculate the amplitudes of the components of the displacement of the S-wave
from the vector potential
8
13
>
>
uS1 c3;2  c2;3 4 p 30:02
>
>
>
3
>
>

>
<
7
6
S
S
p

c

c

4
13:97

ui r  c i )
1;3
3;1
2
>
5
5 3
>
>
p

>

>
>
p
7
>
S
>
: u3 c2;1  c1;2 4
7  p 7:85
15
The modulus is

uS

q
u21 u22 u23 34:03

From the vertical component uS3 we calculate the SV component (Fig. 116) knowing that
r p
p
8
7
7
n3 cos i p ) sin i 1 
p
15
15
15

Then, uS3 uSV cos90  i

) uSV 10:75.

216

Seismology

X3
r
i

SV

90 i
R

Fig. 116

To nd the SH component we use the relation


q
uS uSV 2 uSH 2 ) uSH 32:29

From the SV and SH components we obtain the polarization angle e:


tan e

uSH
) e 71:6
uSV

117. Given the amplitudes of the P- and S-waves (ka 1):


uP1 4

uS1 8

p
uS2 2 2

p
uS3  4 2

uP2 4
uP3 8

determine the angle of incidence i, azimuth az, polarization angle of the S-wave,
and apparent polarization angle g.
Given that the displacements of the P-wave are on the incident plane, in the direction of the
ray r, the angle of incidence i can be obtained from the modulus and the vertical component:
p
p
uP 42 42 82 4 6
cos i

uP3
8
p ) i 35:3
P
u
4 6

The azimuth, the angle between the horizontal projection of the ray and the north (x1), is
obtained from the horizontal components, uP1 and uP2 :
tan az

uP2
) az 45
uP1

217

Wave propagation. Potentials and displacements

90 az

uS
1

az
uS
H

u S2
az

X1

SH
R

X2

Fig. 117

We calculate the SV component from the vertical component uS3 , as in the previous
problem:
uSV

uS3
uS
3 9:37
cos90  i0 sin i0

The SH component is found from the horizontal components (Fig. 117)


uSH uS2 cos az  uS1 sin az 3:66
From SV and SH we nd the polarization angle e (Fig. 117):
tan e

uSH 3:66

uSV 9:37

) e 21:3

To calculate the apparent polarization angle g, the angle between the horizontal component
of S, (uSH), and the radial direction R (Fig. 117), we use the relation
tan e cos i0 tan g

) g 25:5


118. Given the values az 60 , 30 , g 45 , and ub 5, calculate the
amplitudes of the components 1, 2, and 3 of the S-wave.
From the modulus of the displacement of S-waves and the polarization angle, we calculate
the SV and SH components (Fig. 118a):
uSV
uSH

p
5 3
u cos e
2
5
S
u sin e
2
S

218

Seismology

SV

SH

Fig. 118a

X3
r

SV
uS
3
i
i
R
SV
uH

Fig. 118b

From the angles e and g we obtain the incidence angle i:


tan e tan g cos i

1
) cos i p
3

r p
2
1
) sin i 1  p
3
3

The vertical component u3 of the S-wave is obtained from the value of SV (Fig. 118b):
p p
5 3 2
S
SV
p 3:53
u3 u cos90  i
2
3

To calculate the horizontal components we have to take into account the horizontal
component of SV (Fig. 118b):
SV
uSV
cos i
H u

5
2

219

Wave propagation. Potentials and displacements

u SV
H

az
X1
uS
H

az
u SH

az

SH
R

X2

Fig. 118c

From SH and the horizontal component of SV we nd the horizontal components


of S (Fig. 118c):
p


5
uS1  uSH sin az uSV
1 3 3:42
H cos az 
4

p
5
1  3 0:92
uS2 uSH cos az  uSV
H sin az
4
119. For a scalar potential w and a vector potential c i, it is known that A 3, Bi
(2, 2, 0), ka 2/3, T p/2, and Poissons ratio is s 1/4. Calculate the amplitudes
u1, u2, and u3 of the P- and SV-waves.
on the free surface (x3 0) of the components
p


3
1
1
The direction cosines are p ; p ;
.
2 2 2 2 2

The general expressions for the scalar and vector potentials are

A exp ika n1 x1 n2 x2 n3 x3  at
ci Bi exp ikb n1 x1 n2 x2 n3 x3  bt
Since ka o/a then a o/ka 6 km s1.
If Poissons ratio is 0.25 then
1
l
s
4 2 l m

)lm

Substituting this condition in the equation for the P-wave velocity a, we nd for the
velocity b of S-waves is given by
s s
p
p
l 2m
3m
a

b 3 ) b p 2 3 km s1
a
r
r
3
Then the wavenumber of S-waves is

kb

o
2
p
b
3

220

Seismology

Then we can write the complete expressions for the potentials:


p

2
1
1
3
p

3 exp i
x1
x2
x3  6t
3 2 2
2
2 2
p

p
3
2
1
1
x 3  2 3t
ci 2; 2; 0 exp i p p x1 p x2
2
3 2 2
2 2

119:1

To determine the displacements of the P- and S-waves we use the relations


uP r
uS r  c
obtaining
1
uP1 p
2
1
uP2 p
2
p

uP3 3

uS1 2
uS2 2
p
2 2
uS3 p
3

From the components of the displacement of the S-wave we can obtain the SV component.
The values of the angles of incidence and azimuth are found from the direction cosines,
1
n1 sin i cos az p
2 2
1
n2 sin i sin az p
2 2
p
3
n3 cos i
2

) i 30 ;

az 45

and from the angle i we obtain the SV component from uS3:


p
4 2
S
SV
SV
u3 u cos90  i ) u p 3:27
3

From the horizontal components of the S-waves and the azimuth we calculate the SH
component:
uSH uS2 cos az  uS1 sin az 0
120. In an elastic medium of density r 3 g cm3 and Poisson ratio 1/3 there

propagate P- and S-waves of frequency 1 Hz in the direction

7
1 1
; p ; p . Given
3 2 3 2

that the pressure of the P-wave is 5000 dyn cm2, the magnitude of its displacement,
10 mm, is twice that of the S-wave, and the angle g 45 , nd all the parameters
involved in the expression of the potentials and ci. (It is not necessary to solve the
equations to obtain the coefcients Bi)

221

Wave propagation. Potentials and displacements

Given that Poissons ratio is 1/3 the relation between the elastic coefcients l and m is,
s

l
1
) l 2m
2l m 3

and between the velocities of P (a) and S (b)-waves is


s s
l 2m
4m
) a 2b

a
r
r
The bulk modulus K is
2
8
K l m m
3
3
Expressing m in terms of b and a, we get for K:
r
m
8
8 a2 2
b
) m rb2 ) K rb2 r ra2
r
3
3 4
3
Taking into account that the bulk modulus K is dened as the applied pressure divided by
the change in volume per unit volume y,
2
P Ky ra2 y
3
2
y r Aka2
y

120:1

3P
3P
Aka2 ) a2
2ra2
2rAka2

The expression for the scalar potential is


A exp ika n1 x1 n2 x2 n3 x3  at
where the wavenumber for P-waves is
ka

o 2pf

a
a

P
P
P
u jrj ka A ) A ju j ju ja
ka
2pf

By substitution in (120.1) we obtain the values of a and b:


a2
where
P 5000 dyn cm2
r 3 g cm3
f 1 Hz
uP 10 mm

3P
3Pa
3P

)a
2rAka2 2rjuP j2pf
2rjuP j2pf

120:2

222

Seismology
We obtain a 3.98 km s1 and b 1.99 km s1.
Since we know A, a, and ka we can write the complete expression for the scalar
potential
p

7
1
1
10 exp i1:58 x1 p x2 p x3  3:98t
3
3 2
2

The vector potential of the S-waves is given by

ci Bi exp ikb n1 x1 n2 x2 n3 x3  bt

120:3

We calculate kb:
kb

o
3:16 km1
b

The displacements are given by


uS r  c
and
uS1

@c3 @c2

kb B3 n2  B2 n3
@x2 @x3

uS2

@c1 @c3

kb B1 n3  B3 n1
@x3 @x1

uS3

@c2 @c1

kb B2 n1  B1 n2
@x1 @x2

120:4

The incidence angle i is found from n3 and, using tane cosi tang, we nd the polarization
angle e:
p
7
n3 cos i p ) i 31:95 ) e 31:95
3 2
The azimuth is

az tan1

n2
64:76
n1

Since the amplitude of the S-wave displacement is 5 mm, knowing the value of e we can
nd the values of the SV and SH components:
uSV uS cos e 4:24 mm
uSH uS sin e 2:65 mm
From uSV we calculate its vertical and horizontal components:
uS3 uSV cos90  i 2:25 mm
SV
uSV
cos i 3:61 mm
H u

223

Wave propagation. Potentials and displacements


Using uHSV and uSH we nd the two horizontal components:


uS1  uSH sin az uSV
H cos az 3:94 mm
uS2 uSH cos az  uSV
H sin az 2:14 mm

Using the values found for the displacements and Equations (120.4) and c 0 (n1 B1
n2 B2 n3 B3 0) we nd the values of B1, B2, B3. Substituting all the values in (120.3) we
obtain for the vector potential
p

7
1
1
p

ci 1:17; 2:5; 3:44 exp 3:16i x1


x2
x3  1:99t
3
2
3 2

121. At the origin in an innite medium in which s, Poissons ratio, is 0.25, and the
density is 3 g cm3, there is an emitter of elastic plane waves of frequency 0.5 cps.
Calculate:
(a) The equation of the P- and S-waves in exponential form and with arbitrary
amplitudes for the wave arriving at the point A(500, 300, 141) km.
(b) The arrival time.
(a) First we calculate the distance to point A and the direction cosines of the direction
of the ray (r) (Fig. 121):
p
5002 3002 1412 599:90
600 km
500 5

n1
600 6
300 1

n2
600 2p
2
141
n3

6
600
r

X3

r
141
X1
300
500

X2

Fig. 121

224

Seismology

The S-wave velocity is


b
Given that Poissons ratio is 0.25,

r p
m
10 km s1
r

p p
s 0:25 ) l m ) a b 3 30 km s1

The wavenumbers of the P- and S-waves are

2pf
2p  0:5
p

p km1
a
5:5
30
2pf
2p  0:5
p
p p km1
kb
b
10
10
ka

The general expressions for the scalar and vector potentials are
A exp ika n1 x1 n2 x2 n3 x3  at
ci Bi exp ikb n1 x1 n2 x2 n3 x3  bt
Leaving the amplitudes A and Bi in arbitrary form and substituting the obtained values we
have for the potentials,
p

p
p
5
1
2
x1 x2
x3  30t
A exp i p
2
6
30 6
p

p
p
5
1
2
x1 x2
x3  10t
ci Bi exp i p
2
6
10 6
(b) The travel times for P- and S-waves from the origin to the given point are
r
600
p 109:5 s
a
30
r
600
t b p 189:7 s
b
10

ta

Reection and refraction


122. A P-wave represented by the potential

x1
x2
x3
p

 4t
w 4 exp 0:25i
6
3
2

is incident on the surface x3 0 separating two liquids of densities 3 g cm3 and


4 g cm3. If the speed of propagation in the second medium is 2 km s1, write the
expressions for the potentials of the reected and transmitted waves.

225

Reection and refraction

The potentials of the reected and transmitted waves are given by


refl A exp ik  tan e x3 x1  ct
trans A0 exp ik tan e0 x3 x1  ct

122:1

where e 90 i is the emergence angle and i the incidence angle, and k ka cos e is the
wavenumber corresponding to the apparent horizontal velocity, c a /cos e. These
expressions are written for rays contained on the incidence plane (x1, x3). Then, we have
to rotate the given potential to refer it to the incidence plane. First, from the direction
cosines we calculate the incidence angle i and the azimuth az:
1
n1 sin i cos az p
6
1
n2 sin i sin az p
3
1
n3 cos i p ) i 45 e
2
p
1
2
cos az p ) sin az p
3
3
Using the rotation matrix we obtain the direction cosines on the plane of incidence (x1, x3):
1
0
1
10 1
0 01 0
p

n1
n1
cos az
sin az 0
B 2C
C
@ n02 A @  sin az cos az 0 A@ n2 A ) B
B 0 C
@
1 A
n03
1
n3
0
0
p
2
The values of c and k are
p
a
a0
8
p

4
2 km s1

cos e cos e0
2 p
2
km1
k ka cos e ka0 cos e0
8
Then the potential of the incident wave is now given by
p 
1 
inc 4 exp i p x3 x1  4 2 t
4 2
0
The angle i of the transmitted or refracted ray is found from Snells law:
p
sin i sin i0
2
0
0 ) sin i
cos e0
a
4
a
p
p
14
0
0
) tan e0 7
) cos i sin e
4
Using the expressions for the reection and refraction coefcients, V and W, we can
calculate the amplitude of the reected and refracted potentials:
p
p
A
r0 tan e  r tan e0 4  3 7
16  12 7
p

p
1:07
V
0

)
A

A0 r tan e r tan e0 4 3 7
43 7
A0
2rtge0
6
24
p ) A0
p 2:01
0

W
A0 r tan e r tan e0 4 3 7
43 7

226

Seismology

By substitution in (122.1) we obtain


p

8
2
refl 1:07 exp
i x3 x1  p t
8
2
p

2
8
i
7x3 x1  p t
trans 2:01 exp
8
2
 p p p
123. A P-wave of amplitude 5 2; 5 6; 10 2 and frequency v 12 rad s1 in a
semi-innite medium of speed of propagation a 6 km s1 and Poissons ratio 0.25 is
incident on the free surface. Calculate:
(a) The potential of the incident P-wave.
(b) The potential of the reected S-wave.
(c) The components u1, u2, u3 of the reected S-wave.
(a) The displacements of the P-wave can be deduced from its scalar potential:
A exp ika n1 x1 n2 x2 n3 x3  at
123:1
uP r
where A is the amplitude, ka is the wavenumber (P), ni are the direction cosines, and a is the
P-wave velocity. The wavenumber is found from the given angular frequency and velocity:
ka

o 12

2 km1
a
6

Since we know the amplitudes of the components of the displacements we can nd the
incidence angle i and the azimuth az:
p
@
Aka n1 A2 sin i cos az 5 2
@x1
p
@
Aka n2 A2 sin i sin az 5 6
uP2
@x2
p
@
uP3
Aka n3 A2 cos i 10 2
@x3
p
Dividing the two rst equations we obtain 3 tan az ) az 60 , and dividing the
last two,
p
5 2
1
p tan i cos az tan i ) i 45
2
10 2
uP1

The amplitude A is given by

P q
2  2  2
j uP j
u
102 m2
uP1 uP2 uP3 Aka ) A
ka

Finally the potential is given by

inc 102 exp i2

p
p
p

2
6
2
x1
x3  6t m2
x2
4
4
2

227

Reection and refraction

If we express the potential referred to the plane (x1, x3) as the incidence plane, as we did in
Problem 122, then
A exp ik x1 tan e x3  ct
where
p
1
k ka cos e 2 p 2
2
p
a
6
c

6 2 km s1
1
cos e
p
2

e 90  i 45 ;

and the potential is

p
p 
inc 102 exp i 2 x1 x3  6 2t m2

123:2

(b) Since Poissons ratio is 1/4, l m, and

p
a
b p 2 3 km s1
3

The angle f of the reected S-wave is obtained by Snells law:


r
p
cos e cos f
b
1
5
2

) cos f cos e p p ) sin f


a
b
a
6
6
2 3

From the values of e and f we calculate the P-to-S reection coefcient VPS, using equation
VPS

4a1 3a2
4ab 1 3a2 2

where we substitute
a tan e 1

and

b tan f

so

p
5

41 3
VPS p
0:64
4 5 1 3 2

From this coefcient we calculate the proportion of the incident P-wave which is reected
as an S-wave (only with SV component; the negative sign indicates the opposite sense of
the reected ray):
B AVPS 10  0:64 6:4  103 m2
When the ray is contained in the (x1, x3) plane we use a scalar potential for the S-wave
which in this case is given by
c B exp ik x1  tan f x3  ct
p
p 
p
6:4  103 exp i 2 x1  5x3  6 2t m2

228

Seismology

(c) To calculate the amplitudes of the total displacements in terms of the two scalar
potentials, we remember that for this orientation of the axes the displacements are
given by
@ @c
u1

uP1 uSV
1
@x1 @x3
@ @c
u3

uP3 uSV
3
@x3 @x1
The displacements of the SV reected wave in this case are
p
@c
6:4 5
uSV
1 
@x3
p
@c
6:4 2
uSV
3
@x1

If we want to determine the components 1 and 2, referred to the original system of axes, we
SV

project uSV
1 uR using the azimuth 60 :
p
SV
uSV
1 uR cos az 3:2 5
p
SV
uSV
2 u2 sin az 3:2 15

124. An S-wave of vector potential


p
p



p
x1
3
3

x2
x3  4t
ci 10 3; 2; 4 exp 5i
4
4
2

is incident on the free surface x3 0. Find the SV and SH components of the reected
S-wave referred to the same coordinate system as the incident wave, and the coefcient of reection. Poissons ratio is 3/8.
According to the value of Poissons ratio the relation between l and m is
s

l
3
) l 3m
2 l m 8

and the relation between the velocities of the P-waves and S-waves is
ss
l 2m
5m p

5b
a
r
r

The incidence angle i and the azimuth az are obtained from the direction cosines:
p
3
n3 cos i
) i 30
2
1
n1 sin i cos az sin 30 cos az ) az 60
4
Using Snells law we nd the value of the critical angle
sin ic 1
b
1
) sin ic p ) ic 26:5
b
a
a
5

229

Reection and refraction

X3

r
i
u SV
60

Fig. 124a

Since i > ic, there is no reected P-wave. The components of the incident S-wave are
obtained from the potential
8
< uS1 0
S
ui r  c ) uS2 80
: S
u3 40

q
p
The modulus of the displacement is uS 802 402 40 5:
The SV component is given by (Fig. 124a)
uSV

u3
80
cos90  i0

and the SH component is


uSH

q
uS 2 uSV 2 40

The minus sign corresponds to that of uS2


The amplitude of the reected SH wave is equal to that of the incident SH wave. We nd
the 1 and 2 components using the azimuth 60 (Fig. 124b):
p
p
p

3 1
3
3
1
uSH

40

;
;
0
exp
5i
x
x
x


4t

2
3
1
i
2 2
4
2
4
For total reection the amplitude of the reected SV is equal to that of the incident one, but
with a phase shift d. The components are given by (Figs 124a and 124b)
SV
cos i cos az
uSV
1 u
SV
uSV
cos i sin az
2 u
SV
uSV
sin i
3 u

230

Seismology

X1

u SH

30

X2

60
SV

uH

Fig. 124b

The components of the reected SV wave are


p
p



p

3
3
3
3 1
1
SV
; ;
x2 
x3  4t id
exp 5i x1
ui 80 
4
4
2
4 2
4
To determine the phase shift d we have to determine the reection coefcients for a free
surface. Textbooks usually give those for Poissons ratio s = 1/4, but since in this problem
s 3/8 we have to calculate them. On a free surface the boundary conditions are that
stresses are null, which in terms of the scalar potentials and c, are given by


t31 0 m u3;1 u1;3 2;31  c;11 c;33


124:1
t33 0 l u1;1 u3;3 2mu3;3 3;11  5;33  2c;13
where, using (x1, x3) as the plane of incidence, the scalar potential of the reected
P-waves is
A exp ik ax3 x1  ct
and the scalar potential of the incident and reected S-wave is
c B0 exp ik bx3 x1  ct B exp ik bx3 x1  ct
Substituting in (124.1) we obtain for the coefcient of the reected S-waves,
VSS

B
i4^ab  1  b2 3 5^a2

B0 i4ab 1  b2 3 5a2

The phase shift is given by


d tan

1

4^ab
1  b2 3  5^a2

231

Reection and refraction

By substitution of the values of the problem,


s
c2
1
b tan f
 1 p
2
b
3
r r
c2
22
^
a 1 2
a
30
and nally we obtain d 77.33 .
125. An S-wave represented by the potential
p

p
3
1
x3  4 2 t
c 10 exp i3 x1
2
2

is incident from an elastic medium with l 0 onto a liquid with velocity a0 4 km s1
(the two media have the same density). Derive the equations relating the amplitudes of
the potentials of the incident, reected, and transmitted waves.
Given that the wave is incident from an elastic medium onto a liquid medium, there are reected
S- and P-waves in the elastic medium and transmitted P-waves in the liquid (Fig. 125).
If l 0, the P-wave velocity in the elastic medium is
s s
p
pp
l 2m
2m

b 2 4 2 2 8 km s1
a
r
r
Assuming (x1, x3) is the incidence plane, we use the scalar S-wave potential which, for the
incident and reected waves in the solid medium, is given by
c B0 exp ikb x1 cos f x3 sin f  bt
B exp ikb x1 cos f  x3 sin f  bt
X3
P

e
a= 4
b= 0

M
f

X1
a=8
b = 42

Fig. 125

232

Seismology

From the potential of the incident wave given in the problem,


cos f

1
) f 60 ;
2

B0 10;

kb 3

The potential can also be written in the form (Problem 122)


c B0 exp ik x1 tan f x3  ct B exp ik x1  tan f x3  ct
where
3
km1
2
p
b
8 2 km s1
c
cos f

k kb cos f

and the potential is given by


p 
p 
p
p
3
3
c 10 exp i x1 x3 3  8 2t B exp i x1  x3 3  8 2t
2
2

Applying Snells law we determine the angle e of the reected P-wave in the solid medium
and the angle e0 of the transmitted P-wave onto the liquid (Fig. 125):
cos f
cos e
cos 60 cos e

) p
) e 45
b
a
8
4 2
cos e0 cos f
1
) cos e0 p ) e0 69

0
a
b
2 2
r
p
p
7
1
sin e0 1  p ) tan e0 7
8 2 2

The potential of the reected P-wave in the solid medium (M) is


p 
3
A exp ik x1  x3 tan e  ct A exp i x1  x3  8 2t
2

and that of the transmitted P-wave in the liquid (M0 ) is


0 A0 exp ik x1 x3 tan e0  ct A0 exp i

p 
p
3
x 1 x 3 7  8 2t
2

The relation between the amplitude of the potential of the incident S-wave (B0 = 10) and
those of the reected and refracted P-waves A, A0 and the reected S-wave B can be
obtained from the conditions at the boundary between the two media (x3 0), that is,
continuity of the normal component of the displacements (u3) and of the stress (t33) and
null tangential stress (t31):
u3 u03 ) ;3 c;1 0;3
t31 0 ) 2;13  c;33 c;11 0




t33 t033 ) l0 0;33 0;11 2m ;33 c;13

233

Reection and refraction


By substitution of the potentials we obtain the equations (in units of 103 m2)
p
 A 10 B A0 7
A B 10 0
p p
2A0 A 10 3  3B

Solving the system of equations we obtain


p
40 3
0
p p 6:2
A
4 7 21
p
20 21
p 8:2
p

A
4 7 21
p p

10 4  7 21
p p
B
1:8
4 7 21

126. An S-wave incident on the free surface of a semi-innite medium with s 0.25 is
given by (in units of 103 m2)
p
p



  1
p
3
3
ci 10 3; 2; 4 exp 5i x1
x2
x3  4t
4
4
2
Calculate:

(a) The amplitude of the components of the reected P-wave.


(b) The components of the reected S-wave.
(a) From the direction cosines we calculate the incidence i and emergence f angles and
azimuth az of the incident S-wave:
1
sin i cos az
4
p
3
sin i sin az
n2
4
p
3
cos i ) i 30 ) f 60
n3
2
1 1
cos az ) az 60
4 2
p
p
Since Poissons ratio is 0.25 then l m ) a 3b 4 3, and
n1

sin ic 1
1
) sin ic p ) ic 35
b
a
3

Since i < ic we have a reected P-wave. The reection coefcient at a free surface for a
reected P-wave from an incident S-wave is given by
VSP

4b1 3a2
4ab 1 3a2 2

126:1

234

Seismology

X1
f

P
S

X3

Fig. 126

where a tan e and b tan f, and f is the emergence angles of the incident S-wave and e is
that of the reected P-wave (Fig. 126). The relation between f and e according to Snells law is
p 1
cos f
cos e
a

) cos e cos f 3 ) e 30


b
a
b
2

Substituting in Equation (126.1):

VSP

p
4 3 1 1

4 1 1

p
3

We can write the potential of the incident S-wave referred to the incidence plane (x1, x3) by
means of the rotation matrix
0
1
cos az
sin az 0
@  sin az cos az 0 A
1
0
0

and substituting
0

1
B 2
B p
B
@ 3
2
0

p
1
3
8
p
0 C0 10p3 1 0 B1 1
<B1 4 3
2
C@
C
2 A @ B2 A ) B2 16 B0
1
:
0A
B3
B3 4
4
2
0 1

the potential is
c0i B1 ; B2 ; B3 exp ikb cos fx1 sin fx3  4t
p

 p

3
1
x3  4t
4 3;16; 4 exp 5i x1
2
2

The scalar potential of the SV component is

cSV B0 exp ikb cos fx1 sin fx3  4t


p

3
1
x3  4t
16 exp 5i x1
2
2

126:2

235

Reection and refraction

The amplitude of the potential of the reected P-wave is


p
A B0 VSP 16 3

From

kb

o
o
5
b
4

o 20 s1

we have

and
ka

o
20
p km1
a 4 3

The scalar potential of the reected P-wave referred to the original system of axes is
given by


A exp ika nj xj  at
where the direction cosines are now

p
p
31
3

n1 sin i cos az cos e cos az


2
2
4
p p
3 3 3

n2 sin i sin az cos e sin az


2 2
4
1
n3 cos i sin e
2
Substituting these values we obtain
p
5
16 3 exp i p
3

p
3
3
1
x 1 x 2  x 3  4 3t
4
4
2

The components of the displacement in mm are


8
p
P
>
>
< u1 20 3
uP r ) uP2 60
>
>
: P
u3 40

(b) For the reected S-wave we have to separate the SV and SH components. The SV
component can be deduced from the scalar potential
p

1
3
r
cSV B exp ikb cos f x1  sin f x3  4t B exp 5i x1 
x3  4t
2
2

We obtain B by means of the reection Vss:


VSS

4ab  1 3a2
4ab 1

3a2 2

The reected S-wave doesnt have an SV component.

B
)B0
B0

236

Seismology

For the reected SH component we use the displacement of the u2 instead of the
potential. The displacements of the SH component of the incident wave are obtained from
(126.2):
uiSH c01;3  c03;1 40
Referred to the reference of the plane of incidence, the displacement is given by
p

3
1
i
x3  4t
uSH 40 exp 5i x1
2
2
The amplitude of the reected SH wave is equal to that of the incident SH wave. Referred
to the incidence plane system of reference,
p

3
1
r
uSH 40 exp 5i x1 
x3  4t
2
2
The displacement of the reected S-wave referred to the original system of axes is
p
p

1
3
3
uri Br1 ; Br2 ; Br3 exp i5 x1
x2 
x3  4t
4
4
2
Since the SV component is zero, Br3 0; Br1 and Br2 are found using the equations
jur j urSH ) Br1 2 Br2 2 1600
p
1 r
3 r
r
B 0
ui n i 0 ) B 1
4
4 2
resulting in
p
Br1 20 3

Br2 20

127. A wave represented by the potential


x1
x2
x3
w 4 exp 0:25i p p p  4t
6
3
2

is incident on the surface x3 0 of separation between two liquids. If the speed of


propagation in the second medium is 2 km s1, the pressure exerted by the incident
wave on the surface of separation is 5  109 Pa, and the transmitted energy is four
times greater than the reected energy, calculate:
(a) The energy transmitted to the second medium.
(b) The potentials of the transmitted and reected waves referred to the same
coordinate system as the incident potential.
(a) The intensity or energy per unit surface area of the wavefront of an incident
P-wave is given in units of J m2 by
Iinc A20 o2 ka2 ar

127:1

237

Reection and refraction

From the given potential we have


A0 4  103 m2
ka 0:25 km1
a 4 km s1
o
ka ) o 1 s1
a
We need to know the value of the density r. Since the medium is liquid l K (bulk
modulus) and then l P/y, where P is the pressure and y the cubic dilatation (change of
volume per unit volume). For liquids the shear modulus m is zero and from the velocity of
P-waves we obtain,
s s
l 2m
l

) l a2 r
a
r
r
The cubic dilatation is obtained from the potential :
y r2 ka2 A

1
1
4
16
4

Then, we obtain
l a2 r

P
5  109
Pa
5

) r g cm3
2
2
1
y
m s
4
 16  106
4

By substitution in (127.1) the incident energy is


Iinc 16  1 

1
5
 4  5 J m2
16
4

The energy transmitted into the second medium is


Itras A02 o2 ka02 a0 r0 W 2 A20 ka02 a0 r0

127:2

Irefl Ao2 ka02 a0 r0 V 2 A20 ka02 a0 r0

127:3

and the energy reected is

where W and V are the transmission and reection coefcients, respectively:


A0
2r tan e

A0 r0 tan e r tan e0
A
r0 tan e  r tan e0
0
V
A0 r tan e r tan e0

The emergence angle e of the incident wave is


1
n3 p cos i sin e ) i e 45
2

127:4

238

Seismology
and from Snells law the emergence angle of the transmitted wave e0 is
r
p
cos e cos e0
a0
2 1
1
1
7
0
0
0 ) cos e cos e p p ) sin e 1  p
a
a
a
4 2 2 2
8 2 2
Given that the transmitted energy is four times the reected energy,
Itras
Iref

W 2 A20 r0 o4
V 2 2r0 a
0
4 2 a2 4 ) 2
W
5ra0
V A0 ro
a

If we substitute in (127.4)
5 p
7
4
V
5 p
7
r0
4
5
2
4
W
p
5
7
r0
4
r0 

We have three equations for r, V, and W. The solution for positive values of the variables
is

W 0:23
r0 7:7 )
V 0:40
(b) The potential of the reected P-wave is
ref VA0 exp ik a n1 x1 n2 x2  n3 x3  at

1 1
1
1
p x1 p x2  p x3  4t
1:6 exp i
4
6
3
2

To determine the potential of the transmitted wave we have to calculate the direction
cosines of the transmitted ray. The azimuth is the same as that of the incident wave which
can be deduced from the direction cosines and the value of i:
1
1
n1 sin i cos az p ) cos az p
6
3
p
1
2
n2 sin i sin az p ) sin az p
3
3

The direction cosines of the transmitted ray in the medium M0 are


1 1
1
n01 sin i0 cos az p p p
2 2 3 2 6
1
n02 sin i0 sin az p
2 3
p
7
n03 cos i0 p
2 2

239

Reection and refraction

and the potential of the transmitted wave is


p

1
1
1
7
p x1 p x2 p x3  2t
tras 0:92 exp i
2 2 6
2 3
2 2

128. Two liquid media are separated at x3 0, the rst of volumetric coefcient
1
K  109 Pa and density 1 g cm3. The amplitudes of the components of an
p
2

incident wave of frequency 3 Hz
pare
 ui 18p 1;1; 6 mm and those of the wave
63 2p pp p
transmitted to medium 2 are
2; 2; 3 mm. Given that the amplitude of
7
the transmitted potential is twice that of the reected potential, nd expressions for
the incident, reected, and transmitted potentials.
In liquids only P-waves are propagated and their displacements can be deduced from the
scalar potential
A0 exp ika n1 x1 n2 x2 n3 x3  at ) uP r
Then in our case the components of the displacement in mm are
uP1

@
A0 ka n1 A0 ka sin i cos az 18p
@x1

uP2

@
A0 ka n2 A0 ka sin i sin az 18p
@x2

uP3

p
@
A0 ka n3 A0 ka cos i 18p 6
@x3

and we nd az 45 ; i 30 ; e 60 , A0 6  103 m2.


The emergence angle of the refracted wave, e0 , can be found from its displacements,
p p p p

p
2 2 3
3
63p 2
2
0
0
P
p ; p ; p ) n3 sin e p ) cos e0 p
utras p
7 7 7
7
7
7
The P-wave velocity in the medium of the incident wave is
s s
l 2m
K
1
a

p km s1
r
r
2

Using Snells law we nd the velocity of the medium of the refracted wave,
p
cos e
a
2 2
0
) a p
cos e0 a0
7

From the values of the velocities in the two media we calculate their densities:
9
1 0
>
>
r
2
0
>
a
7
Kr
>
2
>
=

02
0
0
3
3
a
16 K r
K
) r0 r g cm3
p

>
2
2
q 2 2
>
8 0>
>
0
K0
0
>
a r0 p ) K r ;
7
7

240

Seismology

The reection V and transmission W coefcients are found using their expressions and from
them we get the relation between the amplitude A0 of the incident wave potential and those
of the reected A and refracted A0 waves, and substituting the value for A0 6, we obtain
A0 6 ) A 3  103 m2
From these values we can write the potentials of the incident, reected, and transmitted
waves:
p

p 1
3
1
1
x3  p t
inc 6 exp i6p 2 p x1 p x2
2
2 2
2 2
2
p

p 1
3
1
1
x3  p t
ref 3 exp i6p 2 p x1 p x2 
2
2
2 2
2 2
p
p

p
p p
2
2
2
3
7
tras 6 exp i3p p p x1 p x2 p x3  2 p t
7
7
7
7
2

129. Two liquids in contact have speeds of propagation of 4 and 6 km s1. The density
of the rst is 2 g cm3 and is less than that of the second. For waves of normal
incidence, the reected and transmitted energies are equal. A wave of v 1 s1 and
with a potential of amplitude A0 2103 cm2 is incident from the rst onto the
second at an angle of 30 . Calculate:

(a) The transmitted and reected energies.


(b) An expression for the transmitted potential.
(a) For normal incidence, the reection and transmission coefcients in terms of the
refractive index m a/a0 and the density contrast m r0 /r are given by
mn
mn
2
Wn
mn
Vn

If the reected energy is equal to the transmitted energy, then


Vn2 mnWn2 )

m  n2
m n2

4mn
m n2

Substituting n a/a0 2/3, from (129.1) we obtain the value of m:



4
m 3:9
4
2
m  4m 0 )
m 0:1
9
Trying both values we obtain for r0
r0
0:1 ) r0 0:2 g cm3
r
m 4 ) r0 8 g cm3
m

129:1

241

Reection and refraction

i = 30

r = 2 g cm3

M
e

a = 4 km s1

a = 6 km s1
M

Fig. 129

But the problem states that r 4 < r0 , so r0 8 g cm3.


For a wave with incidence angle 30 (e 60 ) the emergence angle of the transmitted
wave e0 is, according to Snells law (Fig. 129), given by
r p
cos e cos e0
3
9
7
0
0
0 ) cos e ) sin e 1 

a
4
16
a
4
The partition of energy between the reected and refracted waves is given by
sin e0 2
W V2 1
sin e
The reection V and transmission W coefcients are
p
p
3 2 7
m sin e  n sin e0 4 2  3 4
p 0:8
p
V
m sin e n sin e0
3 2 7
4

2 p3 4
3
2 sin e
p 0:46
W
p
0
m sin e n sin e
3 2 7
4

2
3 4
mn

The incident, reected, and transmitted energies per unit time and surface area are (Problem 127)
ro4 2
A sin e 17:3 erg cm2 s
a 0
ro4 2
ro4 2 2
Eref
A sin e
A V sin e 11:1 erg cm2 s
a
a 0
r0 o4
r0 o4
Etrans 0 A02 sin e0 0 A20 W 2 sin e0 7:5 erg cm2 s
a
a
Einc

242

Seismology

(b) The potential of the transmitted wave is


tras A0 exp ik x3 tan e0 x1  ct
where
A0 A0 W 2  103  0:46 920 cm2
o
1
k ka0 0 cos e0 0 cos e0 km1
a
8
a0
c
8 km s1
cos e0
p

7
1
x3 x1  8t
tras 920 exp i
8 3
130. An SV wave is incident on the free surface of an elastic medium of Poisson ratio
0.25. If the potential of the wave is (in units of 103m2)
p

3
5
5
1
1
c i p ;  p ; 0 exp i p x1 p x2
x3  4t
2
2
2
2 2
2 2

nd the components of the amplitude of the reected P-wave referred to this set of axes.
From the direction cosines we nd the incidence angle i, the emergence angle f, and the
azimuth az of the incident SV wave (Fig. 130):
p
3
n3 cos i sin f
) i 30 and f 60
2
1
1
n1 sin i cos az cos az p ) az 45
2
2 2

Bearing in mind that Poissons ratio is 0.25, from Snells law we nd the emergence angle
e of the reected P-wave:
p
p
s 0:25 ) l m ) a 3b 4 3 km s1
p
cos f
cos e
3

) cos e
) e 30
b
a
2
X3

f
f

SV

Fig. 130

243

Ray theory. Constant and variable velocity

The reection coefcient for the reected P-wave gives us the relation between the
amplitude of the potential, B0, of the incident SV wave and A, that of the reected P wave:
VSP

p
A
4 tan f 1 3 tan2 e

 3
2
B0 4 tan e tan f 1 3 tan2 e

To nd B0 we write the potential of the incident SV wave referred to the (x1, x3) plane of
incidence using the rotation matrix
0 5 1
0
1
0
1
p
0
cos az
sin az 0 B 2 C
B
C
C
C
B
C B
@  sin az cos az 0 AB
B  p5 C @ 5 A
@
2A
0
0
0
1
0
p
B0 5  103 m2 ) A 5 3  103 m2
Referred to this system of axes the potential of the reected P-wave is given by
A exp ik x3 tan e x1  a t
kb
1
1 2 km1
cos f
2

p
p
1
5 3 exp i2 x3 p x1  4 3t
3
k

The amplitudes of the displacements of the reected P-wave referred to this set of axes are
p
@
10 3 mm
@x1
@
u3
10 mm
@x3

u1

Referred to the original set of axes the horizontal components are


p
10 3
0
p
mm
u1 u1 cos az
2
p
10 3
u20 u1 sin az p mm
2

Ray theory. Constant and variable velocity


131. Assume that the Earths crust consists of a single layer of thickness H and a
constant speed of propagation of seismic waves of v1 on top of a mantle of velocity of
propagation 20% greater than the crust. Given that a focus on the surface produces a
reected wave that takes 17.2 s to reach a distance of 99 km, and that this is the

244

Seismology

xc
F

H
ic

ic

v1

v2
Fig. 131a

critical distance, calculate the values of H, v1, and v2. Plot the travel-time curve (t, x)
for this specic case with numerical values.
The critical distance xc is the distance at which a ray that is reected with the critical angle
at the top of the mantle arrives at the surface and is given by the equation (Fig. 131a)
xc 2H tan ic 99 km

131:1

where H is the thickness of the crust. Since we know the relation between the velocities in
the crust and the mantle, we can calculate the critical angle
v2 1:2v1 )

sin ic
1
1
) ic 56:44
) sin ic
v1
v2
1:2

If we substitute in Equation (131.1) we obtain the thickness of the crust, H:


99 2H tan 56:44 ) H 32:8 km
The travel time of the critically reected ray is
H
H
2
17:2
v1 cos ic
v1 cos 56:44
32:84
) v1 2
6:9 km s1 ) v2 8:3 km s1
17:2 cos 56:44
t2

To draw the travel-time curve for different distances of the direct, reected, and critically
refracted waves we use the equations
x
t1
v1
r
2 x2
H2
t2
v1 4
p
x 2H v22  v21
t3
v2
v1 v2

245

Ray theory. Constant and variable velocity

We obtain the following values


x(km)

t1 (s)

t2 (s)

t3 (s)

0
30
60
90
99
120
150

0
4.3
8.7
13.0
14.3
17.4
21.7

9.5
10.4
12.9
16.1
17.2
19.8
23.7

17.2
19.7
23.4

The travel-time curves are drawn in Fig. 131b.

40
3

t (S)

30

20
1

10

ti

0
0

Fig. 131b

50

xc

150
x (km)

200

250

300

246

Seismology

x
P

Fig. 132

132. In a seismogram recorded at a regional distance, the S-P time lag is 5.5 s, and the
focus is at a depth x/2, where x is the epicentral distance. The
p model1Earth has a single
layer of Poisson ratio 0.25 and constant S-wave velocity 3 km s . Calculate:

(a) The depth of the focus.


(b) The epicentral distance.

(a) For a direct wave from point F to point P (Fig. 132) the difference of the arrival
times of the P- and S-waves (the S-P interval) is
t S-P 5:5

FP FP

b
a

The distance FP can be expressed in terms of x as (Fig. 132)


r
p
 x 2
p
5
FP x2 h2 x2
x
2
2
The S-P interval is given by

p
5ab
5:5 x
2 ab
Since Poissons ratio is 0.25 and knowing the S-wave velocity we obtain
p p
p
5 31
pp
s 0:25 ) a b 3 ) 5:5 x
2
3 3
x 21 km
x
h 10:5 km
2
133. The Earth consists of a layer of thickness 20 km and seismic wave velocity
6 km s1 on top of a medium of speed of propagation 8 km s1. A seismic focus is

247

Ray theory. Constant and variable velocity

x
h
H

F
ic

ic

ic
i

v1
v2

Fig. 133

located at a depth of 10 km. Calculate the difference in travel times between the
reected and the critical refracted waves observed on the surface at a distance of
150 km from the epicentre.
This problem is similar to Problem 131, but now the focus is at depth h 10 km.
The critical distance in this case is given by (Fig. 133)
xc 2H  h tan ic

v1
1 6
sin ic ) ic sin
48:6
v2
8
) xc 2  20  10 tan48:6 34:0 km
Since the distance 150 km is greater than the critical distance there arrive critically
refracted rays. The travel times of the reected (t2), and critically refracted (t3) rays at that
distance are
q q
x2 2H  h2
1502 2  20  102
25:5 s
t2

v1
6
p
p
x 2H  h v22  v21 150 2  20  10 82  62
t3

22:1 s
v2
8
v1 v2
86
The time difference between the travel times of the two rays is
t3  t2 22:06  24:49 3:4 s
134. Consider a crust of thickness H and constant speed of propagation v1 on a mantle
of constant speed of propagation v2. A seismic focus is located at depth H/2, the
critical distance is 51.09 km, the delay time is 4.96 s, and the critical angle is 48.59 .
Calculate the values of H, v1 and v2, and the depth of the focus.
For a focus at depth h H/2, the travel times of the critically refracted (t3) rays and the
critical distance are given by the expressions (Fig. 133).

248

Seismology

p
x 2H  h v22  v21
t3
v2
v1 v2
xc 2H  h tan ic ) 51:09

H
2H 
tan48:59
2

) H 30 km; h 15km

Knowing the depth and thickness of the crust, using Snells law, the value of the critical
angle, and the delay time ti, we nd the velocities v1 and v2:
sin ic
1
) v1 0:75v2
v1
v2
q
p
2
2

2

30

15

v22  0:75v2 2
2H  h v2  v1
ti
) 4:96
v1 v2
0:75v22
) v2 8 km s1

v1 6 km s1
135. In a seismogram, the S-P time difference is equal to 5.31 s, and corresponds to a
regional earthquake that occurred at a depth h 2H, where H is the thickness
of the crust. Given that the crust is formed by a layer of constant P-wave velocity
of 3 km s1, that below it there is a semi-innite mantle of double that speed of
propagation, and that Poissons ratio is 0.25, determine:
(a) An expression for the travel-time of the P- and S-waves.
(b) The epicentral distance for an emerging P-wave with a take-off angle of 30 at the
focus.
(a) The travel time corresponding to the ray given in Fig. 135 is given by
t

FA AS

2v
v

i0

i0

2H
A
ih

Fig. 135

2v

249

Ray theory. Constant and variable velocity

where v a for P-waves and v b for S-waves. Using Snells law we nd the relation
between the incidence angle at the focus ih and at the station i0:
sin ih sin i0
1

) sin i0 sin ih
2v
v
2
From Fig. 135 we obtain
H
H
) FA
cos ih
FA
H
H
cos i0
) AS
cos i0
AS

cos ih

From these equations we deduce the expression for the travel time:
t

H
H

2v cos ih v cos i0

135:1

For the epicentral distance we obtain


x F0 A AS0 H tan ih H tan i0

135:2

Using Equations (135.1) and (135.2) and putting v a we obtain the travel time and
epicentral distance for P-waves and putting v b for S-waves.
(b) For a P-wave with take-off angle at the focus (ih) of 30 , we rst nd the value of
the angle at the station i0,
sin ih sin i0

) i0 14:47
2a
a
Substituting in (135.1) we obtain
tP

H
H
H
p
1:61
a0:97
a
3
2a
2

The travel time of the S-wave with take-off angle jh 30 can also be calculated using
(135.1). Since Poissons ratio is 0.25, the velocity of the S-wave is
p
p
a
s 0:25 ) a 3b ) b p 3 km s1
3

The incidence angle at the station, j0, using Snells law, is given by
sin jh sin j0

) j0 14:47
2b
b
and the travel time is
tS

H
H
H

2:79
2b cos jh b cos j0
a

Since we know the S-P time interval we can obtain the value of h:
t S-P 5:31 2:79

H
H
 1:61 ) h 13:61 km
3
3

250

Seismology

The epicentral distance is found using (135.2):


x 13:61 tan 30 13:61 tan 14:47 11:41 km
136. Consider a crust composed of two layers of thickness 12 and 18 km, and constant
P-wave speeds of propagation of 7 and 6 km s1, respectively, on top of a semi-innite
mantle of constant speed of propagation 8 km s1. There is a seismic focus at a depth
of 6 km below the surface. For a station located at 100 km epicentral distance,
calculate the travel time of the direct, reected, and critical refracted waves (neglecting waves with more than a single reection or critical refraction).
The travel times of the direct ray t1 and the ray reected on the bottom of the rst layer
t2 are given by (Fig.136)
p p
h2 x 2
62 1002
14:3 s

t1
v1
7
q q
2H1  h2 x2
2  12  62 1002
t2

14:5 s
v1
7
As the velocity of the second layer is less than that of the rst layer there is no critical
refraction at that boundary. There is critical refraction at the boundary between the second
layer and the mantle where the velocity is greater. Using Snells law, we can calculate the

h
F
2 (H1 h)

H1

n1
i1

i1

F
H2

n2

ic

ic

n3

Fig. 136

251

Ray theory. Constant and variable velocity

critical angle ic and from this value the incidence angle at the focus i1 for the critically
refracted ray:
sin i1 sin ic
1
6

) sin ic ) ic 48:6
v1
v2
v3
8
7
sin i1 ) i1 61:0
8
The travel time t3 of the critically refracted ray at the bottom of the second layer is given by
FA AB BC CD DS

v1
v2
v3
v2
v1
If the epicentral distance x is 100 km, the different segments of (136.1) are
t3

136:1

H1  h H1

) FA 12:4 km; DS 24:8 km


FA
DS
H2
cos ic
) AB CD 27:2 km
AB
BC x  2H2 tan ic  H1  h tan i1  H1 tan i1 26:8 km

cos i1

Finally, by substitution in (136.1) we obtain,


t3 17:7 s
137. Consider a two-layered structure of thickness H and speed of propagation v and
3v on top of a half-space medium of speed of propagation 2v. At a depth 3H below the
surface there is a seismic focus. Write the expressions (as functions of H, v, and ih) for
the travel times of waves that reach the surface without being reected. Give the
range of values of ih.
In this problem the focus is located at the half-space medium at depth h3H under its
boundary. Applying Snells law we can nd the relation between the velocities, the
incidence angles at the focus and at the bottom of each layer, and the critical angle at the
boundary between the second layer and the half-space (Fig. 137):
sin ih sin i2 sin i1

2v
3v
v
sin ic
1
) ic 41:8
2v
3v

137:1

The rays which leave the focus and arrive at the surface at a distance x are only those with
angles less than the critical angle (Fig. 137). The travel time for these rays is
t

FA AB BS
H
H
H

2v
3v
v
2v cos ih 3v cos i2 v cos i1

137:2

According to Equation (137.1) we have the relation between the incidence angles:
q
q
3
1
4  9 sin2 ih
sin i2 sin ih ) cos i2 1  sin2 ih
2
2
q
q
1
1
sin i1 sin ih ) cos i1 1  sin2 ih
4  sin2 ih
2
2

252

Seismology

i1
H

i0

H
i2

3v

A
ic
ih

2v

Fig. 137

Substituting in (137.2) we write the travel time as function of the take-off angle ih:
!
H
1
2
2
t
p p
v 2 cos ih 3 4  9 sin2 ih
4  sin2 ih
We nd a similar expression for the epicentral distance x:
x H tan ih H tan i2 H tan i1
!
sin ih
3 sin ih
x H tan ih p p
4  sin2 ih
4  9 sin2 ih

The range of values of the take-off angle for rays which arrive at the surface is
0 < ih <42 .
138. A semi-innite medium consists of two media of velocities v and 3v separated by a
vertical surface. In the rst medium there is a focus of seismic waves at a depth a
below the free surface and at the same distance a from the surface separating the two
media. Write the expressions for the direct, reected, and transmitted waves arriving
at the free surface, and plot the travel time curve (t, x) in units of a/v and a (neglecting
waves with more than a single reection).
In this situation we have the following rays arriving at the surface: direct in the rst
medium, reected at the boundary, and critically refracted and refracted to the
second medium. We consider two cases for rays arriving at distances 0 < x < a and
distances x > a.

253

Ray theory. Constant and variable velocity

S
x

a
3v

Fig. 138a
S

E
x

ic

a
A

ic
ic

a
v

3v

Fig. 138b

(a) For 0 < x <a, the travel time t1 of the direct wave is (Fig. 138a)
p
FS
x 2 a2
t1

v
v
The travel time t2 of the reected ray is
q
2a  x2 a2
FP PS AP PS AS

t2
v
v
v
v
The reected rays exist also for negative distances, but we will not consider them.
The travel time t3 of the critically refracted ray (Fig. 138b) is
t3

FA AB AS

v
3v
v

138:1

254

Seismology

E
xc
ic

a
ic
ic
F

ic
a

3v

Fig. 138c

Using Snells law the critical angle is given by


sin ic
1
) ic 19:47
3v
v
and
cos ic

a
ax

FA
AS

The distance AB is given by


AB a  a tan ic  a  x tan ic a  tan ic x  2a
By substitution in (138.1) we obtain
a 0:94
x
t3 2:22 
v
v
The critical distance (distance of the ray reected with the critical angle) is given by
(Fig. 138c)
xc a  SB

138:2

From Fig. 138c we obtain


tan ic

BP a  a tan ic

) SB 1:83a
SB
SB

and substituting in (138.2)


xc 0:83a
Critically refracted rays exists for distances 0.83a < x < a. Here we consider only those
for 0 < x < a.

255

Ray theory. Constant and variable velocity

S
X

a
A

i
a
3v

Fig. 138d

2.0
2

t (a/v)

1.5

1
1.0

0.0

0.0

0.0

0.2

0.4

0.6

0.8

1.0

x (a)

Fig. 138e

(b) For distances x > a, we have the rays refracted at the boundary between the two
media when the incidence angle is less than the critical angle (Fig. 138d), that is,
i <19.47 :

256

Seismology

FA AS

v
3v
a
FA
cos i
xa
AS
cos e
t4

Using Snell`law
sin i sin e

) sin e 3 sin i
v
3v
The travel-time is given by
t4

a
xa
p
v cos i 3v 1  9 sin2 i

The travel-time curves for direct (1), reected (2), critically refracted (3) and transmitted
(4) waves are given in Fig. 138e.
139. Given the structure in the diagram, calculate the arrival times of the direct and
(non-reected) transmitted waves for x 0, where x 0 is a point on the free surface
in the vertical above the focus.
At x 0 )ih 0 , the travel-time of the vertical ray is (Fig. 139a)
t

a a 3a

2v v 2v

For rays arriving at x > 0 and leaving the focus with take-off angles 0 < ih < 45 , the
travel-times are given by
FA AS
a
a

2v
v
2v cos ih v cos r
a
cos ih
FA
a
cos r
AS
t

a
2a

2v

Focus

Fig. 139

139:1

257

Ray theory. Constant and variable velocity

X
E

r
2a

2v
a
ih
F
a

Fig. 139a

Applying Snells law,


sin ih sin r
1

) sin r sin ih
v
2
2v

139:2

Substituting in (139.1):
a
1
2
t
p
v 2 cos ih
4  sin2 ih

139:3

The relation between the epicentral distance x and the incidence angle ih is
x F0 A A0 S
F0 A a tan ih
A0 S a tan r
sin ih
x atan i tan r a tan i p
4  sin2 ih

From Fig. 139a we deduce:

tan ih

F0 A
a

tan r

x  F0 A x
 tan ih
a
a

Using Equation (139.2),


x
sin ih
2
sin r
x
1
2
a
2

 tan ih )


cos r cos r a
cos r sin ih cos ih

258

Seismology

x
E

n
2a
r
2n

a
ih

ih
F

F
a

Fig. 139b

By substitution in (139.3),
t

2x
3a

v sin ih 2v cos ih

For ih 45 the epicentral distance is


1
x a 1 p
7

This is the limit of the epicentral distance at which these rays arrive. The corresponding
time limit is
t

a 1
4
p p
v
2
14

For angles ih > 45 (Fig. 139b), the travel time and epicentral distance, as a function of the
take-off angle ih, are
t

FA AS
a
xa
a
xa

sin ih
2v
v
2v sin ih v sin r 2v sin ih
v
2

x a A0 S
A0 S
2a  F0 A
a
F0 A
tan ih

a
sin ih
p
x a 2a 
tan ih
4  sin2 ih
tan r

259

Ray theory. Constant and variable velocity

Using
cos ih cos r
1

) cos r cos ih
2v
v
2
1 p
4  cos2 ih
sin r
2p
4  cos2 ih
tan r
cos ih
we obtain for x and t,

p
a
4  cos2 ih
tan ih
cos ih
a
2x  a
p
t
2v sin ih v 4  cos2 ih

x a 2a 

For i 90 , as expected the ray doesnt arrive at the free surface.
140. For the structure in Fig. 140a, write the equations of the travel times of the
direct, reected, and transmitted waves (neglecting waves with more than a single
reection) as a function of the epicentral distance. Determine the times of intersection,
and the minimum and maximum distances in each case in terms of a/v and v. Plot the
travel-time curves.
The travel-time of the direct wave for distance 0 < x < 1 is (Fig. 140a)
p
4x2 a2
t
2v
For the ray reected on the horizontal surface at depth a the travel time is
s

2
3a
p
x2
2
4x2 9a2

t
2v
v
x
S
a /2
F

Fig. 140a

260

Seismology

x
S
a/2

a
a

2v

Fig. 140b

The range of distances for this ray is


xmin 0
a xmax
xmax ) a
) xmax 3a
3a
2
2
For the reected ray on the surface at depth 2a the travel time is (Fig. 140b)
s

2
7a
2
p
x
2
4x2 49a2
t

2v
v
The range of distances is
3a
3
a
2

xmin

7
2

) xmin 7a ) tmin

p
7 5a

2 v

xmax 1
The critically refracted ray on the surface at depth a, using the general expression, is given by
p
p
x 2H  h v22  v21
x a3 3

t
v2
v 1 v2
2v v 4
The minimum distance for this ray corresponds to the critical distance:
sin ic
1
) ic 30

v
2v
p
3
a
a
xc tan ic a tan ic
2
2
and the maximum distance is
xmax

p

a 3 3
a
a p
3
3

261

Ray theory. Constant and variable velocity

10

t (a/v)

2
2
4

6
X (a)

10

12

Fig. 140c

On the surface at depth 2a there is no critically refracted ray, since the minimum take-off
angle ih at that surface is
a
tan ih a 2 ) ih 63:4
2

greater than the critical angle 30 .
The travel-time curves are drawn after rewriting the equations in units of a/v and a, and
are represented in Fig. 140c
tv2  x 2
x
1. Direct ray:

14 ; 0 < 1 (a hyperbola)
a
a
a  2  
tv
x 2 9
x
2. Reected ray on the surface at depth a:

; 0 3 (a hyperbola)
a
a
4
a
tv2  x2 49
x

; 7 1 (a hyperbola)
3. Reected ray on the surface at depth 2a:
a
a
4 p
a
tv 1 x 3 3
x
4. Critically refracted ray on surface at depth a:

; 0:87 1:58 (due


a 2a
4
a
to the short range of distances this is not noticeable in the gure)

262

Seismology

141. For the structure in the diagram, assume a seismic focus at the surface, and
calculate the travel time of the direct, reected, and critical refracted waves for
epicentral distances between 0 and a. Calculate the critical distance, and the expression for the transmitted wave.
Since the focus is at the free surface, the travel time of the direct ray is simply given by
(Fig. 141a)
t

x
v

0<x<a

a
F

45

Fig. 141

a
x
S

45
P
a
a 2

45

Fig. 141a

263

Ray theory. Constant and variable velocity

a
x
S

45

45

ic

ic

B
A

2
a

45

Fig. 141b

The travel time of the reected wave is (Fig. 141a)


t

FP PS F0 P PS 1 p

x2 2a2  2ax
v
v
v
v
v

The critical angle is given by

p
sin ic
1
1
2 2
) ic 19:47
) sin ic ) cos ic
v
3v
3
3
The travel-time of the critically refracted ray is (Fig. 141b)
FA AB BS

v
3v
v
a
p
3
2
a
FA
cos ic 4
t

The distance BS can be found using the sine law in triangle SBD:
sin90  ic sin 45

ax
BS

BS a  x

3
4

and the distance AB (calling d BD) is


p
p
p
p
2
2
2
tan ic  d a
a
tan ic  d
AB a 2  a cos 45  a
2
2
2

264

Seismology

a
xc
s

D
45

ic

45

ic
A

2
2

45

Fig. 141c

The distance d, using the sine law, is given by


sin45 ic sin90  ic
sin45 ic

) d a  x
d
ax
sin90  ic
p
sin45 ic
2
AB a
0:96x  0:50a
1  tan ic  a  x
sin90  ic
2
The travel time is given by
x
a
t 0:43 1:33
v
v
Finally we determine the critical distance xc from the triangle AS0 D (Fig. 141c):
b 90  ic 45 180 ) b 64:47
sin90  ic
sin b
) xc 0:52a
a
a
a  xc
p  p tan ic
2
2

142. A medium consists of a at crust of thickness H and constant speed of propagation


v1 on a semi-innite mantle of constant speed of propagation v2. For a focus at the
surface, at a distance x the direct wave arrives at a time t1 x/a, the critical distance is
p
2a
xc p, and the direct and critical refracted waves intersect at the distance x 2a 3.
3
(a) Calculate the crusts thickness, its speed of propagation, the mantles speed of
propagation, and the critical angle.

265

Ray theory. Constant and variable velocity


(b) Assume now that this is a layer which dips downwards at 45 with the parameters
of the model being those determined in the previous part. Calculate the travel
times of the reected and critical refracted waves at x a, 3a, and 5a.
(a) We determine the velocity of the crust from the travel time of the direct ray:
t1

x
x
) v1 a
v1 a

The critical distance is given by


v1
a
a
2 p
xc 2H tan ic 2H p
2H p
2
2
2
2
3
v2  v1
v2  a

142:1

Equating the travel


p times of the direct and critically refracted ray for the value of the
distancex 2a 3 we obtain
p
x
x 2H v22  v21
142:2
t1 t3 )
v1 v2
v1 v2
From Equations (142.1) and (142.2) we obtain the values of H and v2:
H a
v2 2a
The critical angle may be estimated from
2a
x0c 2H tan ic ) p 2a tan ic ) ic 30
3

(b) We now consider the case of a dipping layer with dip angle y 45 . The critical
distance is now given by the equation (Fig. 142)

xc

F
*

ic

xc sin

Fig. 142

266

Seismology

xc cos y H tan ic H xc sin y tan ic

142:3

so
xc

2H tan ic
cos y  sin y tan ic

The critical distance along the horizontal free surface xc is found by substituting H, y, and ic
in (142.3)
xc 3:86a
The travel times of the reected and critically refracted rays for a dipping layer are given by
the equations
p
4H 2 x2 4Hx sin y
t2
v1
q
x cos y x sin y 2H
t3

v22  v21
v2
v1 v2
By substitution of the values of H, v1, v2, and y, we obtain
p
p
4a2 x2 2 2ax
t2
pa
2
p
p
p p
x 2 x 2 2a p x 2 
3

1 3 3
t3
2a
4a
4a

For the required values of x, we obtain the following values of the travel time in units
of a/v:
x

t2 (a/v)

t3 (a/v)

a
3a
5a

2.80
4.64
6.57

6.56

Since the critical distance is 3.86a, it only exists for x 5a.


143. In a at medium, the velocity increases linearly with depth according to the
expression v v0 kz, where v0 is the velocity at the surface, k is a constant, and z is
the depth. For a focus at depth h, calculate an expression for the take-off angle at the
focus ih in terms of the epicentral distance x, and v0, h, and k.
At the maximum depth of penetration of the ray r the incidence angle of the ray with the
vertical is equal to 90 (Fig. 143). Using Snells law we can relate the angles at the focus ih,
at the bottom of the ray 90 , and i0, the incidence angle at the station on the surface:
sin i0
1
sin ih
v0 kh

) sin ih

v0 kr v0 kh
v0 kr
v0

143:1

267

Ray theory. Constant and variable velocity

x
a

0 / k

R
S

E
V0

r
F

i0

ih

Fig. 143

The problem is solved if we can express r as a function of x, v0, h, and k. The epicentral
distance x is the sum of a and b (Fig. 143):
r
v 2
0
b R2 
k
But we know that for a distribution of velocities which increases linearly with depth the
rays are circular and their radius R is

r
2 v 2
v0
v0
0
r 
R r )b
k
k
k
and

r
v
2 v
2
0
0
a
r 
h
k
k

Therefore,

r
r

2 v 2
2 v
2
v0
v0
0
0
bxa)
r 
x
r 
h
k
k
k
k

Solving for r we obtain

v
u0
12
u x 2  h2  2 v 0 h


u
k A v0 h 2  v0
r t@
2x
k
k

By substitution of this expression for r in (143.1) we obtain the required expression


for ih.

268

Seismology

144. Consider a semi-innite medium in which the velocity increases linearly with depth
according to the expression v 4 0.1 z. There is a seismic focus at a depth of 10 km.
Calculate the epicentral distance reached by a wave leaving the focus at an angle of 30 .
The velocity at the focus is found directly by putting in the equation for the distribution of
velocity, z h:
vh 4 0:1  10 5 km s1
According to Snells law we nd the velocity at the point of greatest depth penetration
(i 90 ) of the ray (Fig. 144):
sin ih
1
5
10 km s1

) vm
vm
sin 30
vh
From this value we nd the depth to that point:
vm 10 4 0:1  r ) r 60 km
Knowing that the rays are circular of radius R,
v0
R r 60 40 100 km
k
As in the previous problem, the epicentral distance (from point E to S) is (Fig. 144):
xab
where
r
v
2
0
h 86:60 km
a R2 
k
r
v 2
0
b R2 
91:65 km
k
so
x 178:25 km
x
a

0/k

R
S

V0
h
r
F

i0

ih

Fig. 144

269

Ray theory. Constant and variable velocity

145. A at medium consists of a layer of thickness H and constant speed of propagation v on top of a medium of variable speed of propagation v v0 k(z  H) where z
is the depth and k is a constant. If there is a focus at the surface:
(a) Write expressions for the epicentral distance x and the travel time t as functions of
the angle of incidence i0 at the surface.
(b) If H 10 km, k 0.1 s1, and v0 6 km s1, calculate the angle of incidence of a
wave that reaches an epicentral distance of 140 km.
(a) As we saw in Problem 143, for a distribution with a linear increase of velocity with
depth, now in the medium under the layer, v v0 k(z  H), the rays are circular with
v0
radius R r where r is the maximum depth of penetration of the ray (Fig. 145).
k
The travel-time of the ray that crosses the layer and penetrates the medium is given by
t2

FP 2
kx0
sinh1
v0 k
2v0

145:1

In the layer of constant velocity the path is a straight line and in the medium it is circular.
The epicentral distance x (from F to S) is (Fig. 145)
x x0 2H tan i0
The length of the straight ray in the layer is
FP

H
cos i0

Substituting in (145.1):
0

2H
2
1 kx
sinh
t
v0 cos i0 k
2v0

145:2

Since the layer has constant velocity the angle i0 is the same at the focus as at the bottom of
the layer at the boundary with the medium. According to Snells law
x
F

i0

i0
x

i0

0/k
P

Fig. 145

Q
r

270

Seismology

sin i0
1
v0

) vm v0 kr
v0
v0 kr
sin i0
where r is the maximum depth reached by the ray in the medium and vm the velocity at that
depth. According to Fig. 145,
0
2   
 2 v2
x
v0 2
v0
v2
v0

r m2 2 0 2 ) x0 2 cot i0
145:3
2
k
k
k
k
k sin i0
The epicentral distance x is given by
x x0 2H tan i0

2v0
cot i0 2H tan i0
k

145:4

Substituting in (145.2) the expression for x 0 in terms of i0 (143.3) we obtain


t

2H
2
sinh1 cot i0
v0 cos i0 k

(b) By substituting the given values in (145.4), we obtain



i0 45
26
140
cot i0 2  10 tan i0 )
0:01
i0 80:5
146. Beneath a layer of thickness H of velocity distribution v v0 kz there is a semiinnite medium of speed of propagation v1 2(v0 kH).
(a) Determine expressions (as functions of the above parameters) for the critical
distance, the time of intersection of the reected wave, and the maximum distance
of the direct wave.
(b) For H 10 km, v0 1 km s1, and k 0.1 s1, calculate these parameters and
plot the travel-time curves.
(a) In a layer of thickness H with variable velocity the epicentral distance x for a
reected ray is given by
H
x 2 tan i dz
0

Using the ray parameter p sin i/v we can write


pv
sin i vp ) tan i p
1  p2 v 2

The epicentral distance for a ray reaching the bottom the layer is given by
H

vp
2 H
kpv0 kz
qdz
x 2 p dz
2
2
k
1p v
0
0
1  p2 v kz2
0

For a ray incident at the bottom of the layer at the critical angle, we have
p

sin ic
1

v0 kH 2v0 kH

146:1

271

Ray theory. Constant and variable velocity

xmax
xc

n0/k
S

H
ic

n1

Fig. 146a

Substituting this expression in (146.1) and evaluating the integral, making the change of
variable u v0 kz, we obtain the critical distance
"p s#
3
4v0 kH
v20
 1
146:2
xc 
2
k
4v0 kH 2
The intercept time for x 0, corresponding to the time of the reected vertical ray (p 0),
is given by
H
dz
2 v0 kH
ln
146:3
ti 2
k
v0
0 v0 kz
The maximum distance xmax corresponds to the last ray propagated inside the layer
v0
without penetrating into the medium and has a circular path of radius R H
k
(Fig. 146a):
r
x 2 v 2 
v0  2
2v0
max
0

H
) xmax 2H 1
146:4
2
k
k
kH
(b) For the particular case with the values, H 10 km, v0 1 km s1, and k 0.1 s1,
the velocity at the bottom of the layer H is
vH v0 kH 1 10  0:1 2 km s1
The velocity of the medium is
v1 2v0 kH 21 10  0:1 4 km s1
The critical distance, using Equation (146.2), is
42
xc 
0:1

"p s #
p
p
3
1

20
3
2

5 8:2 km
 1
2
422

272

Seismology

30

25

t (s)

20

15

10

10

20

30

40

x (km)

Fig. 146b

The intercept time (146.3) of the reected ray is


2 v0 kH
2 1 0:1  10
ti ln
ln
13:9 s

k
v0
0:1
1
and the maximum distance for the ray in the layer (146.4) is
xmax

r
p
21
2  10 1
20 3 34:6 km
0:1  10

The travel-time curve for rays inside the layer is calculated using the expression

2
1 kx
t sinh
20 sinh1 0:05x
k
2v0
and is represented in Fig. 146b.

273

Ray theory. Constant and variable velocity

Fig. 147

147. A medium has a distribution of velocity with depth of the form v v0 eaz, with
0 < a <1. Write as functions of the epicentral distance x the expressions for the ray
parameter, travel-time, and maximum depth reached.
If r is the maximum depth reached for a ray with ray parameter p (Fig. 147), the epicentral
distance x is given by
r
r
pvdz
pv0 eaz dz
x p p
1  p2 v20 e2az
1  p2 v 2
0
0

r
2
2
sin1 pv0 eaz sin1 pv0 ear  sin1 pv0 
a
a
0
1
, we have
and, as p
v0 ear
i
2 hp
 sin1 pv0
x
a 2
From this expression we obtain
1
ax
p cos
v0
2
The travel-time is given by

x
1 x
ax
2
ax
cos dx
sin
t pdx
v0 0
2
v0 a
2
0
To nd the maximum depth of penetration r of a ray arriving at distance x, we write
1
1
ax
1  ax
p
)
r


ln cos

cos
v0 ear v0
2
a
2
z
148. In a semi-innite medium of speed of propagation v 6 exp
, the P-wave
2
emerges with an angle of incidence of 30 . Calculate the difference in arrival times at
a given station of the P-wave and the PP-wave (the wave reected once at the free
surface). At what angle of incidence does the PP-wave emerge?
For a velocity distribution increasing with depth of the type v v0ea z (in our case with
v0 6, a 1/2) rays follow a curved path. For a focus on the free surface the ray parameter
p and the travel times t are given by (Problem 147)
ax
1
p cos
v0
2
148:1
ax
2
sin
t
av0
2

274

Seismology

Fig. 148

For a ray with incidence angle at the surface i0 30 , the ray parameter of the direct
P-wave is given by
p

sin i0
1
1

v0
2  6 12

Substituting this value in (148.1) we obtain, for the epicentral distance x,


1
1  x 
4p
)x
cos
km
12 6
22
3
The corresponding travel time is

tP

2
1 4p 1
sin
0:58 s
1
2 3 2
6
2

The travel time of the reected PP-wave (Fig. 148) is double that of the direct P-wave
arriving at the distance x/2:

2
1 4p 1
t PP 2 sin
0:67 s:
1
2 6 2
6
2
The difference between the two times is
t PP  t P 0:67  0:58 0:09 s:
To calculate the incidence angle of the PP-wave, we determine rst the ray parameter
p corresponding to the distance x/2:
x
4p
4p

2 32
6
so

p
1
1 4p 1
3

p cos
12
6
2 6 2

275

Ray theory. Constant and variable velocity

From the value of p, using Snells law, we obtain i0:


p
3
sin i0
p
6 ) i0 60
) sin i0
12
v0
149. A layer of thickness H has a velocity distribution v v0 exp(az) where a <1.
Beneath it there is a semi-innite medium of speed of propagation v1 2v0 exp(aH).
Determine in terms of v0, v1, a, and H:
(a)
(b)
(c)
(d)

The distance and critical angle.


The time of intersection of the reected wave.
The maximum distance of the direct wave.
Calculate the values of these parameters if v 1 km s1, H 10 km, and
a 0.1 km1.
(a) The distance for a ray reaching a depth H is given by (Fig. 149)
H
x 2 tan idz

149:1

and using the denition of the ray parameter p,


sin i
) sin i vp
p
pv
cos i 1  v2 p2
vp
tan i p
1  v 2 p2
Substituting in (149.1) we obtain
H
H
vp
pv0 eaz
x 2 p dz 2 p
dz
1  p2 v20 e2az
1  v 2 p2
0
0

149:2

xmax

xc
S

ic

n1

Fig. 149

276

Seismology

Introducing the change of variable


u v0 eaz
du v0 eaz adz
we nd
x


2 1
sin pv0 eaH  sin1 pv0
a

For the critical angle ic at the bottom of the layer we have


p

sin i0 sin ic
1
vH
v0 eaH

) ic sin1
sin1
v0
vH
v1
v1
v1

By substitution in (149.2) we nd, for the critical distance xc,








2
v0 eaH
v0
2
vH
v0
2
xc sin1
 sin1
sin1
 sin1
ic  i0
a
v1
v1
a
v1
v1
a
(b) The intercept time of the reected ray corresponding to the vertical ray (x 0 and p 0) is
H
H

dz
1 az
2
ti 2
2
e dz
1  eaH
av0
0 v
0 v0

(c) The maximum distance of a ray contained in the layer is given by (149.2)
H

vp
2
149:3
xmax 2 p dz sin1 pv0 eaH  sin1 pv0
a
1  p2 v 2
0

At the point of greatest depth penetration the incidence angle is 90 and, according to
Snells law,
p

sin 90 eaH


1

p
vH
v0
vH

By substitution in (149.3),
xmax

i 2
2 hp
 sin1 eaH cos1 eaH
a 2
a

(d) Substituting the data of the problem we obtain

1  e0:110
30
5:62

2
ti
1  e0:110 12:6 s
0:1  1
2
cos1 e0:110 23:9 km
xmax
0:1


2
1 1
1 1
sin
 sin
6:8 km
xc
0:1
2
5:62
ic sin1

277

Ray theory. Spherical media

Ray theory. Spherical media


150. Assume that the Earth consists of two concentric regions of constant velocity: the
core of radius R/2 and the mantle. The speed of propagation in the core is twice that of
the mantle. Calculate:
(a) The maximum angular distance of the direct ray in the mantle.
(b) The critical angular distance of the refracted ray in the core.
(c) Plot the paths of the waves that propagate through the Earths interior, and the
travel-time curves of these waves in units of R/v, where v is the speed of propagation in the mantle.
(a) The travel time of the direct ray in the mantle in terms of the angular distance is
given by
R 
t1 2 sin
v
2
where 0 D Dmax and Dmax is the maximum distance for a ray contained in the mantle.
According to Fig. 150a the last ray which propagates in the mantle without entering the
core corresponds to angular distance Dmax which in our case is
R
max 2
cos
) max 120
R
2
(b) The critical angle for a ray incident at the core is
sin ic
1

) ic 30
v
2v
To calculate the critical distance Dc we consider the relation (Fig. 150b)

R
n
2n
R/2

Fig. 150a

/2

max
R

278

Seismology

ic
P
n

2n

q
R/2

Fig. 150b

/2
R/2

Fig. 150c

y b a 180
sin a sin b

R
R
2
ic b 180 ) b 150 ) a 14:5 ) y 15:5
c 2y 31
(c) The travel time of a ray reected at the mantlecore boundary is (Fig. 150c)
t2

FP PS
FP

2
v
v
v

150:1

279

Ray theory. Spherical media

S
i1
Q
P

i1

i2

i2
2n

R/2

n
R

Fig. 150d

According to the cosine law,


2

FP
Substituting in (150.1):


2
R
R

R2  2 R cos
2
2
2

r
R

t2
5  4 cos
v
2
The travel times for the minimum and maximum angular distances are
R
min 0 ) t
v
p R
max 120 ) t 3
v
The travel time for a ray which enters the core, that is, with i1 < ic, can be determined
according to Fig. 150d using Snells law:
sin i1 sin i2

v
2v
sin i2 2 sin i1

150:2

Adding the times of the paths through the mantle and the core:
FP PQ QS 2FP PQ

v
2v
v
v
2v

2
R
R

a
2
FP R2  2R cos
2
2
4
R a
PQ 2 sin
2
2

t3

Because a 180 2i2, we obtain


s

2R 5

R
 sin
i2 cos i2
t3
v
4
2
2v

150:3

for values of the incidence angle 0 < i < ic, corresponding to distances Dc < D < 180 .
The relation between the incidence angle i1 and angular distance D is given by

280

Seismology

i 1 ( )

i 2 ( )

D ( )

t3 (R/v)

0
10
20
30

0
20.3
43.2
90

180.0
169.9
153.4
89.0

1.50
1.53
1.60
1.46



sin i1 sin 90 i1  i2  2

R
R=2


sin i1
sin i1 2 cos i1  i2 
)  2 i1  i2  cos1
2
2

150:4

Using Equations (150.2), (150.3), and (150.4) we can calculate the travel times of the direct
ray in the mantle, the reected ray, and the transmitted ray through the core. Some values
for the transmitted rays in the core are given in the table.
The travel-time curves for direct rays (1), reected rays (2), and rays refracted in the core
(3) are shown in Fig. 150e.
2.0

3
1.5

t (R/v)

1.0

1
0.5

0.0
0

50

100

Fig. 150e

150

281

Ray theory. Spherical media

151. Assume that the Earth consists of two concentric regions of constant velocity: the
core of radius R/2 and the mantle. The speed of propagation in the core is half that
of the mantle. Plot the travel-time curves of the waves that propagate in the interior of
the Earth in units of R/v where v is the speed of propagation in the mantle.
This problem is similar to Problem 150, but now the velocity of the core is less than that of
the mantle. In the mantle we have direct and reected rays. As in Problem 150 the
maximum angular distance for the direct wave is 120 . The travel times for the direct
(t1) and reected (t2) rays are
R 
t1 2 sin
v
2
r
R

5  4 cos
t2
v
2

151:1
151:2

Since the velocity of the core is less than that of the mantle there is no critical angle. All
rays incident at the core are refracted into it. According to Snells law the refracted angle i2
is less than the incident angle i1 (Fig. 151a):
sin i1 sin i2
v
v
2
1
sin i2 sin i1
2

i1

i2
i2

R/2
n

n/2

Q
i1

Fig. 151a

282

Seismology

The travel-time for a ray crossing the mantle and the core is (Fig. 151a)
t3

FP PQ QS 2FP 2PQ
v

v
v
v
v
2

where

R2
R
R2

 2R cos a R2  R2 cos 90 i2 
4
2
4
2

2
5R

2
R2 sin i2 
FP
4
2
R
PQ 2 cos i2 R cos i2
2
2

FP R2

so
2R
t3
v

5

2R
sin i2 
cos i2

4
2
v

151:3

3.5
3

3.0

2.5

t (R / n )

2.0

1.5
2
1.0

0.5
1

0
0

25

50

75

100
()

Fig. 151b

125

150

175

283

Ray theory. Spherical media

The relation between the incidence angle at the mantlecore boundary, i1, and the angular
distance, D, of a ray which crosses the core is




90
sin
i

i

1
2
sin180  i1
2

R=2
R


 sin i1
1
sin i1 2 cos i1  i2
 i1 sin1 sin i1
)  2 cos1
2
2
2

151:4

The range of distance for this ray is 120 < D <180 .


From Equations (151.1), (151.2), and (151.3) we can calculate the values for the traveltimes of the direct, reected, and refracted rays. Some values for t3 are given in the table.
i 1 ( )

i2 (  )

D ( )

t3 (R/v)

0
10
20
30

0
20.3
43.2
90

180.0
149.4
114.0
31.0

1.50
1.48
1.40
1.07

The travel-time curves for rays that are direct (1), reected (2), and refracted in the core (3)
are shown in Fig. 151b.
152. Consider a spherical Earth of radius R formed by two hemispherical media of
constant velocities of propagation v and 2v. For a focus on the surface of the
hemisphere of velocity v at the point of intersection of the diameter perpendicular
to the plane that separates the two media, calculate the travel times and travel-time
curves of the direct, reected, and critical refracted waves at the surface of separation
of the two media, in units of R/v. Calculate the expression for the travel time of waves
that propagate through the medium of speed of propagation 2v.
The travel time for angular distances D 90 are given by (Fig. 152a)
t1

FS
R 
2 sin
v
v
2

152:1

The travel time of the ray reected at the plane boundary between the two hemispheres is
(Fig. 152a)
t2

FP PS F0 S

v
v
v

According to Fig. 152a (triangle OSF0 ) the relation between the angles a and D is
180   2a 180 ) a
0


2

SS0
sin a
FS
R sin 

SS0 R sin  ) F0 S
2R cos
sin a
2
sin a

SS
0

) F0 S

284

Seismology

2
R

Fig. 152a

Then for D < 90


t2

2R

cos
v
2

152:2

The critical angle, according to Snells law, is given by


sin ic
1
) ic 30
v
2v
The travel-time of the critically refracted ray is (Fig. 152b)
t3

FA AB BS

v
2v
v

152:3

According to Fig. 152b


FA

R
cos ic

BS

SS0
R cos 

cos ic
cos ic
0

AB OS  OA  BS
0

OS R sin 
OA R tan ic
0

BS BS sin ic

285

Ray theory. Spherical media

ic
ic

ic

ic
P
O

2n

Fig. 152b

Substituting ic 30 :
2R
FA p ;
3

BS

2R cos 
p ;
3

R
R cos 
AB R sin   p  p
3
3

and substituting in (152.3) we obtain, for Dc D 90 ,


i
p
Rh
sin  31 cos 
t3
2v

152:4

The critical distance can be calculated from ic 30 using Fig. 152a (triangle OSP) and a D/2
90  c a ic 90 180 ) c ic a ) c 60
The travel time of the rays that cross the boundary and penetrate in to the medium of
velocity 2v is given by (Fig. 152c)
t4

FP PS

v
2v

According to Snells law,


sin i sin i0

) sin i0 2 sin i
v
2v
and we have that
R
cos i
SP0
PS
sin i0
0
SP R sin180    R tan i
FP

152:5

286

Seismology

2n
i
R

Fig. 152c

2.0
2

t (R / n )

1.5

1.0
1

0.5

0
0

Fig. 152d

20

40
()

60

80

287

Ray theory. Spherical media


Substituting in Equation (152.5) we obtain, for D > 90 ,
t4

3R
R sin 

4v cos i 4v sin i

152:6

Travel-time curves for t1, t2, and t3 are shown in Fig. 152d.
153. Assume an Earth formed by a mantle of thickness R and a core of radius R/2,
with velocities v and 2v. There is a seismic focus at depth R/4 below the surface.
Calculate the travel-time curves of the direct and reected waves.
The travel time of the direct ray is (Fig. 153a)
t1

FP
v

The distance FP can be expressed in terms of R and D using the cosine law in triangle FOP:

3R 2
3R
FP
R2  2R cos 
4
4
Then, we obtain
R
t1
v

r
25 3
 cos 
16 2

153:1

The maximum distance for the direct ray is


max 1 2
R
cos 1 2 ) 1 48:2
3R
4
R
cos 2 2 ) 2 60:0
R
max 48:2 60:0 108:2
S

F
R/4

Fig. 153a

R/2

MAX
O

2n
R

288

Seismology

i
r

b
P

r
F
R /4

2
a
i

2n

1
R/2

Fig. 153b

The travel time for the reected ray (Fig. 153b) is


t2

FP PS

v
v

The distances FP and PS are expressed in terms of R, D1, and D2 using the cosine law in
triangles FOP and SOP (Fig. 153b):
s

2
2
3
R
R 3R
R
cos 1
2
FP
4
2
2 4

s

2
R
R
PS R2
 2 cos 2
2
2
Then, we obtain
r
r
13 3
5
 cos 1 R
 cos 2
R
16 4
4
t2

v
v

153:2

Now we need to express D1 and D2 in terms of the take-off angle i at the focus (F). Using
Snells law for a spherical medium, we relate i and i0 , the incidence angle at the station (S):
3R
R
sin i
sin r R sin i0
3
4
2

) sin i0 sin i
v
v
v
4

153:3

According to Fig. 153b for triangle FOP we have


i a 1 180
sin a sin i

R
3R
2
4
and we obtain
2 sin1 i 3 sin i

153:4

289

Ray theory. Spherical media

1.4

0.8

t (R / n )

0.8

0.8
1
0.6

0.4

40

20

60

80

100

()

Fig. 153c

and for triangle POS


b 180  2  i0
sin i0 sin b

R
R
2
and
2 sin i0 sin2 i0

153:5

Equations (153.3), (153.4), and (153.5) allow us to calculate D1 and D2 from the take-off
angle i at the focus. The travel-times are given in the following table.
i ( )

i 0 ( )

D1 ( )

D2 ( )

D ( )

t2 (R/v)

0
10
30
40
41.8

0
7.5
22.0
28.8
30.0

0
5.1
18.6
34.6
47.1

0
7.6
26.6
45.8
58.9

0
12.7
45.2
80.4
106.0

0.75
0.76
0.92
1.19
1.41

For angular distance D greater than 108.2 there are no reected rays. The travel-time
curves corresponding to the direct (1) and reected (2) rays are shown in Fig. 153c.

290

Seismology

154. Consider a spherical Earth of radius R 3000 km and constant P-wave speed of
propagation of 4 km s1. Within it there is a core of radius R/2 and constant velocity
v1. At a station at epicentral distance D from an earthquake with focus at the surface,
the observed time interval is tS-P 547.0 s. Given that Poisson's ratio is 1/6, and that
the arrival of the P-wave is at 12 h 23 m 20.4 s, calculate:
(a) The epicentral distance.
(b) The time of the earthquake.
(a) For a spherical Earth of constant velocity the travel time of the direct ray is given by
t2

FO0
R 
2 sin
v
v
2

154:1

Taking into account the presence of the core the maximum distance for the direct ray is (Fig. 154)
R
max 2
cos
) max 120
R
2
Since Poissons ratio is 1/6 we have
s

1
l

) m 2l
6 2l m

and we can calculate the relation between the velocities of the P-wave (a) and the S-wave (b),
r 9
l 2m >
r
>
a
=
5

rr
)a
b ) b 2:53 km s1
m >
2
>
b
;
r

Using (154.1) and assuming the same path for P- and S-waves, from the time interval S-P
we obtain the distance:

 1 1
S-P

)  77:7
t
2R sin
2 b a
O

max
n1
F

Fig. 154

R/2

291

Ray theory. Spherical media

Notice that D < Dmax.


(b) To calculate the time of origin we subtract from the arrival time of the P-wave the
value of the travel time for that distance:
tP

2R 
3000 77:7
sin 2
sin
940:9 s
a
2
4
2

The time of origin is then given by


t0 12 h 23 m 20:4 s  940:9 s 12 h 07 m 39:5 s
155. Consider the Earth formed by a sphere of radius R 4000 km, Poisson's ratio
1/8, and constant S-wave speed of propagation 3 km s1. Within it, there is a liquid
core of radius R/2. There occurs an earthquake with a focus in the interior of the
Earth. At a station of epicentral distance D the observed time interval is tS-P 600 s.
This focus may be at a depth of either R/10 or 2R/5. Calculate the correct depth of the
focus, and the epicentral distance.
First we determine the maximum distance for direct rays which dont penetrate into the
core, which according to Fig. 155 corresponds to Dmax D1 D2:
R
cos 1 2
Rh
R
cos 2 2 ) 2 60
R
If the depth of the focus is R/10 then
h

R
5
) cos 1 ) 1 56 ) max 116
10
9

and if it is 2R/5, then


2
5
h R ) cos 1 ) 1 33:56 ) max 94
5
6
S

F
h

Fig. 155

2 1

R/2

292

Seismology

For a point on the surface at distance D, the S-P time interval implies, assuming the same
path for P- and S-waves,
t S-P

FP FP FP


a  b
b
a
ab

155:1

From the value of Poissons ratio the relation between the P and S velocities is given by
s s r
1
l
m
l 2m
7m
7
)l )a

b
s
8 2 l m
3
r
3r
3
Substituting in (155.1) the S-P interval equal to 600 s we obtain the length of the ray:
FP 600

ab
5212 km
a  b

Using the cosine law for triangle FOP


2

FP R  h2 R2  2RR  h cos 
2

FP  R  h2  R2
cos 
2RR  h
2
If h R then D 106 , but this result is not possible because the maximum distance of
5
R
the direct ray for that depth is 94 . If h
then D 86 , this result is possible because
10
this distance is less than the maximum distance. The depth is, then, 400 km.
156. Consider the Earth of radius R and constant velocity v with a core of radius 6R/10
and constant speed of propagation 2v. An earthquake occurs with focus at 8R/10 from
the centre of the Earth. A wave emerges from that focus with a take-off angle of 15 .
(a) Will it pass through the core?
(b) What epicentral distance will it reach?
(c) What will be the travel time of the wave (in units of R/v)?
(a) First we calculate the maximum epicentral distance for a ray which doesnt
penetrate the core. According to Fig. 156a the maximum distance is
max 1 2
6
R
10
) 1 41:4
cos 1
8
R
10
6
R
10
cos 2
) 2 53:1
R
max 41:4 53:1 94:5
From this value we calculate the take-off angle ih for this ray:
1 ih 90 ) ih 48:6

293

Ray theory. Spherical media

ih

6R/10

2n

8R/10

Fig. 156a

S
i
B
i0
i2

A
i
F

ih

i2

6R/10

2
O

3
2n

8R/10

Fig. 156b

For take-off angles less than 48.6 the rays travel through the core.
Since the velocity in the core is greater than in the mantle, to nd out which rays
penetrate into the core, we also need to know the critical angle. Rays with incidence angle
at the coremantle boundary with i > ic are totally reected and dont penetrate into the
core. According to Snells law the critical angle is given by
sin ic
1
) ic 30:0
v
2v
We calculate, using Snells law, the angle of incidence i corresponding to the take-off angle
of 15 (Fig. 156b):
8
6
R sin ih
R sin i
10
10

) i 20:2
v
v
Since the incidence angle i (20.2 ) is less than the critical angle (30 ) and less than the
angle corresponding to the maximum distance (48.6 ), the ray with take-off angle of 15
penetrates into the core.

294

Seismology

(b) Applying Snells law we nd the angle of the transmitted ray in the core i2 (Fig. 156b):
8
6
R sin ih
R sin i2
10
10
v
2v

i2 43:7

By consideration of triangles FOA and AOB, we determine D1 and D2 (Fig. 156b):


ih 1 180  i 180 ) 1 5:2
i2 2 i2 180 ) 2 92:6
and using Snells law we determine i0 the incidence angle at the surface and D3:
6
6
R sin i2
R sin i R sin i
o
10
10

2v
v
v
3 i  io 8:3

) io 11:9

The epicentral distance of the ray is


 1 2 3 106
(c) The travel time is
t

FA AB BS

v
2v
v

where

s

2
2
8R
6R
8 6
FA

 2 R R cos 1 0:21 R
10
10
10 10
s

2
6
6
6 6 2
AB
R
R 2
R cos 2 0:87 R
10
10
10 10
s

2
6
6
R R2  2 RR cos 3 0:42 R
BS
10
10
so
t 1:07

R
v

157. Assume a spherical Earth of radius R 6000 km and constant S-wave speed of
propagation 4.17 km s1. Poisson's ratio is 1/4. At a station at epicentral distance 60
an earthquake is recorded with a time interval t S-P 554 s. Calculate the depth of the
earthquake.
Given that Poissons ratio is 0.25, the P-wave velocity is
s

p
1
l

) l m ) a 3b 7:22 km s1
4 2l m

295

Ray theory. Spherical media

R
Rh

Fig. 157

From the time interval S-P we can calculate the length of the ray FS (Fig. 157):
FS FS
ab

FS
b
a
ab
S-P
t
ab
FS
5469 km
ab

t S-P

157:1

The distance along the ray in terms of the angular epicentral distance D, using the cosine
law, is
q
FS R2 R  h2  2RR  h cos 
Substituting FS from (157.1):

5370

q
R2 R  h2  2RR  h cos 

157:2

We substitute in (157.2) the values R 6000 km and D 60 and solve for h, nding two
possible solutions:
h1 4706 km
h2 1294 km
158. Assume a spherical Earth of radius R and P-wave velocity which can be
expressed by the equation v(r) a  br2. The speed of propagation at the surface
of the Earth is v0 and at the centre of the Earth it is 2v0. What angular distance D does
a wave reach which penetrates to a depth equal to half the Earths radius?
If the velocity distribution inside the Earth is v(r) a  br2, the ray paths are circular with
radius given by (Fig. 158)

296

Seismology

7/4 R
7/4 R

/2 /2

R/2

Fig. 158

r
dv
p
dr

158:1

From the conditions of the problem


r R ) v v0 a  bR2
r 0 ) v 2v0 a
and

v0 )
r2
2
R
) v v0 2  2
R
a 2v0
b

The radius of curvature of the ray which penetrates to r R/2 is that corresponding to the
ray parameter
r0
R
p0 0 0
158:2
2v
v
The velocity at depth R/2 is
0
0 12 1
R
B
7
B
0
2C C
C
v v0 B
@2  @RA A v0 4

Substituting the ray parameter in (158.2):


p0

R
2R

7 7v0
v0
2

297

Ray theory. Spherical media

The derivative of the velocity is



dv
d
v0

a  br2 2br 2 2 r
dr dr
R

Substituting in (158.1) we obtain, for the radius of curvature (Fig. 158),


r

r
7R

p2br
4

The epicentral (angular) distance corresponding to this ray is found by applying the cosine
law to the triangle POS:

2
2
7
9
9R

R
R R2  2 R cos )  96:4
4
4
4
2
p
159. The Earth consists of a mantle of radius R and speed of propagation v a= r,
and a core of radius R/2 and speed of propagation 4v0, where v0 is the speed of
propagation at the Earths surface. Calculate:

(a) The maximum epicentral distance corresponding to a wave that travels only
through the mantle.
(b) The critical angle of the wave reected at the core, and the angle at which it leaves
the surface.
(c) The epicentral distance Dc corresponding to the critical angle.
(d) Plot the travel-time curve, specifying Dc and Dm.
(a) The value of a in the velocity distribution is found from the value of velocity at the surface:
1
1
r R ! v v aR 2 ) a v R 2
0

The velocity distribution is



a

12
R
R
v v0
v0
r
r
For this general type of distribution of velocity with depth a < 1, the angular distance for a
surface focus (Fig. 159a) is given by

2
1 p

159:1
cos
1a
0
where
r
R
 ) 0
v
v0
The maximum distance for a ray which travels only through the mantle, that is that reaches
depth R/2, can be calculated from the velocity at that depth, vm:
0 11
R
R
2
p
1 R
BRC
vm v0 @ A v0 2 ) p 2 2p p p
R
vm v0 2 2 2 v0
2

298

Seismology

c
F

i0

ic

/2

/2

R/2

Fig. 159a

By substitution in (159.1) we obtain, for the maximum distance,


1
1 R
p
B2 2 v0 C
2

C
cos1 B
m
@ R A 92:4
1
1
v0
2
0

(b) The critical angle of a reected ray at the mantle-core boundary applying Snells law is
sin ic
1
p
) ic 20:7
v0 2 4v0
The take-off angle at the surface i0 for this ray is found by again applying Snells law:
R

R sin i0 2 sin20:7
p

) i0 7:2
v0
v0 2

(c) To nd the critical distance we use the expression


r0
p
dr
p
c 2
2
r
  p2
rp
where
R
2
r0 R
r0 sin i0
R

p
v0
8v0
r
r
r3=2
  1 p
v v R 2 v0 R
0
rp

Substituting the values of the problem in (159.2):


c

1
4

dr
r
R=2 r
R3
3
r
64
R3=2

159:2

299

Ray theory. Spherical media

This integral is of the type

so we can write the solution


c

dx
2
p p cos1
n
n
x x a
n an

r
an
xn





8
1
1
18
cos1
 cos1 p
6
8
8

(d) The travel time of the rays in the mantle is given by




20

R 4 3
sin 1 a
sin
for 0  92:4

t
1a
2
v0 3
4
By substitution of values of D we nd the travel time curve given in Fig. 159b.
1.2

1.0

t (R/n)

0.8

0.6

0.4

0.2

0
0

c
20

m
40

60

80

()

Fig. 159b

160. A spherical medium of radius R has a constant speed of propagation v0 from the
surface down
R/2, and from R/2 to the centre a core of variable speed of propaga to

1=2
R
tion v v0
.
r
(a) What value should i0 have for the waves to penetrate into the core?
(b) Calculate the epicentral distance reached by a wave leaving a focus at the surface
at angle i0.
(a) The velocity at the top of the core (r R/2) is

300

Seismology

S
F

l
i0

n0
R/2

Fig. 160

0 11
2

p
BRC
v1 v0 @ A v0 2
R
2

Applying Snells law we nd the critical angle ic for incident rays at the core (Fig. 160):
R
R sin ic
p2
v0
2v 0

) ic 45

The take-off angle i0, for a focus at the surface corresponding to the critical angle, using
Snells law, is
R
R sin i0 2 sin ic

v0
v0

) i0 20:7

The rays that penetrate into the core must leave the focus with take-off angles less than
20.7 .
(b) For a ray with take-off angle i0 which penetrates the core the epicentral distance is the
sum of that corresponding to the part that has travelled through the mantle, D1, plus the
part that has travelled through the core, D2:
 21 2
Since in the core the velocity varies with depth with the law given in the problem, the
epicentral distance is given by

2
1 p
cos
2
1a
1
where a is the exponent of the velocity distribution
v v0 ra ) a

1
2

301

Ray theory. Spherical media

and p is the ray parameter, which can be obtained using Snells law:

and

R
R sin i0 2 sin i1
p
p
v0
2v 0

Then, we nd

R
r
 ) 1 p2
v
2v 0
0

1
R sin i0

B
C 4 1  p
2
1 B v0 C
2
cos
2

cos
2
sin
i
0
@ R A 3
1 12
p
2 2v 0

The distance D1 can be determined using the sine and cosine laws for the triangle FOP
(Fig. 160):
R

2 l )  sin1 2l sin i0
1
R
sin i0 sin 1
l 2 R2  2Rl cos i0

R2
1 p
) l R cos i0 
4R2 cos2 i0  3R2
4
2

and we nd the expression in terms of i0:



 p


1 p
4
 2sin1 2 cos i0 
4cos2 i0  3 sin i0 cos1 2 2 sin i0
2
3

161. Consider a spherical Earth of radius 6000 km and surface velocity of 6 km s1, with
p
a velocity distribution of the type vr a= r. At the distance reached by a wave
emerging at a take-off angle of 45 from a focus on the surface, calculate the interval
between the arrival times of the direct P- and reected PP-waves (the PP-wave is one that
is reected at the surface at the midpoint between
focus and the point of observation).

the
a
R
The velocity distribution is of the type v v0
where
r
p
a
r R ) v v0 p ) a v0 R
R

12
R
v v0
r
and the ray parameter p for a ray with take-off angle of 45 is
1
6000  p
R sin i0
2
p
707 s

6
v0

302

Seismology

For this type of velocity distribution the relation between the ray parameter and the
epicentral distance is



1a
p 0 cos

161:1
2
In this problem the value of 0 is
r
R
 ) 0 1000 s
v
v0
Substituting the values in (161.1) we obtain the distance for the ray with take-off angle
of 45 :

3
707 1000 cos  )  60
4
The PP-wave which arrives at D 60 travels twice the distance which a P-wave does for a
distance of 30 . For this type of velocity distribution the travel time for a distance D is
given by


20
1 a
sin
t
1a
2
In our case for the P- and PP-waves at distance 60 we substitute the values of the problem
and nd
tP 60 943 s
tPP 60 2tP 30 1021 s
The time interval between the PP- and P-waves 60 is
tPP  tP 1021  943 78s
162. In an elastic spherical medium of radius r0, the velocity increases with depth
according to v arb. If v0 6 km s1, r0 6000 km, and, at a point at distance
D 90 , the slope of the travel-time curve is 500 s, determine:
(a) The value of b.
(b) The value of rp and of vp of the wave reaching an epicentral distance of 90 .
(a) For this type of velocity distribution the travel time in terms of the epicentral
distance is given by


20

t
sin 1 b
162:1
1b
2
As we know the velocity at the surface,
r
r0
 ) 0 1000 s
v
v0

303

Ray theory. Spherical media

The ray parameter p is known, because it is equal to the slope of the travel-time curve
which for D 90 is given as 500 s. Using the relation between p and D for this type of
velocity distribution,


dt

0 cos 1 b
p
d
2
h
h
pi
pi 1
p p
1
500 1000 cos 1 b ) cos 1 b ) 1 b ) b
4
4
2
4 3
3
(b) At the point of greatest penetration rp for D 90 , we have the relation
p

rp sin 90 rp
) rp pvp
vp
vp

and also

1
r0 3
vp v0
rp
From these two equations we obtain rp and vp:
1

3 1 3

rp v0 pr03 rp 3 ) rp v40 r04 p4 3564 km


rp 3564
vp
7:1 km s1
p
500
163. Consider a spherical Earth of radius R, the northern hemisphere with a constant
speed of propagation v0, and the southern hemisphere with a speed of propagation of

1
R 2
.
v v0
r
(a) Calculate the travel time of seismic waves for a focus on the equator and stations
on the same meridian.
(b) In which hemisphere does the wave at a distance of 60 arrive rst?
(a) In the northern hemisphere the velocity is constant and the rays have straight paths
and their travel time is (Fig. 163)
2R 
sin
0 <  < 90
163:1
v0
2
In the southern hemisphere the velocity increases with depth and the rays have curved
paths. Their travel time is given by
tN

tS

20

sin1 b
2
1b

where

b
R
R
1
0
and v v0
)b
v0
r
2
Substituting in (163.2) we obtain for the travel time in the southern hemisphere

163:2

304

Seismology

S1

n0

S2

Fig. 163

tS

4R
3
sin
3 v0
4

0 <  < 90

163:3

(b) The travel times for waves in the northern and southern hemisphere are given by
Equations (163.1) and (163.3). By substitution of D 60 we obtain
R
v0
p
R2 2
R
S
0:47
t
v0 6
v0
The waves arrive rst in the southern hemisphere.
tN

164. Consider a spherical medium of radius R consisting of two concentric regions


(mantle and core), the core of radius R/2. The speeds of propagation are v ar1=2
for the mantle and v aR1=6 r1=3 for the core. The surface velocity is v0. For a wave
leaving a focus with angle of incidence 14.5 , calculate the angular distance D at
which it reaches the surface.
We calculate a by applying the boundary conditions
1

r R ) v v0 ) v0 aR 2 ) a v0 R2

1
R 2
Mantle : v v0
r

1
R 3
Core : v v0
r
We determine the ray parameter corresponding to the ray with take-off angle i0 14.5 ,
using Snells law (Fig. 164a):
p

r sin i R sin 14:5


R

v
v0
4v0

305

Ray theory. Spherical media

S
F

i0

i0

R/2

i2

Fig. 164a

At the bottom of the mantle at depth R/2 the velocity is


0 11
2

1
BRC
v1 v0 @ A v0 22
R

The incident angle i of this ray on the mantlecore boundary, applying Snells law, is given by
R
sin i R sin i
0
2
) i 45

1
v
0
v 22
0

On the top of the core the velocity is


0 11
3

1
BRC
v2 v0 @ A v0 23
R

which is less than at the bottom of the mantle and there is no critical angle. Applying
Snells law again we obtain the angle i2 of the refracted ray in the core:
sin i sin i2

) i2 39
v1
v2
The take-off angle of the last ray which travels only in the mantle is given by
R
R sin i0
2 1 ) i0 20:7
v0
v 0 22
In our case the angle 14.5 is less and the ray penetrates into the core.
The epicentral distance is the sum of the distances corresponding to the paths in the
mantle and in the core:
 21 2

306

Seismology

S
F

Fig. 164b

The distance corresponding to the path in the core is given by


2
p
2
cos1
1b
0
where
p

R
4v0

R
0 2 1
v0 23
and where p is the ray parameter and  r/v. Substituting the values we obtain
2 76:4
To calculate D1 we suppose that there is no core and a ray with take-off angle i0 14.5
would arrive at distance D3 which is related with D1 by (Fig. 164b)
21 3  4
The distances D3 and D4 can be determined using the equation

2
1 p
cos

1b
0
where for D3
0 R


2

R
v0
R
3
R 2
2
2

0 11 v 0
2
BR C
v0 @ A
R
2

307

Surface waves

and we obtain
0

1
R
B4v0 C
2

C
3
cos1 B
@ R A 100:6
1
1
2

and by similar substitutions for D4

1
R
C
B
B
C
2
1 B 4v0 C

cos B
4
C 60:0
1
3
@
A
1
R 
2
2 2
v0

Then, 2D1 100.6 60 40.6 and the epicentral distance is


 40:6 76:4 117:0

Surface waves
165. A Rayleigh wave in a semi-innite medium has a 20 s period. If the P-wave
velocity is 6 km s1 and Poissons ratio is 0.25, calculate the depth at which u1 0,
and at which depth the particle movement becomes prograde.
Since Poissons ratio is 0.25, we nd the relation between P- and S-waves:
s s
1
l
l 2m
3m p
s
)lm)a

3b
4 2 l m
r
r
6
a 6 ) b p 3:4 km s1
3

For a half-space the velocity of Rayleigh waves is


cR 0:919b 3:2 km s1
The displacement u1 is given by
u1

@ @c

@x1 @x3

where the potentials are given by


A expikrx3 ik x1  cR t
c B expiksx3 ik x1  cR t

1=2
c2
r i 1  R2
0:85i
a

1=2
c2
0:39i
s i 1  R2
b

308

Seismology

Then,
u1 0 ) ikA exp0:85kx3  0:39ikB exp0:39kx3 0

165:1

so
k

2p
2p

0:1 km1
l
TcR

We can write B in terms of A using the boundary condition of zero stress at the free surface:


t31 0jx3 0 ) 2rA  1  s2 B 0
so

B 1:47iA
Substituting in (165.1) we obtain the value of x3:
exp0:85kx3 0:39  1:47 exp0:39kx3
x3 12 km
At 12 km depth u1 is null and for greater values of depth the particle motion is prograde
while for lesser values of depth it is retrograde.
166. Given a layer of thickness H and shear modulus m 0 on top of a half-space or
semi-innite medium in which l 0, study (without expanding the determinant)
whether there exist surface waves that propagate in the x1-direction. Are they
dispersive waves?
In the liquid layer ( m 0) the P- and S-velocities are
s s
s
0
m
l0 2m0
l0
b0 0
) a0

r
r
r
and in the solid half-space
s
2m p
2b
l0)a
r

The relation between the stress and strain is

tij lydij 2meij


where eij

1
2

ui; j uj;i .

t0ii l0 e011 e022 e033


t0ij 0
In the half-space: l 0 ) tij 2 meij
If there are surface waves propagating in the x1-direction, their displacements in terms of
the potentials are given by (Fig. 166)
In the layer: m0 0 )

u1 ;1  c;3
u2 u 2
u3 ;3 c;1

309

Surface waves

x3
x1

= 0
0
=0

Fig. 166

The boundary conditions at the free surface are null normal stresses:
 0
t33 0
x3 H )
t031 0
and at the boundary between the liquid layer and the solid half-space continuity of the
normal component of the displacement and stress and zero tangential stresses,
8
u3 u03
>
>
>
< t t0
33
33
x3 0 )
0
>
t

t
>
32 0
> 32
:
0
t31 0

In the liquid layer there is only the P-wave potential . Taking (x1, x3) as the incidence
plane
0 A expikr0 x3 ik x1  ct B expikr0 x3 ik x1  ct
r
c2
0
r
1
a02
where c is the velocity of wave propagation in the x1-direction
In the half-space
c C expiksx3 ik x1  ct
u2 E expiksx3 ik x1  ct
D expikrx3 ik x1  ct
r
c2
1
r
a2
s
c2
s
1
b2
In the layer we have only guided P-waves and r0 is real, while in the half-space for surface
waves, r and s must be imaginary. Then a > b > c > a0 must be satised.

310

Seismology

From the boundary conditions we obtain the following equations:


x3 H

 0


0
t033 0 ) A 1 r02 eikr H B 1 r02 eikr H 0
x3 0
t31 0 ) 2Dr  C Cs2 0
u03 u3 ) Ar0  Br0 Dr C




t033 t33 ) l0 1 r02 A B 2m Dr2 Cs

For a solution of the system the determinant must



0
0

eikr H
eikr H


0
0

0
0

r
r

l0 1 r02 l0 1 r02

be zero:
0
s2  1
1
2ms


0
2r
0
r
2mr2

Expanding the determinant and working r0, r, and s in terms of the variable c, we obtain c
(k), the velocity of waves in the x1-direction. They have the form of guided waves in the
liquid layer and surface waves in the half-space. Since the velocity c(k) is a function of the
wavenumber the waves are dispersive.
167. There is a liquid layer of density r and speed of propagation a on top of a rigid
medium (half-space). Derive the dispersion equation of waves in the layer by boundary conditions and by constructive interference in terms of v. Plot the dispersion
curve for the different modes.
Given that the layer is liquid the only potential is :
A exp ikrx3 B expikrx3 exp ik x1  ct

167:1

The boundary conditions at the free surface are zero normal stress and at the boundary
between the liquid layer and the rigid half-space zero normal component of displacement
(Fig. 167a):
x3 H ) t33 0
x 3 0 ) u3 0
where
t33 ly r2 ro2 0
u3 ;3
r
c2
1
r
a2
By substitution of (167.1) we obtain
AeikrH BeikrH 0
AB0)AB

311

Surface waves

x3

liquid layer
x1

0
rigid half-space

Fig. 167a
x3
O

B
A

x1

0
P

Fig. 167b

Then, 2A cos krH 0. For waves propagating in the layer, r must be real and c > a. The
solution is given by

1
krH n p;
n 0; 1; 2; . . .
167:2
2
The solution can also be found by the method of constructive interference. The condition
of constructive interference implies that waves coinciding at a given wavefront (AB) are in
phase, that is, the distance along the ray must be an integer multiple of the wavelength,
taking into account possible phase shifts (Fig. 167b). In our case on the free surface,
x3 H, there is a phase shift of p (l/2) and we write the condition as (Fig. 167b)
AP PQ QB 

la
nla
2

or
2p
A P P Q Q B  p 2pn
la

312

Seismology

Substituting
A P P Q Q B 2H cos i
we obtain
2p
2H cos i  p 2pn 1
la

167:3

According to Snells law,


a
sin i ) cos i
c

r
r
a2 a c 2
a
1 2
1 r
2
c
c a
c

a
and ka k.
c
Substituting in (167.3), we obtain the same solution obtained in (167.2):

a
1
ka Hr n p
c
2
This expression can also be written as
"
#12
r

1
oH c2
1
1
1 2 p2
167:4
1 n p)c 2  n
krH n p )
2
c
2
a
2 H 2 o2
a2
The fundamental mode (FM) corresponds to n 0, and n 1 to the higher modes (HM).
In the FM and the higher modes, the frequency oc corresponding to the zero in the
denominator in (167.4) is called the cut-off frequency, as there are no values of c for o <
oc. For a mode of order n the cut-off frequency is given by

1
p n a
2
oc
H
The dispersion curve is shown in Fig. 167c.
c
FM

1 HM

2 HM

2H

Fig. 167c

3
2H

5
2H

313

Surface waves

168. Consider an elastic layer of coefcients l and m, thickness H, and density r on a


rigid semi-innite medium. Derive the dispersion equation c(v) for P-SV and SH-type
channelled waves for the fundamental mode (FM) and the rst higher mode (1HM).
Plot the dispersion curve for the SH motion.
For a SH-wave which propagates in the x1-direction its displacement is given by
u2 E expiksx3 F expiksx3 exp ik x1  ct
The P- and SV-waves are given by their scalar potentials and c:
A expikrx3 B expikrx3 exp ik x1  ct
c C expiksx3 D expiksx3 exp ik x1  ct
where r and s were dened in Problem 166.
The boundary conditions for SH-waves are null stress at the free surface and null
displacement at the boundary with the rigid medium (Fig 168a):
x3 H ) t32 0 m

@u2
@x3

x 3 0 ) u2 0
By substitution we have
EeiksH  FeiksH 0
E F 0 ) F E
Ee

iksH

iksH

0 ) cosksH 0 ) ksh

Substituting s and putting k o/c:


Ho
c

c2
1

1

p)c
2
b2

1
n p
2

! 1

2
1
1 2 p2

n

2 H 2 o2
b2

168:1

This equation give us, for the SH component, the frequency dependence of the velocity c(o).
The boundary conditions for P and SV are similarly

x3

x1

, ,

0
rigid medium

Fig. 168a

314

Seismology

x3 H ) t31 0; t33 0
x3 0 ) u1 0; u3 0
where
t33
t31
u1
u3

le11 e33 2me33


2me31
;1  c;3
;3 c;1

Substituting the expression for the potentials we obtain


l 2mr2 AeikrH r2 BeikrH lAeikrH BeiksH 2msCeiksH  DeiksH 0
2rAeikrH  BeikrH 1  s2 CeiksH DeiksH 0
A B  sC  D 0
rA  B C D 0
For a solution we put the determinant of the system of equations equal to zero:




1
1
s
s




r
r
1
1


ikrH
ikrH
2 iksH
2 iksH 0 168:2

2re
2re
1

s
e
1

s
e


l r2 l 2meikrH l r2 l 2meikrH
2mseiksH
2mseiksH

Expanding the determinant and putting it equal to zero, we obtain the dependence with
frequency of the velocity c(o) which gives us the dispersion curve.
For the wave with SH component the dispersion curve is given in Fig. 168b:
!12

1
1 2 p2
c
 n
2 H 2 o2
b2
c
FM

1 HM

2 HM

2H

Fig. 168b

3
2H

5
2H

315

Surface waves

For n 0 the curves correspond to the fundamental mode and for 1 n to the higher
modes. For all modes, including the fundamental mode, there is a cut-off frequency oc
(n1)pb/2H, with n 0 for the fundamental mode and n 1, for higher-order modes.
169. For a liquid layer of thickness H with a rigid medium above and below, derive
the dispersion equation c(v) of the fundamental and higher modes. For the FM, at
what height above the layer is the motion circular?
Given that the medium is a liquid, motion is represented only by the scalar potential f:
169:1
A exp ikrx3 B expikrx3 exp ik x1  ct
r
c2
 1.
where r
a2
The boundary condition at the two boundaries between the liquid and rigid solid is that
the normal component of the displacement is null (Fig. 169):
x 3 0 ) u3 0
x 3 H ) u3 0
Substituting u3 ;3 we have
AB0

169:2

AeikrH  BeikrH 0
which leads to the equation


A eikrH  eikrH 0

169:3

Consider rst that r is real, that is, c >a. Then, from (169.1)
2iA sin krH 0 ) krH np; n 0; 1; 2; :::
with n 0, fundamental mode (FM), and n 1 for higher modes.
For the FM, n 0 and r 0, and then
r
c2
10)ca
Hk
a2
The displacements from (169.1) and (169.2) are
x3
rigid medium
H

liquid layer
0

x1
rigid medium

Fig. 169

316

Seismology

@
Aikrexp ikrx3  exp ikrx3 exp ikx1  ct
@x3
@
Aik exp ikrx3 exp ikrx3 exp ikx1  ct
u1
@x1
For the FM r 0, then u3 0 and this is a P-wave, with only a u1 component, which
propagates in the x1-direction. For all HM the displacements have both components
For the rst higher mode (1HM), n 1:
r
o c2
H
1p
c a2
1
c2
1
p2

a2 o2 H 2
If
1
p2
ap
)c!1

0
then
o
a2 o 2 H 2
H
u3

The cut-off frequency is oc > ap/H. For each higher mode there is a cut-off frequency onc
> nap/H.
_
_
If r ir is imaginary, then c < a and this implies that 2 sinhkr H 0 which is
impossible (1< sinhx <1).
The particle motion inside the layer is circular when
u1 u3

169:4

so
u1 u3 ) 1  r exp ikrx3 1 r exp ikrx3 0
Taking only the amplitudes of the displacements,
1  r cos krx3 1 r cos krx3 0 ) cos krx3 0

1
1 p
krx3 n p ) x3 n
2
2 kr
For the FM, we have seen that u3 0, so there is no circular motion. For the 1HM, the
height in the layer at which the motion is circular is
3p
x3
2kr
The height inside the layer at which the motion is circular depends in each higher mode of
the frequency.
170. In the hypothetical case of a layer of thickness H and speed of propagation b0 on
top of a semi-innite medium of speed of propagation b, the phase shifts at the free
surface and the contact plane are
2 s 3
c2
7
6
6 b0  1 7
p
7
1 6
and
tan 6s7:
6
4
27
4 1 c 5
b2

317

Surface waves

Determine:
(a) The dispersion equation using constructive interference.
(b) The cut-off frequency of the fundamental mode and rst higher mode.
(c) Plot the dispersion curve of the FM and 1HM using units of c/b and H/l for
b 2b0 .
(a) The distance from A to B along the ray path is (Fig. 170a)
AB 2H cos i
According to Snells law
b0
c s
b0 c2
1
cos i
c b02

sin i

and the wavenumber k associated with velocity c is


kb0 sin i kb0

b0
k
c

As explained in Problem 167, the condition for constructive interference is that the distance
AB along the ray path be an integer multiple of the wavelength, taking into account the
phase shift at the free surface and the boundary surface between the two media:
0

p
s
2pn
170:1
2kB0 H cos i  tan1
4
s
where
s
c2
s0
1
b02
s
c2
s 1 2
b
By substitution in (170.1) we obtain the dispersion equation

B
H
A

Fig. 170a

318

Seismology

v
u 2
u c


 u
ub02  1
1
tan 2kHs0  2n  p u
u
4
c2
t
1 2
b

(b) For the fundamental mode (FM), n 0:


s
c2
s
"
#
1
b02
c2
p
tan 2kH
1
s > 0
where
b0 < c < b
4
2
b02
c
1 2
b
p
p
Given that tan B > 0 ) B so that b< c < b, then
4
2
If k 0 then
s
c2
1
p 0
 p
b02
2bb
1 s ) c p
tan
2
4
2
b b02
c
1 2
b
so

If c b then


p
p
tanB 1 ) B
tan 2kHs0
4
2
s
s
c2
p p
p
b2
0
2kH

)
k

)
s


1

1
4 2
8Hs0
b02
b02

so

If c b0 then


p
0
tan 2kHs0
4

s

c2
p
p
0
s0

1

0
)
tan
kHs
0 ) k 1 ) kHs0 

02
4
4
b

But kHs0 must be positive, so this last solution is not possible.


For the rst higher mode (1HM), n 1:


7
s0
0
tan 2kHs  p
4
s
The tangent function is positive for the range [0, p/2]:
0 2kHs0 
For c b0

7p p

4
2

170:2

319

Surface waves



7p
7p
tan 2kHs0 
)k1
0 ) kHs0
4
8
For c b


7p
7p p
9p
1 ) 2kHs0 
)k
tan 2kHs0 
4
4
2
8Hs0
(c) Taking b b/2, the dispersion equation for the FM is
s
4c2
s
"
#
1

2
b2
4c2
p
c
2

s
#
tan 2kH
)
1

"
s
2

2
4
b
2
b
c
H
c
1 2
5  3 sin 8p
4
1
l
b
b
But we have seen that
p p
H
1
)
4 2
l 16s0
p
c b ) s0 3

2kHs0

H
1
p
l 16 3

Giving values to H/l we can calculate the corresponding values of c/b by means of a
numerical method. For example, we obtain,
H/l

c/b

0.0
0.01
0.02
0.036

0.63
0.68
0.78
1

For the rst higher mode (1HM), we arrive at the same equation, given that tan(a p/4)
tan(a 7p/4), but vary the intervals of H/l and kHs0 :
7p
9p
kHs0
8
8
7
H
9

16s0
l 16s0
H
7
at one limit we have the value s0 0 which corresponds to c b0 and c/b
For
l 16s0
0.5.
H
9
H
9
we have s 0 and c b, and consequently p.
For
0
l 16s
l 16 3
The dispersion curves for the fundamental mode and the rst higher mode are shown in
Fig.170b.

320

Seismology

1.0
1 HM

FM

0.5

0.10

1
16

0.20

0.30
16

9
3

Fig. 170b

171. In a structure with a layer of thickness H and speed of propagation b0 on top of a


medium of speed of propagation b, the phase shifts at the free surface and the contact
plane are:
8 s 9
2>
>
>
>
> 1 c >
>
>
>
2>
=
<
b
p
d1  and d2 sin1 s
>
>
2
>
>
c2
>
>
>
>

1
>
>
;
: b02
Calculate:

(a) The dispersion equation of the Love wave.


(b) For the fundamental and rst higher mode, and the minimum and maximum
frequencies as functions of H, b, and b0 .
(c) For this mode, given b0 b/2, the maximum and minimum frequencies, and the
corresponding values of c.
(a) As in Problem 170, the condition of constructive interference with the phase shifts
given in this problem results in (Fig. 171)
s
4p
p
H cos i   sin1 0 2pn
l
2
s
where

171:1

321

Surface waves

H
A

Fig. 171

s
c2
s 1 2
b
s
c2
s0
1
b0 2
According to Snells law

b0
c s
b0 c2
1
cos i
c b02
sin i

Substituting in (171.1) we nd the dispersion equation for Love waves:


s
c2
s!
1

s
b2
p
c2


1
2Hks0   2pn sin1 0 )  cos 2Hk
2
s
b02
c2
1
b02
(b) The fundamental mode (FM) corresponds to the values
s
c2
1 2
b
p
0 2Hks0 ) 1 s 0
2
2
c
1
b02
The velocity at the limit of lowest frequencies, k 0, is given by
s
c2
1 2
b
s
2b02 b2
k 0 ) 0 s 1 ) c2 02
s
b b2
c2

1
b02
q
p
s
1
0
b2  b02
)

0
)
s

0
)
c

b
)
s

k
4Hs0
s0
b0

322

Seismology

For the rst higher mode (1HM)


s
c2
1 2
b
3p
2Hks0 2p ) 0 s 1
2
2
c
1
b02
In the lowest limit
k1

3p
s
1
) 0 0 ) s 0 ) c b ) s0 0
4Hs0
s
b

The minimum value of the frequency is


k1
In the highest frequency limit

q
b2  b02

3pb0
p
4H b2  b02

p
s
2b02 b2
k2 0 ) 0 1 ) s s 0 ) c2 2
) s0
Hs
s
b b02

s
b2  b02
b2 b02

The maximum value of the frequency is given by


p
p b2 b02
k2 p
H b2  b02

(c) If we put b0 b=2 in the 1MS,

b
p
p
3p
p 3 2p
H
3
2
r

)
;cb
k1
2
H4
l
l
8
b
2
4H b 
4
r
b2
r
p
p
p b2
p 5 2p
H
2
5
4
r p
) p ; c b
k2
2
l
l
5
H
3
2
3
b
4H b2 
4
172. Consider a layer of thickness H and parameters m0 and r0 on top of a semic
H
innite medium of parameters m 4m0 and r r0 . If a and b:
b
l
(a) Write the dispersion equation of the Love wave in terms of a and b.
1 3
(b) Calculate the values of b corresponding to a ; and 1 for the fundamental
2 4
mode and rst higher mode.
(c) For which values of b is the node of the amplitude of the rst higher mode at a
depth of H/2?
(a) The dispersion equation for Love waves is

323

Surface waves
q2
s)
m
1  bc 2
2
c

q
tan kH

1

2
b02
m0 c02  1
(

In this problem,

172:1

m 4m0
r r0
then,
r
r
m
m
b
0
!b

b
r
4r 2
We now introduce a and b:
a
b

c
c
) 0 2a
b
b

H
2p 2pb
)k

l
l
H

Substituting in (172.1):
p
p
1  a2
2
tan 2pb 4a  1 4 p
4a2  1


(b) For the FM

a 1; b 0 and c b
1
a ; b ! 1 and c b0
2
r
9
"
r#
1
3
9
16
a ) tan 2pb 4  1 4 r ) b 0:17
4
16
9
4 1
16
For the 1HM,
1
a ; b!1
2


p
p
1
a 1 ) tan 2pb 4a2  1 0 ) 2pb 3 p ) b p 0:29
2 3
0
1
s
r
r!

2
p
3
3
7
7
1
 1A 4
4
) pb 5 tan
p ) b 0:61
a ) tan@2pb 4
4
4
20
20
(c) Inside the layer the amplitude of the displacements of the Love wave are given by
h

x3 i
u02 2A0 cos ks0 H 1 
cos k s0 H x1  ct
H

324

Seismology

The nodes are the points where the amplitude is zero. For the 1HM the node is located at
the value of x3 which satises the relation

x3  3p

ks0 H 1 
H
2
If we want a node located at x3 H/2 then
k

3p
3
) b p
s0 H
2 4a2  1

If we substitute the values of a, , 1, and we obtain for b innity, 0.86, and 1.34. The
innite value of b corresponds to l 0.

Focal parameters
173. Consider three stations with coordinates:
St1 36.2 N, 4.8 E;
St2 37.0 N, 2.4 E; St3 38.6 N, 4.0 E
An earthquake is recorded at the three stations with the following respective S-P
time intervals:
t 1SP 26:7 s
t SP
27:0 s
2
t SP
22:5 s
3
Given that the focus is at the surface, the P-wave velocity is constant and equal to 6
km s1, and Poissons ratio is 1/3, calculate the coordinates of the epicentre.
Given that Poissons ratio is 1/3,
1
l
s
) l 2m ) a
3 2l m

s
l 2m
2b
r

where a and b are the velocities of the P- and S-waves, respectively.


From the S-P time intervals, calling x the epicentral distance (Fig. 173):
x x
x
) x 2t SP b
t SP 
b a 2b
Given that a 6 km s1 then b 3 km s1.
The epicentral distances corresponding to each station in kilometres and degrees are
x1 26:7  6 160 km
x2 27  6 162 km

160
1:44
111:11

162
1:46
111:11

x3 22:5  6 135 km

135
1:22
111:11

325

Focal parameters

40N

39N
St3

38N
E2
E1
St2

37N

St1
36N

35N
1E

2E

3E

4E

5E

6E

Fig. 173

If xe and ye are the geographical coordinates of the epicentre,


36:2  xe 2 4:8  ye 2 1:442
37:0  xe 2 2:4  ye 2 1:462
38:6  xe 2 4:0  ye 2 1:222
Solving the system we nd two possible solutions:

37:40 N; 3:83 E
xe ; ye
37:63 N; 4:71 E
174. A focus is at a depth h, in a medium with constant P- and S-wave speeds of
propagation a and b. Calculate an expression for the depth h in terms of the time
interval between the P and sP phases.
If L is the distance along the ray, the travel-time of the P-wave from the focus to the station
is given by (Fig. 174)
tP

L
a

326

Seismology

jh

ih

F
ih
A

Fig. 174

An sP-wave leaves the focus upward as an S-wave is reected at the Earths surface and
converted into a P-wave that travels to the station. Its travel time is (Fig. 174)
0

t sP

FS SF L

b
a
a

If we consider the length along the ray L from F and F0 to the station to be the same for both
waves, then the sP-P time interval is (Fig. 174)
t sP  t P t 0

FS SF

b
a

174:1

At the focus at depth h, the takeoff angle of the direct P-wave is ih and the take-off angle
of the sP-wave is jh. We can write
FS

h
cos jh
0

SF
h
0
cosjh ih
) SF
sin jh sin ih  cos jh cos ih
cos
jh
FS
a sin jh
sin ih
b
By substitution in (174.1) we obtain
t 0 h

cos jh cos ih

b
a

175. The displacement vector l of an earthquake is (0, 1, 0) and the vector normal to
the plane of displacement n is (0, 1, 0). Determine:
(a) The components of the P-wave displacement at the point of azimuth 45 and angle
of incidence 30 .

327

Focal parameters

(b) The kind of mechanism it represents.


(a) The elastic displacements due to a dislocation Du in the direction of li on a plane of
normal ni (Fig. 175) are given by



uPk u lnk lk dij m li nj ni lj Gki;j

where Greens function corresponding to the P-wave in the far eld for an innite medium
is given by

1
r
g
g
d
t

Gki
i
k
4pra2 r
a
and its derivative is

Gki;j



1
_ tr
g
g
g
d
4pra3 r i k j
a

where gi are the direction cosines of the line from the focus to the observation point.
The amplitude of the displacement is then
uPk



u
lns ls dij m li nj lj ni gi gk gj
3
4pra r

175:1

In our problem the direction cosines of the ray of the waves arriving at the point are
p
2
g1 sin i cos az
4
p
2
g2 sin i sin az
4
p
3
g3 cos i
2
X3

P
r
i
li= ni

e
az

X2

Fig. 175

X1

328

Seismology

The orientation of the source is given by li (0, 1, 0) and ni (0, 1, 0), and substituting in
(175.1) gives


uP1 A l 2mg22 g1


uP2 A l 2mg22 g2


uP3 A l 2mg22 g3
u
A
4pa3 rr
Substituting the direction cosines of the ray we obtain,

p
2
1
uP1 A l m
4
4

p
2
1
P
u2 A l m
4
4

p
3
1
P
u3 A l m
2
4
(b) The mechanism corresponds to a fault on the (x1, x3) plane which opens in the
direction of its normal, x2, under tensional forces in that direction.
176. The focal mechanism of an earthquake can be represented by a double-couple
(DC) model. The orientation of the fault plane is azimuth 30 , dip 90 , and slip angle
0 . Calculate:
(a)
(b)
(c)
(d)

What kind of fault it is. Sketch it, indicating the direction of motion.
The auxiliary plane.
The azimuth of the stress axis.
A wave incident at a station has azimuth 180 and angle of incidence at the focus
of 90 . Calculate the amplitude of the components of the P-wave at that station.
(a) Given that the dip of the plane is 90 , the fault plane is vertical, and since the slip
angle is 0 , the motion is horizontal. Thus, it corresponds to a right lateral strikeslip fault (Fig. 176).
N

A
u
l

A
n

Fig. 176

329

Focal parameters
(b) From the azimuth ( 30 ), dip (d 90 ), and slip (l 0 ) we can calculate the
unit vectors li and ni which give the direction of the fracture and of the normal to
the fault plane:
1
n1  sin d sin  sin Yn cos Fn
p 2
3
sin Yn cos Fn
n2 sin d cos
2
n3  cos d 0 cos Yn
p
3
l1 cos l cos cos d sin l sin
sin Yl cos Fl
2
1
l2 cos l sin  cos d sin l cos sin Yl cos Fl
2
l3  sin l sin d 0 cos Yl
A Fn 90 ) Fn 120
dA Yn 90

cos Yl
lA sin1
0 ) Yi 90
sin Yn
where Y and F are the spherical coordinates for the vectors n and l (r 1, unitary vectors)
and for the auxiliary plane,
B 30 90 120
dB 90
cos Yn
lB sin1
0
sin Yl
(c) The T-axis is on the same plane as ni and li at 45 between them, and the direction
cosines are
0 1 0
10 1 1
p
T1
n1 n2 n3
2
@ T2 A @ l1 l2 l3 A@ p1 A
176:1
2
Z1 Z2 Z3
T3
0

where Zi is the axis normal to ni and li, that is, Zi ni  li which results in Zi (0, 0, 1).
p
p

31 31
p ; p ; 0 . The azimuth of the T-axis is
Substituting in (176.1) we obtain Ti
2 2
2 2
p

31
1 T2
1
p
75
FT tan
tan
T1
31
(d) The direction cosines of the direction from the focus to the station are
g1 sin ih cos az 1
g2 sin ih sin az 0
g3 cos ih 0

330

Seismology

The amplitude of the displacements for a shear fracture or double-couple (DC) source in an
innite medium is given by
Rih ; az ) uPj Ani lk nk li gi gk gj
uP1 0:84A
uP2 0
uP3 0
A

M0
4pa3 rr

177. An earthquake is caused by a shear fracture. The vectors n and l (normal and
direction of travel) are, in terms of the angles F and Y,
n 57:13 ; 66:44
l 305:96 ; 50:09
Calculate the orientation of the fault plane, the auxiliary plane, and the tension (T )
and pressure (P) stress axes.
The orientation of the fault plane and auxiliary plane in terms of the angles , d, and l (azimuth,
dip, and slip) are found directly from the given values (Fig. 177), using the following relations:
A Fn 90 147:13
dA Yn 66:64

1 cos Yl
135:68
lA sin
sin Yn
B Fl 90 35:96
dB Yl 50:09

1 cos Yn
lB sin
31:12
sin Yl
Z

Y
qp

X
n

l
qn
T

P
fn
N

Fig. 177

331

Focal parameters

To calculate the T and P axes, we calculate rst the direction cosines of the l and n axes
from the given angles:
x1 sin Y cos F

n1 0:50; l1 0:45

x2 sin Y sin F ) n2 0:77; l2 0:62


x3 cos Y

n3 0:40; l3 0:64

Given that the T and P axes are on the same plane as n and l at 45 between them, we can
write, as in Problem 176,
0
1
0 1 0
1 p1
T1
l1 n1 Z1 B 2 C
@ T2 A @ l2 n2 Z2 AB p1 C
177:1
@
A
2
T3
l3 n3 Z3
0

where the Z-axis is normal to n and l and is found by Z n  l. Its direction cosines
are (Z1, Z2, Z3) (0.72, 0.14, 0.66). By substitution of ni, li, and Zi in (177.1) we
obtain
9
T1 0:67 sin YT cos FT >
=
T2 0:11 sin YT sin FT
) T YT 42:27 ; FT 9:32
>
;
T3 0:74 cos YT

In the same way for the axis P


1 0
l1
P1
@ P2 A @ l2
P3
l3
0

n1
n2
n3

0 1 1
p
1
Z1 B 2 C


1 C
Z2 AB
@  p A ! PYP 80:02 ; FP 268:03
2
Z3
0

178. The seismic moment tensor relative to the geographical axes (X1, X2, X3) (north,
east, nadir) is
0
1
2 1 1
Mij @ 1 0 1 A
1
1 2
Find the values of the principal stresses, and the orientation of the tension and
pressure stress axes.

First we calculate the eigenvalues of Mij. Since Mij is a symmetric tensor its eigenvalues are
real and the corresponding eigenvectors mutually orthogonal (Problem 111):


0
1
2  s 1
3
1

1
0s
1 0 ) s @ 2 A

1
1
1
2  s

Ordered by magnitude, the three eigenvalues are

s1 3; s2 2; s3 1

332

Seismology

The diagonalized matrix is


0

3
Mij @ 0
0

0
2
0

1
0
0 A
1

In this form Mij is referred to the coordinate system formed by the eigenvectors or principal
axes. Given that the sum of the elements of the principal diagonal is not zero, the source
has net volume changes. Then, we can separate Mij into two parts: an isotropic part with
volume changes (ISO) and a deviatoric part without volume changes. The second part can
be separated into two parts: a part corresponding to a double-couple or shear fracture (DC)
and a part corresponding to a non-double-couple source usually expressed as a compensated linear vector dipole (CLVD). Thus the moment tensor is separated into three parts,
namely
M M ISO M DC M CLVD
The isotropic part is given by
1
4
M ISO s0 s1 s2 s3
3
3
The deviatoric part (DCCLVD) is given by
Mij0 Mij  dij so
and in our case

Mij0

s1
@ 0
0

05

0
B3
0
B
2
0 AB
B0 3
@
s3
0 0
1

0
s2
0

Now we separate this part into two parts, DC and CLVD:

C
C
0 C
C
A
7

3

M 0ij M DC M CLVD
1
s  s3
B2 1
B
0
Mij B
0
@
0
0

Mij0

2
B
@0
0

0
0
0

0
0
0

s2
C B 2
C B
CB 0
0
A @
1
0
 s1  s3
2
1

1
3

B
B
C B
0 AB 0
B
@
2
0

C
C
C
0 C
C
A
1
0 
3
2
3

0
s2
0

C
C
0 C
s2 A

2

333

Focal parameters
The orientation of the P and T axes is calculated from the double-couple part MDC:
0
1
1 0 0
MijDC 2@ 0 0 0 A
0 0 1

We can nd the n and l axes:

MijDC M0 li nj  lj ni

1
1
~
) nYn 45 ; Fn 0
n : p ; 0;  p
2
2

1
1
~
p

; 0;
) lYl 45 ; Fl 0
l:
2
2

In the same way as in Problems 176 and 177, we determine T and P from n and l, nding
rst Z n  l:
0
1
0 1 0
1 p1
T1
l1 n1 Z1 B 2 C
C
@ T2 A @ l2 n2 Z2 AB
B 1 C
@ p A
T3
l3 n3 Z3
2
0
T1 1 sin YT cos FT
T2 0 sin YT sin FT ! T YT 90 ; FT 0
T3 0 cos YT
For the P-axis,
0

P1
l1
@ P2 A @ l2
P3
l3

n1
n2
n3

1 1
p
Z1 B 2 C
B
1 C
Z 2 AB
CPYP 0 ; FP 0
@  p A
Z3
2
0
1

179. The magnitude Ms of an earthquake as calculated for surface waves of period


20 s is 6.13.
(a) Calculate the amplitude of these waves at a station 3000 km away. If the instruments amplication is 1500, what will be the amplitude of the seismograms
waves and the seismic energy?
(b) If Ms Mw, and the area of the fault is 12 km  8 km with m 4.4104 MPa, nd
the fault slip Du.
(a) The surface wave magnitude Ms is given by
Ms log

A
1:66 log  3:3
T

334

Seismology

where A is the ground motion amplitude, T is the period of the wave, and D is epicentral
distance in degrees. Knowing the magnitude and period of the waves we can calculate the
wave amplitude:
A
3000
1:66 log
3:3
T
111:11
A
) log 0:454 ) A 2:84  20  1500 8:5 cm
T

6:13 log

We have reduced the ground motion to the amplitude of the seismogram using the
amplication of the instrument (1500).
Knowing the magnitude we can calculate the seismic energy:
log Es 11:8 1:5Ms ) Es 1021 ergs 1014 J
(b) Mw 6.13 2/3 log M0 6.1; M0 2.191018 Nm
If M0 mDu S, with S 12 8 9.6107 m2, then
u

M0
2:19  1018 Nm

0:52 m
mS 4:4  1010 N m2  9:6  107 m2

Heat ow and geochronology

Heat ow
180. Assume that the temperature variation within the Earth is caused by gravitational forces under adiabatic conditions. Knowing that the coefcient of thermal
expansion at constant pressure is aP 2105 K1 and the specic heat at constant
pressure is cP 1.3 kJ kg1 K1, determine an expression for the gradient of the
temperature with depth. Compare it with the value observed at the surface which is
30 K km1, knowing that, at 200 km depth, T 1600 K.
Under adiabatic conditions, there is no heat ow and the variation of pressure with depth z
is a function of gravity g and density r:
dP rgdz

180:1

Using the rst and second laws of thermodynamics


dU dQ  PdV
dQ TdS
where Q is the heat, U is the internal energy, S is the entropy, T is the absolute temperature,
P is the pressure, and V is the volume.
If we use the specic variables (variables divided by mass) we can write
du dq  pdv
dq Tds
Considering that
dS

@S
@T

dT

@S
@P

dP

we can write

@s
dq Tds T
@T

@s
dT T
@P
P

dP

According to the denition of specic heat at constant pressure, cP and the increase in heat
dq are given by
335

Heat ow and geochronology

336

cP T

@S
@T

dq cP dT T

@s
@P

180:2
dP

The Gibbs function G is dened as


G u  Ts pv
Taking the differential in this expression, and taking into account the second law of
thermodynamics
du Tds  pdv
we obtain
dG vdp  sdT
If we compare this expression to the differential of the Gibbs function

@G
@G
dG
dp
dT
@p T
@T p
we differentiate again and using the Schwartz theorem we obtain

@v
@s

@T P
@P T
But the coefcient of thermal expansion is dened as

1 @v
ap
v @T p
In consequence we can write Equation (180.2) as
dq cp dT  Tvap dp
In our case the process is adiabatic and in consequence using this equation and Equation
(180.2) we obtain
dT T ap g

cp
dz

180:3

where we have taken into account that the variables are by unit mass so rv 1, and
substituting the values we obtain:
dT 1600  2  105  10 K K1 ms2

0:25 K km1
dz
1:3  103 J kg1 K1
We observe that this result is two orders of magnitude lower than the observed values. This
shows that observations correspond to heat ow at the lithosphere and are not satised by
purely adiabatic conditions.

Heat ow

337

181. If the Earths temperature gradient is 1  C/30 m, calculate the heat loss per
second due to conduction from its core. Compare this with the average power
received from the Sun.
Data:
Thermal conductivity K 4 W m1  C1.
Earths radius R 6370 km.
Solar constant: 1.35 kW m2.
The heat ow is given by
q_ KA

dT
dr

where A is the area of the Earths surface


A 4pR2 5:10  1014 m2
and
dT
1 C

dr
30 m
Substituting the values
q_ 6:80  1013 J s1 1:63  1013 cal s1
the average power received on the Earths surface by the radiation from the Sun is
1:35  103

W
5:10  1014 m2 6:89  1017 J s1 1:65  1017 cal s1
m2

In consequence, from these values we can see that on the Earths surface the average solar
power is much larger than that due to the heat ow from inside the Earth.
182. At the Earths surface, the heat ow is 60 mW m2 and T0 0  C. If all the heat is
generated by the crust at whose base the thermal conductivity is K 4 W m1  C1, and
T is 1000  C, determine the thickness of the crust and the heat production per unit volume
If we assume that all the heat is generated at the crust and there is no heat ow from the
mantle at the crust base, then we can write
_ zH 0
qj
Using the temperature equation for a at Earth for one-dimensional heat-ow and the
stationary case we can write
T 
e

e 2 q_ 0
z z T0
K
2K

q_ 0
H

For z H:
TH

q_ 0 H
TH  T0 2K
T0 ) H
2K
q_ 0

Heat ow and geochronology

338

Substituting the values given in the problem we obtain


H 133:3 km
and the heat production by unit of volume is
e

q_ 0
4:5  104 mW m3
H

183. Consider the crust to be H 30 km thick and the heat ow at the surface to be
60 mW m2.
(a) If all the heat is generated in the crust, what is the value of the heat generated per
unit volume? (Take K 3 W m1 K1)
(b) If all the heat is generated in the mantle with a distribution Aez=H mW m3 , what
is the value of A? What is the temperature at 100 km depth?
(a) We solve the heat equation for a stationary one-dimensional case, assuming a at
Earth with one-dimensional ow in the z-direction (vertical) positive downward.
In this case the solution of the heat equation is given by
T 

e 2 q_ 0
z z T0
2K
K

where e is the heat generated by unit volume and time, K is the thermal conductivity, q0 and
T0 are the heat ow and temperature at the surface of the Earth, respectively
If all the heat is generated at the crust we can write (Fig. 183)
z H ! q_ 0 0
The heat generated by unit volume is


dT
e2z q_ 0
_ zH 0 K
K 

qj
zH 0
dz zH
2K K
e

q_ 0 60  103

2  106 W m3
H
30  103

(b) If all heat is generated in the mantle with distribution


e Aez=H
the heat equation is
d2 T
e
A
  ez=H
dz2
K
K
and the solution is given by
T 

A 2 z=H
H e
Cz D
K

183:1

where C and D are constants of integration. They may be estimated from the boundary
conditions at the surface

Heat ow

339

q0
T0

qH
TH

Fig. 183

A 2
A
H D ! D T0 H 2
K
K
A
z 0 ! q_ q_ 0 0 ! C  H
K

z 0 ! T T0 

Substituting in (183.1) we obtain


T 


A 2 z=H AH
A AH 
H e
z T0 H 2
Hez=H  z H T0

K
K
K
K

183:2

If the heat has its origin in the mantle, the ow at the base of the crust is

dT
q_ 0
_ zH q_ o K
)A
qj
dz zH
H e1  1

The temperature at z = 100 km may be estimated from (183.2) assuming that T0 = 0:




q_ 0
z=H
He

z

H
T z
K e1  1
T z 100


60  103 
3 100=30
3
3
30

10
e

100

10

30

10
2249 K
3e1  1

184. Calculate the thickness of the continental lithosphere if its boundary coincides
with the 1350  C geothermal, knowing that the surface temperature is 15  C,
the heat ow at the surface is q_ 0 46 mW m2 , the lithospheric mantles thermal
conductivity is K = 3.35 W m1 K1, and the radiogenic heat production is
P = 0.01  103 mW m3
The geothermal equation at depth z is given by:
Tz T0

q_ 0
P0
z  z0 
z  z0 2
K
2K

184:1

Heat ow and geochronology

340

where K is the thermal conductivity, T0 is the temperature at the surface of the Earth (in K), q_ 0
is the heat ow at the surface, and P0 is the radiogenic heat production at the Earths surface.
At the Earths surface z0 = 0, and Equation (184.1) becomes:
Tz T0

q_ 0
P0 2
z
z
K
2K

P0 2 q_ 0
z  z Tz  T0 0
2K
K
So

q_ 0

K

s

2
q_ 0
P0
p
4
Tz  T0
K
2K
q_ 0  q_ 20  2P0 KTz  T0

P0
P0
2
2K

Substituting the data given in the problem:


K 3.35 W m1 K1
q_ 0 46  103 W m2
P0 0.01106 mW m3
Tz 1623 K
T0 288 K
we obtain two solutions, but only z = 98.27 km is realistic (the second one gives a depth
larger than the Earths radius).
185. On the surface of an Earth of radius 6000 km, the temperature is 300 K, the heat
ow is 6.7 mW m2, and the thermal conductivity is 3 W m1 K1. If the heat production per unit volume inside the Earth is homogeneously distributed, what is the
temperature at the centre of the planet?
We begin solving the problem of heat conduction inside a sphere with constant internal
heat generation per unit volume e and conductivity K. The differential equation for heat
conduction with spherical symmetry is

1 d
e
2 dT
r
185:1

2
r dr
dr
K
Integrating twice and using the boundary conditions:
Surface: r = R ! T = T0
Center: r = 0 ! T nite
we obtain the solution

e  2
T T0
R  r2
6K

The heat ow is given by


dT
q_ K
) q_ 0 r R 6:7 mW m2
dr
 e r
d
e  2
T0
R  r2
K
dr
6K
3

185:2

Heat ow

341

Solving for the heat production e


e 3:35  109 Wm3
From Equation (185.1) the temperature at the Earths centre is
Tr0 300

3:35  109
 62  1012 7000 K
63

186. Consider a spherical Earth of radius R = 6000 km and a core at R/2, in which
there is a uniform and stationary distribution of heat sources per unit volume. The
heat ow at the surface is 5 mW m2, the thermal conductivity is 3 W m1 K1, and
the temperature at the coremantle boundary is 4000 K. Calculate the temperature at
the Earths surface.
We consider the problem as one of heat conduction inside a sphere with conductivity K and
constant heat generation per unit volume e inside the core (radius R/2). We begin with
Equation (185.1)

1 d
e
2 dT
r

r2 dr
dr
K
The boundary conditions at the coremantle boundary and its centre are
r R=2 ! T TN
r 0 ! T finite
where TN is the temperature at the coremantle boundary
Integrating twice we obtain
e
T r TN
6K

!

2
R
 r2
2

The heat ow is given by



dT
eR
q_  3

)e 0
q_ 0 r R K
2:5  109 W m3
dr rR
3
R

Then, the temperature at the Earths surface is


2:5  109
T r R 4000
63

6000  103
2

 6000  10

250 K

187. Consider the Earth of radius R0 = 6000 km formed by a spherical crust


with its base at 500 km and constant thermal conductivity K. If the temperature
at the base of the crust is T1 and at the surface of the Earth is T0 = 0  C,
determine:
(a) An expression for the heat ow through the crust.

Heat ow and geochronology

342

(b) An expression for the temperature distribution within the Earth.


(c) The temperature at the base of the crust if, at that depth, q_ 5:5  1013 W and
K = 4 W m1  C1.
(a) We assume a spherical Earth where the temperature varies only in the radial
direction. Then we can solve the problem as one of spherical unidirectional ow.
For the stationary case, when the conductivity and heat generation are constant, the
Fourier law may be written as
q_ KA

dT
dr

187:1

where A 4pr2 is the area in the normal direction to the heat ow. Integrating this
equation:
q_
4p

R0
r1

dr
K
r2

T0

dT

T1

where the conditions at the Earths surface are, r = R0 !T = T0 and at the base of the crust,
r = r1 !T = T1.
Solving Equation (187.1), assuming that K is constant, we obtain
4pK
4pKr1 R0
T0  T1
T0  T1
q_  1
1
R0  r1

r1
R0

187:2

(b) The temperature distribution inside of the Earth may be obtained by integration of
Equation (187.1):

T
q_ r dr

K
dT
4p r1 r2
T1
T r T1

R0 r  r1
T0  T1
R0  r1 r

(c) The radial distance to the base of the crust is r2 = 5500 km, so, using expression
(187.2), we obtain
T1 T0

_ 0  r1
qR
5:5  1013  500  103
0
16579  C
4pKr1 R0
4p  4  5500  103  6000  103

This result implies a constant increase of temperature from the Earths surface
of 1 C each 33.2 m similar to the observed gradient in the real Earth of 1 C
per 30 m
188. Assume that the heat ow inside the Earth is due to solar heating of the
Earths surface. Calculate the maximum penetration of this ow in the diurnal and
annual cycles. Take as typical values for the Earth K = 3 Wm1 K1, r = 5.5 g cm3,
Cv = 1 kJ kg1 K1.

Heat ow

343

We assume the heat propagation inside the Earth coming from the solar radiation on its
surface as unidirectional ow thermal diffusion (inside the Earth) with periodic variation of
surface temperature. The diffusivity equation is
k

@ 2 T @T

@z2
@t

188:1

K
, K is the thermal conductivity, r is the density,
rCv
and Cv is the specic heat at constant volume. We solve Equation (188.1) using the
separation of variables
where the thermal diffusivity is k

T z; t Z zyt
Substituting in (188.1) we obtain the solution
Z z Aeaz Beaz
yt Ceka

where a is the constant of separation of variables. Using the boundary condition of periodic
ow and the temperature T0 at the Earths surface,
z 0 ) T T0 eiot
and as Z(z) exists only inside the Earth, B = 0. At the surface, z = 0, so
2

ACeka t T0 eiot
Then
AC T0
ka2 io
But putting, i 12 1 i2 , we have
r
o
a 1 i
2k
Then, we can write the temperature variation inside of the Earth as:
 r
r


o
o
zi 
z ot
T z; t T0 exp 
2k
2k
This equation corresponds to a periodic wave, with angular frequency o propagating for
positive z values (to the Earths interior) and with the amplitude decreasing with depth. The
propagation velocity and wavelength are given by
r
2k
v
o
r
8k
l 2pv p
o

Heat ow and geochronology

344

The values of l corresponding to the daily and annual cycles give their maximum penetration:
Daily cycle:
2p
7:2  105 s1
24  60  60
K
3 W m1 K1

0:5  106 m2 s1


k
rCv 5:5  103 K gm3  103 J Kg1 K1

Then
r
8  0:5  106
0:74 m
lp
7:2  105
Annual cycle:
2p
2  107 s1
365  24  60  60
l 14 m

The penetration of the solar radiation as periodic heat conduction inside the Earth is very
shallow due to the poor heat conduction.
189. Consider a lithospheric plate of 100 km thickness created from asthenospheric
material originating from a ridge in the asthenosphere with constant temperature Ta and
in which no heat is generated. Given that k = 106 m s1, that the temperature at the base
of the lithosphere is 1100  C, and in the asthenosphere is 1300  C, calculate the age of the
plate, and, if the velocity of drift is 2 cm yr1, how far it has moved away from the ridge.
The heat propagation inside the plate is given by:

K @2T @2T
@T
2 u
rcv @x2
@z
@x

189:1

where T is the temperature, r is the density, cv is the specic heat at constant volume, and u
is the horizontal velocity of the plate in the x-direction (normal to the plate front). If we
assume that the horizontal conduction of heat is insignicant in comparison with the
horizontal advection and vertical conduction, we can write, using the following change
of variable t = x/u,
K @ 2 T @T

rcv @z2
@t
Integrating this equation and using the boundary conditions at the ridge and surface:
x 0 ! T Ta
z0!T 0
we obtain for the temperature distribution
T z; t Ta erf

z
p
2 kt

345

Geochronology

where
K
rcv

2 x
2
erf x p ey dy
p 0
k

Substituting the data of the problem,


L
L
1100 1300erf p ) erf p 0:846
2 kt
2 kt

Values of the error function, erf(x), may be obtained from tables. If erf(x) = 0.846, x = 1.008, then

L
L2
1010
p 1:008 ) t

4k  1:0082 4  106  1:0082


2 kt
2:5  1015 s 79 Myr
If the displacement velocity is 2 cm yr1, the plate has moved 1580 km.
190. If the concentrations of 235U and 235Th in granite are 4 ppm and 17 ppm,
respectively, and the respective values of heat production are 5.7  104 W kg1
and 2.7  105 W kg1, respectively, calculate the heat ow at the base of a granite
column of 1 m cross-section and 30 km height (the density of granite is 2.65 g cm3).
We estimate rst the mass of the granite column:
M rV 2:65  103  1  30  103 7:95  107 kg
If the concentration of

235

U in the granite is 4 ppm, its quantity in the column is


235

Then the heat ow due to the

U:

235

4  7:95  107
318 kg
106

U is

q_ 5:7  104  318 181:26 mW m2


For 235Th, the heat ow is
235

Th : 7:95  107  17  106 1351:5 kg

q_ 1351:5  2:7  105 36:49 mW m2

Geochronology
191. The mass of 1 millicurie of 214Pb is 3  1014 kg. Calculate the value of the decay
constant of 214Pb.
The mass of the sample is

Heat ow and geochronology

346

3  1014 kg 3  1017 g

N
M
N0

where N is the number of atoms in the sample, N0 is Avogadros number 6.02  1023,
and M is atomic number 214. Solving for N we obtain
N

3  1017  6:02  1023


8:44  1038 atoms
214

The correspondence of a curie is


1 curie

dN
lN 3:7  1010 disintegrations s1
dt

where l is the decay constant and t is the time. Then


l

3:7  107
0:44  1031 s1
8:44  1038

192. The isotope 40K decays by emission of b particle with a half-life of 1.83  109
years. How many b decays occur per second in one gram of pure 40K?
The average life t of a radioactive material is a function of the decay constant l:
t

1
1
!l
0:55  109 yr1 1:73  102 s1
l
1:83  109

The number of atoms N contained in 1 g of 40K may be estimated from Avogadros number
N0 and the atomic number M:
1

N
1  6:02  1023
M !N
0:15  1023 atoms
N0
40

The rate of disintegration is given by


dN
lN 0:26  1021 Bq
dt
193. The half-life of 238U is 4468  106 yr and of 235U is 704  106 yr. The ratio
235 238
U/ U in a sample is 0.007257. Given that the ratio was 0.4 at the time of
formation, calculate the samples age.
From the half-life T1/2 we can obtain the decay constant l:
0:693
l
0:693

1:5510  1010 yr1


4468  106
0:693
9:8434  1010 yr1

704  106

T1=2
l238
l235

The number N of disintegrating atoms at time t is given by


N N0 elt

193:1

347

Geochronology

where N0 is the number of atoms at time t 0. Then


N235 N0235 el235 t
N238 N0238 el238 t
If we divide these equations:
N235 N0235 l238 l235 t

e
N238 N0238
Substituting the values given in the problem we obtain
 4:0095 8:2924  1010 t ! t 4:8  109 yr
194. Date a meteorite which contains potassium knowing that its content of 40K is
1.19  1014 atoms g1, of 40Ar is 4.14  1017 atoms g1, and that the half-life of
40
K!40Ar is 1.19  109 years.
We obtain the decay constant of the

40

K from its half-life T1/2:

0:693
T1=2
0:693
l
0:58  109 yr1
1:19  109

We can solve the problem considering it as a case of radioactive parent atom disintegrating to a daughter stable atom. At time t 0 we have n0 parent atoms at the sample,
and at time t there remain NR radioactive atoms in the sample and NE daughter atoms,
from the disintegration of the n0 parent atoms:
n0 NR NE
But
n0 NR NE
NE
1

nt
NR
NR
From Equation (193.1) we obtain the age of the sample:
nt n0 elt !

n0
NR
elt 1
nt
NR

194:1

so

1
NE
t ln 1
l
NR
NE, the number of atoms of 40K in the sample, can be estimated from the number of atoms
contained in 1 g of potassium:
1

N
40 ! N 1:506  1022 atoms g1
6:023  1023

Heat ow and geochronology

348

so
NE 1:506  1022  1:19  1014 1:792  1036 atoms g1
Then the age of the meteorite from (194.1) is

1
1:792  1036
t ln 1
7:4  1010 yr
l
4:41  1017
195. At an archaeological site, human remains were found and assigned an age of
2000 years. One wants to conrm this with 14C dating whose half-life is 5730 yr. If the
proportion of 14C/12C in the remains is 6  1013, calculate their age. (Assume that at
the initial time the 14C/12C ratio was 1.2  1012.)
The decay constant l may be obtained from the half-life:
T1=2

0:693
0:693
!l
1:2094  104 yr1
l
5730

The activity in a sample is given by


R R0 elt
where R0 (14C/12C)t 0 and R (14C/12C).
Then the age of the remains is

1
R0
1
1:2  1012

ln
2379 yr
t ln
l
R
1:2094  104
9  1013
196. Mass spectrometry of the different minerals in an igneous rock yielded the
following table of values for the concentrations of 87Sr originating from the radioactive decay of 87Rb and of 87Rb, with the concentration expressed relative to the
concentrations of 86Sr of non-radioactive origin.
Mineral

87

Sr /86Sr

A
B
C
D
E
F

0.709
0.715
0.732
0.755
0.756
0.762

87

Rb /86Sr

0.125
0.418
1.216
2.000
2.115
2.247

Express on a 87Sr/86Sr87Rb/86Sr diagram the isochron corresponding to the formation of the rock, and calculate the age of the rock. Take l 1.42  1011 yr1.
For the decay of

87

Rb




87
Sr now 87 Sr 0 87 Rb now elt  1

196:1

349

Geochronology

0.80

0.78

86Sr

87Sr

0.76

0.74

0.72

0.70
0.0

0.5

1.0

1.5

2.0

2.5

87Rb
86Sr

Fig. 196

where [87Sr]now and [87Rb]now are the number of atoms of each isotope at time t, [87Sr]0 is the
amount of original number of atoms of the isotope 87Sr [87Sr]now, and l is the decay constant.
Equation (196.1) may be written as



87
87

Sr
Sr 87 Rb  lt
86 86
e 1
196:2
86 Sr
Sr 0
Sr now
now

87
Sr
This equation corresponds to a line (y a bx) with intercept 86 and slope (elt  1),
Sr 0
which is called an isochron.
If we plot the values given in the problem (Fig. 196) we can obtain the equation of the
line by least-squares tting:
y 0:025x 0:705
The age of the sample can be obtained from the slope b 0.025:
el t  1 b
ln1 b
t
l
Substituting the values of b and l:
t 1:72  109 yr:

Heat ow and geochronology

350

197. Magma with a material proportion of 87Sr/86Sr equal to 0.709 crystallizes


producing a series of rocks with different concentrations of 87Rb with respect to the
content of 86Sr:
Sample

87

Rb/86Sr

A
B
C
D

1.195
2.638
4.892
5.671

(a) Calculate the proportions of 87Sr/86Sr and 87Rb/86Sr that these rocks will have
after 500 Myr. Take l 1.421011 yr1.
(b) Express in a 87Sr/86Sr87Rb/86Sr diagram the isochrons corresponding to t 0
and t 500 Myr.
(a) Using the same method as in the previous problem, we can write


87
87

Sr 87 Sr 87 Rb  lt
Rb  1:421011 5108

86
e  1 0:709 86
e
1
86 Sr
86 Sr
Sr
Sr
0
0.78

0.76

86

87

Sr
Sr

t = 500 Ma

0.74

0.72
t=0

0.70

4
87Rb
86Sr

Fig. 197

351

Geochronology

The results for each rock are given in the following table
Sample

87

Sr/86Sr

A
B
C
D

0.717
0.728
0.744
0.749

87

Rb/86Sr

1.187
2.619
4.857
5.631

(b) For t 500 Myr, we carry out a least-squares tting to obtain the isochron,
which results in
y = 0.007x 0.709
In Fig. 197 the isochrones corresponding to t 0 and t 500 Myr are shown.

Bibliography

General geophysics
Berckhemer, H. (1990). Grundlagen der Geophysik. Wissenschaftlische Buchgeselschaft,
Darmstadt.
Buforn, E., C. Pro, and A. Udas (2010). Problemas Resueltos de Geofsica. Pearson, Madrid.
Cara, M. (1989). Geophysique. Dunod, Paris.
Coulomb, J. and G. Jobert (1973 and 1976). Trait de Gophysique Interne (I and II).
Masson, Paris.
Fowler, C.M. R. (2005). The Solid Earth: An Introduction to Global Geophysics. (2nd
edn). Cambridge University Press, Cambridge.
Garland, G.D. (1979). Introduction to Geophysics. W.B. Saunders, Philadelphia.
Kaufman, A.A. (1992). Geophysical Field Theory and Method. Academic Press, San Diego.
Larroque, C. and J. Virieux (2001). Physique de la Terre solide. Observations et Thories.
Gordon and Breach, Paris.
Lillie, R.J. (1999). Whole Earth Geophysics. An Introductory Textbook for Geologists and
Geophysicists. Prentice Hall, New Jersey.
Lowrie, W. (2007). Fundamentals of Geophysics (2nd edn). Cambridge University Press,
Cambridge.
Lowrie, W. (2011). A Students Guide to Geophysical Equations. Cambridge University
Press, Cambridge.
Ofcer, C.B. (1974). Introduction to Theoretical Geophysics. Springer, New York.
Poirier, J. P. (2000). Introduction to the Physics of the Earths Interior. (2nd edn).
Cambridge University Press, Cambridge.
Schick, R. and G. Schneider (1973). Physik des Erdkrpers. Eine Einfrung fr
Naturwissenschaer und Ingenieure. Ferdinand Enke, Stuttgart.
Sleep, N. H. and K. Fujita (1997). Principles of Geophysics. Blackwell Science, London.
Stacey, F. D. (1992). Physics of the Earth (3rd edn). Brookeld Press, Brisbane.
Udas, A. and J. Mezcua (1997). Fundamentos de Geofsica. Alianza Universidad, Madrid.

Gravimetry
Blakely, R. J. (1995). Potential Theory in Gravity and Magnetic Applications. Cambridge
University Press, Cambridge.
Bomford, G. (1980). Geodesy (4th edn). Clarendon Press, Oxford.
Heiskanen, W.A. and H. Moritz (1985). Physical Geodesy. Freeman, San Francisco.
Hofmann-Wellenhof, B. and H. Moritz. (2006). Physical Geodesy. Springer, New York.
Lambeck, K. (1988). Geophysical Geodesy. Clarendon Press. Oxford.
352

353

Bibliography

Levallois, J.J. (1970). Godesie Gnrale (Vol. 3, Le Champ de la Pesanteur). Edition


Eyrolles, Paris.
Pick, M., J. Picha, and V. Vyskocil (1973). Theory of the Earths Gravity Field. Elsevier,
Amsterdam.
Torge, W. (1989). Gravimetry. Walter Gruyter, Berln.

Geomagnetism
Backus, G., R. Ladislav, and C. Constable (1996). Foundations of Geomagnetism.
Cambridge University Press, Cambridge.
Basavaiah, N. (2010). Geomagnetism: Solid Earth and Atmospheric Perspectives.
Springer, Berlin.
Blakely, R. J. (1995). Potential Theory in Gravity and Magnetic Applications. Cambridge
University Press, Cambridge.
Campbell, W.H. (2003). Introduction to Geomagnetic Fields. Cambridge University Press,
Cambridge.
Campbell, W.H. (2001). Earth Magnetism: A Guided Tour through Magnetic Fields.
Cambridge University Press, Cambridge.
Delcourt, J.J. (1990). Magntisme Terrestre. Introduction. Masson, Paris.
Jacobs, J.A. (19871991). Geomagnetism (4 volumes). Academic Press, London.
Miguel, L. de (1980). Geomagnetismo. Instituto Geogrco Nacional, Madrid.
Parkinson, W.D. (1983). Introduction to Geomagnetism. Scottish Academic Press, Edinburgh.
Rikitake, T. and Y. Honkura (1985). Solid Earth Geomagnetism. Terra Scientic
Publishing, Tokyo.
Wait, J.R. (1982). Geo-electromagnetism. Academic Press, New York.

Seismology
Aki, K. and P.G. Richards (2002). Quantitative Seismology. (2nd edn). University Science
Books, Sausalito, California.
Bullen, K.E. and B.A. Bolt (1985). An Introduction to the Theory of Seismology (4th edn).
Cambridge University Press, Cambridge.
Chapman, C. (2004). Fundamentals of Seismic Wave Propagation. Cambridge University
Press, Cambridge.
Dahlen, F.A. and J. Tromp (1998). Theoretical Global Seismology. Princeton University
Press, Princeton
Gubbins, D. (1990). Seismology and Plate Tectonics. Cambridge University Press,
Cambridge.
Lay, T. and T.C. Wallace (1995). Modern Global Seismology. Academic Press, San Diego.
Kennnett, B.L.N. (2001). The Seismic Waveeld. Cambridge University Press, Cambridge.
Pujol, J. (2003). Elastic Wave Propagation and Generation in Seismology. Cambridge
University Press, Cambridge.
Shearer, P.M. (1999). Introduction to Seismology. Cambridge University Press, Cambridge.
Stein, S. and M. Wyssesion (2003). An Introduction to Seismology, Earthquakes and Earth
Structure. Blackwell Publishers, Oxford.
Udas, A. (1999). Principles of Seismology. Cambridge University Press, Cambridge.

354

Bibliography

Heat ow and geochronology


Bunterbarth, G. (1984). Geothermics. Springer, Berlin.
Faure, G. (1986). Principles of Isotope Geology. John Wiley, New York.
Jaupart, C. and J.-C. Mareschal (2011). Heat Generation and Transport in the Earth.
Cambridge University Press, Cambridge.