1

1
2.1. GENERAL SOLUTION TO WAVE EQUATION
1.138J/2.062J/18.376J, WAVE PROPAGATION
Fall, 2006 MIT
Notes by C. C. Mei
CHAPTER TWO
ONE DIMENSIONAL WAVES IN SIMPLE SYSTEMS
General solution to wave equation
It is easy to verify by direct substitution that the most general solution of the one
dimensional wave equation:

t
2
=c
2

x
2
(1.1)
can be solved by
(x,t) = f(xct) +g(x+ct) (1.2)
wheref() and g()arearbitraryfunctionsof. In the x, t(space,time)planef(xct)
is constant along the straight line xct = constant. Thus to the observer (x, t) who
moves at the steady speed c along the positivwe x-axis., the function f is stationary.
Thus to an observer moving from left to right at the speed c, the signal described
initially by f(x) at t = 0 remains unchanged in form as t increases, i.e., f propagates
to the right at the speed c. Similarly g propagates to the left at the speed c. The lines
xct=constantandx+ct=contantarecalled thecharacteristiccurves(lines)along
which signals propagate. Note that another way of writing (1.2) is
(x, t) = F(tx/c) + G(t+x/c) (1.3)
Let us illustrates an application.
Branching of arteries
References: Y C Fung : Biomechanics, Circulation. Springer1997
M.J. Lighthill : Waves in Fluids, Cambridge 1978.
2
2 2.2 BRANCHING OF ARTERIES
Recall the governing equations for pressureand velocity

2
p
2

2
p
=c (2.1)
t
2
x
2

2
u
2

2
u
=c (2.2)
t
2
x
2
The two are related by the momentum equation
u p
= (2.3)
t x
The general solutions are :
p=p
+
(xct) +p

(x+ct) (2.4)
u=u
+
(xct) + u

(x+ct) (2.5)
Since
p

x
=p
+
+p

,
and

=cu

+cu

t
+
where primes indicated ordinary dierentiation with repect to the argument. Equation
(2.3) can be satised if
p
+
=cu
+
, p

=cu

(2.6)
Denote the discharge by Q=uA then
ZQ

=u

A=p

(2.7)
where
p

c
Z = = (2.8)
Q

A
is the ratio of pressure to ux rate and is call the impedance. It is the property of the
tube.
Now we examine the eects of branching; Referening to gure x, the parent tubes
branches into two charaterized by wave speeds C
1
and C
2
and impedaces Z
1
and Z
2
.
An incidentwave approaching the junction will cause reection in the same tube
p=p
i
(tx/c) + p
r
(t+x/c) (2.9)
_ _
2.2. WAVES DUE TO INITIAL DISTURBANCES 4
and transmitted wavesinthebranchesarep
1
(t x/c
1
) and p
2
(t x/c
2
). Atthejunction
x=0 we expect the continuity of pressure and uxes, hence
p
i
(t) +p
r
(t) = p
1
(t) = p
2
(t) (2.10)
p
i
p
r
p
1
p
2
= + (2.11)
Z Z
1
Z
2
DenethereectioncoecientRtobetheamplituderatioofreectedwavetoincident
wave, then
1

1
+
1
p
r
(t) Z Z
1
Z
2
R= = _ _ (2.12)
p
i
(t) 1
+
1
+
1
Z Z
1
Z
2
Similarly the tranmission coecients are
p
1
(t) p
2
(t)
Z
2
T = = = _ _ (2.13)
p
i
(t) p
i
(t) 1
+
1
+
1
Z Z
1
Z
2
Notethatbothcoecientsareconstantsdependingonlyontheimpedances. Hencethe
transmitted wavesaresimilarinformtotheincidentwavesexceptsmallerbythefactor
T. The total wave on the incidence side is however very dierent.
3 Waves in an innite domain due to initial distur-
bances
Recall the governing equation for one-dimensional waves in a taut string

2
u
2

2
u
c = 0, < x < . (3.1)
t
2
x
2
Let the initial transverse displacement and velocity be given along the entire string
u(x,0)=f(x) (3.2)
u
(x, t) = g(x), (3.3)
t
wheref(x) and g(x)arenon-zeroonlyinthenitedomainofx. Atinnitiesx, u
andu/tarezeroforanynitet. Theseconditionsarebestdisplayedinthespace-time
diagram as shown in Figure 1.
_
_
_
5 2.2. WAVES DUE TO INITIAL DISTURBANCES
t
2
u =c u
tt xx
x
u=f(x) u =g(x)
t
Figure 1: Summary of the initial-boundary-value problem
In (3.1.1) the highest time derivative is of the second order and initial data are
prescribed for u and u/t. Initial conditions that specify all derivatives of all orders
lessthanthehighestinthedierentialequationarecalledtheCauchyinitialconditions.
Recall that the general solution is
u=() + () = (x+ct) + (xct), (3.4)
where and are so far arbitrary functions of the characteristic variables = x+ct
and =xctrespectively.
From the initial conditions we get
u(x,0) = (x) +(x) = f(x)
u
(x,0) = c

(x)c

(x) = g(x). (3.5)

t
The last equation may be integrated with respect tox
1
x
= g(x

)dx

+K, (3.6)
c
x
0
where K is an arbitrary constant. Now and can be solved from (3.1.6) and (3.1.7)
as functions of x,
1 1
x
(x) = [f(x) + K] + g(x

)dx

2 2c
x
0
1 1
x
(x) = [f(x)K] g(x

)dx

,
2 2c
x
0
_
_
6 2.2. WAVES DUE TO INITIAL DISTURBANCES
x,t
c c
c c
x
(
0
influence
Range of
1 1
1 1
)
x-ct
x+ct x
t
Domain of
dependence
Figure 2: Domain of dependenceand range of inuence
where K and x
0
are some constants. Replacing the arguments of by x+ct and of
by xct and substituting the results inu, we get
1 1
_
xct
u(x, t) = f(xct) g dx

2 2c
x
0
1 1
x+ct
+ f(x+ct) + g dx

2 2c
x
0
1 1
x+ct
= [f(xct) +f(x+ct)]+ g(x

)dx

, (3.7)
2 2c
xct
which is dAlemberts solution to the homogeneous wave equation subject to general
Cauchy initial conditions.
Toseethephysicalmeaning,letusdrawinthespace-timediagramatriangleformed
by two characteristic lines passing through the observer at x, t, as shown in Figure 2.
Thebaseofthetrianglealongtheinitialaxist=0beginsatxctand endsatx+ct.
The solution (3.1.9) depends on the initial displacement at just the two corners xct
and x+ct, and on the initial velocity only along the segment from xct to x+ct.
Nothing outside the triangle matters. Therefore, to the observer at x, t, the domain
of dependence is the base of the characteristic triangle formed by two characteristics
passingthroughx, t. Ontheotherhand,thedataatanypointxontheinitiallinet= 0
mustinuenceallobserversinthewedgeformedbytwocharacteristicsdrawnfromx,0
into the region of t > 0; this characteristic wedge is called the range of inuence.
Let us illustrate the physical eects of initial displacement and velocity separately.
_
_
7 2.2. WAVES DUE TO INITIAL DISTURBANCES
u, t
x
O
Figure 3: Waves due to initial displacement
Case (i): Initial displacement only: f(x),= 0 and g(x)=0. The solution is
1 1
u(x, t) = f(xct) + f(x+ct)
2 2
and is shown for a simple f(x) in Figure 3 at successive time steps. Clearly, the initial
disturbanceissplitintotwoequalwavespropagatinginoppositedirectionsatthespeed
c. Theoutgoingwavespreservetheinitialprole,althoughtheiramplitudesarereduced
by half.
Case(ii): Initialvelocityonly: f(x) = 0,andg(x),=0. Considerthesimpleexample
where
g(x) = g
0
when [x[< b, and
= 0 when [x[>0.
ReferringtoFigure4,wedividethextdiagramintosixregionsbythecharacteristics
withBandClyingonthexaxisatx=band +b,respectively. Thesolutioninvarious
regions is:
u= 0
in the wedge ABE;
u=
1
x+ct
g
0
dx

=
g
o
(x+ct+b)
2c
b
2c
in the strip EBIF;
1
x+ct
u= g
o
dx

=g
o
t
2c
xct
8
4
3.2. REFLECTION FROM THE FIXED END
u, t
I
E F G
H
A B
O
C D
x
Figure 4: Waves due to initial velocity
in the triangle BCI;
u=
1
2c
_
b
b
g
0
dx

=
g
o
b
c
in the wedge FIG;
u=
1
2c
_
b
xct
g
0
dx

=
g
o
2c
(bx+ct)
in the strip GICH; and
u= 0
in the wedge HCD. The spatial variation of u is plotted for several instants in Figure
4. Note that the wave fronts in both directions advance at the speed c. In contrast to
Case (i), disturbance persists for all time in the region between the two fronts.
Reection from the xed end of a string
LetususethedAlembertsolutiontoaprobleminahalfinnitedomainx > 0. Consider
a long and taut string stretched fromx=0 to innity. How do disturbances generated
near the left end propagate as the result of initial displacement and velocity?
At the left boundaryx=0 must now add the condition
u= 0, x = 0, t > 0. (4.1)
In the space-time diagram let us draw two characteristics passing through x, t. For
an observer in the region x > ct, the characteristic triangle does not intersect the time
_
_ _
_ _
_
3.4. FORCED WAVES IN AN INFINTE DOMAIN 9
axisbecausetisstilltoosmall. Theobserverdoesnotfeelthepresenceofthexedend
at x=0, hence the solution (3.1.9) for an innitely long string applies,
1 1
x+ct
u= [f(x+ct) + f(xct)]+ g()d, x > ct. (4.2)
2 2c
xct
But for x < ct, this result is no longer valid. To ensure that the boundary condition
is satised we employ the idea of mirror reection. Consider a ctitious extension of
the string to < x 0. If on the side x < 0 the initial data are imposed such that
f(x) = f(x), g(x) = g(x), then u(0, t) = 0 is assured by symmetry. We now
have initial conditions stated over the entirex axis
u(x,0)=F(x) and u
t
(x, 0)=G(x) < x < ,
where
_
_
f(x) if x > 0
F(x) =
_
f(x) if x < 0
_
_
g(x) if x > 0
G(x) =
_
g(x) if x < 0.
These conditions are summarized in Figure 5. Hence the solution for 0< x < ct is
1 1
_
0
_
x+ct
u = [F(x+ct) + F(xct)] + + G(x

)dx

2 2c
xct 0
1 1
_
0
_
x+ct
= [f(x+ct)f(ctx)] + + g(x

)dx

2 2c
ctx 0
=
1
[f(x+ct)f(ctx)] +
1
ct+x
g(x

)dx

. (4.3)
2 2c
ctx
ThedomainofdependenceisshownbythehatchedsegmentonthexaxisinFigure6.
5 Forced waves in an innite domain
Consider the inhomogeneous wave equation

2
u
2
u
=c
2
+h(x, t) t > 0, [x[<,
t
2
x
2
10 3.4. FORCED WAVES IN AN INFINTE DOMAIN
t
u=0 2
u =c u
tt xx
x
u=-f(-x) u=f(x)
u =-g(-x) u =g(x)
t t
Figure 5: Initial-boundary-value problem and the mirror reection
x+ct=x +ct
0 0
t
x-ct=x -ct
0 0
(x ,t
0
)
0
x
( x -ct ,0) ct -x ,0) ( x +ct ,0)
0 0
O
(
0 0 0 0
Figure 6: Reection from a xed end
_ _
_

11 3.4. FORCED WAVES IN AN INFINTE DOMAIN

where h(x, t) represents forcing. Because of linearity, we can treat the eects of initial
dataseparately. Letusthereforefocusattentiononlytotheeectsofpersistentforcing
and let the initial data be zero,
u
u(x,0)=0, = 0, (5.1)
t
t=0
The boundary conditions are
u0, [x[ . (5.2)

Let the Fourier transform of any functionf(x) and its inverse f() be dened by
_

() = f(x)e
ix
dx (5.3)

1
_

f(x) = f

()e
ix
d (5.4)
2

Thetransformedwaveequationisnowanordinarydierentialequationforthetransform
of u(x, t), i.e., u(, t),
d
2
u
+c
2

2
u =h

t > 0
dt
2
whereh

(, t)denotesthetransformoftheforcingfunction. Theinitialconditionsfor u
are:

du(,0)
u(,0)=f() = 0, = g() = 0
dt
Let us hide the parametric dependence on for the time being. The general solution
to the the inhomogeneous second-order ordinary dierential equation is
t
h()
u =C
1
u
1
(t) +C
2
u
2
(t) + [ u
1
() u
2
(t)u
2
() u
2
(t)]d, (5.5)
W
0
where u
1
and u
2
are the homogeneous solutions
u
1
=e
ict
u
2
=e
ict
and W is the Wronskian
W = u
1
u
2

u
2
u
1

= 2ic=constant.
The two initial conditions require thatC
1
=C
2
=0, hence the Fourier transform is
t
u =
h(, )
_
e
ic(t)
e
ic(t)

d. (5.6)
2ic
0
_ _ _
_ _
_
_ _
2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING 12
The inverse transform is
1
_
t
1
_

(, )
_
i(x+c(t))
e
i(xc(t))

u(x, t) = d e d (5.7)
2
0
2c

i
The integrand can be written as an integral:
1
t
1
x+c(t)
u= d dh

(, ) de
i
2
0
2c
xc(t)
By interchanging the order of integration
u=
1
_
t
d
_
x+c(t)
d
1
_

de
i
h

(, )
2c
0 xc(t)
2

and using the denition of Fourier inversion, we get

1
t x+c(t)
u(x, t) = d d h(, ) (5.8)
2c
0 xc(t)
Theright-hand sideistheintegrationofhoverthecharacteristictriangledenedby
the two characteristics passing through (x, t). in the xt plane. Thus the observer is
aected only by the forcing inside the characteristic triangle.
For non-zero initial data u(x,0)=f(x) and u
t
(x,0)=g(x), we get by linear super-
position the full solution of DAlambert
1 1
x+ct
u(x, t) = [f(x+ct) +f(xct)]+ dg()
2 2c
xct
1
t x+c(t)
+ d d h(, ), (5.9)
2c
0 xc(t)
The domain of dependence is entirely within the characteristic triangle.
6 String embeded in an elastic surounding
Reference : Gra : Waves in Elastic Solids
If the lateral motion of the string is restrained by elastic springs along the entire
length,thegoverningequationcanbefound from1.1byreplacingtheexternalpressure
by the elastic restoring force KV per unit length,

2
V
2
V
=T KV (6.1)
t
2
x
2
_ _
13 2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING
which can be written as
1
2
V
2
V K
c
2
o
t
2
=
x
2

T
V (6.2)
where

T
c
o
= (6.3)

6.1 Monochromatic waves

For any linearized wave problem, if the range of the spatial corrdinate x is (,),
andall coecients are independentofx, t, thenthersttaskis to examine the physics
of sinusoidal wave train of the form:
V(x, t) = [A[ cos(kxit
A
) = ' Ae
ikxit
(6.4)
whereA=[A[e
i
A
isacomplexnumberwithmagnitude[A[and phaseangle
A
. After
examining the physical meaning of this special type of waves, it is possible to use the
principle of superposition to construct more general solutions. It is customary to omit
the symbol '= the real part of for the sake of brevity, i.e.,
V(x, t) = Ae
ikxit
(6.5)
First of all a few denitions about sinusoidal waves in general. We shall call
(x, t) = kxt (6.6)
thewave phase. Clearlythetrigonometricfunctionisperiodicinphasewiththeperiod
2. In the x, tplane,V hasaconstantvaluealongalineofcontantphase. Inparticular,
= 2n,(n= 0,1,2, ) correspond to the wave crests where V =[A[ is the greatest.
On the other hand, = (2n+ 1),(n = 0,1,2, ) correspond to the wave troughs
where V = [A[ is the smallest. [A[ is half of the separation between adjacent crests
and troughs and is called the wave amplitude; we also call A the complex amplitude.

Clearly
x
represents the number of phase lines per unit distance, i.e., the density of
phase lines, at a given instant; it is called thewavenumber,

wavenumber=k= (6.7)
x

_ _
_ _
14 2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING
On the other hand

represensts the number of phase lines passing across a xed x

t
per unit time; it is called the wave frequency.

wave frequency == (6.8)

t
To stay with a particular line of contantphase, say a crest, one must have
d=kdxdt= 0
namely one must move at thephase velocity,
dx

c= = (6.9)
dt
=constant
k
Now back to the string. Substituting(6.10) in (6.1) we get
K
2
k
2
+ V =0 (6.10)
T c
2
o
or
_
K
=c
o
k
2
+ (6.11)
T
The phase speed is

_
K
_
1/2
c= =c
o
1 + > c
o
(6.12)
[k[ Tk
2
See gure 6.1. Note that, due the stiening by the lateral support, the phase speed
is always greater than c
o
, and decreases monotonically with the wave number. Longer
waves(small[k[)arefaster,whileshorterwaves(larger[k[)areslower. As[k[increases,
capproachesthenitelimitc
o
fortheshortestwaves. Ingeneralasinusoidalwavewhose
phase velocity depends on the wavelength, i.e., is a nonlinear function of k, is called
a dispersive wave, and (6.11) or its equivalent (6.20) is called thedispersion relation.
An interesting physical feature for dispersive waves in general can be found by su-
perposing two trains of sighltly dierent frequencies and wave numbers:
V =A e
i(k
+
x
+
t)
+Ae
i(k

t)
(6.13)
where
k

=kk

=(k

), k

k. (6.14)
_ _
15 2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING
Figure 7: Phase and group velocities of a string in elastic surrounding. Point of Sta-
tionary phasek
0
.
Let us approximate

by
d

=(k)k

+O(k
2
)
dk
then
V =Ae
ikxit
e
ik

(x(d/dk)t)
+e
ik

(x(d/dk)t)
=/(x, t)e
ikxit
(6.15)
where
/= 2 cos k

(x(d/dk)t) (6.16)
Thefactorexp(ikxit)iscalled thecarrier waveand/(x, t) the envelope. Thus the
result is a sinusoidal wave train with a slowly varying envelope which has a very long
wavelength 2/k

2/kand moves at thegroup velocity

d
group velocity=c
g
= (6.17)
dk
which is in general dierent from the phase velocity for dispersive waves.
Inourstringproblem,thegroupvelocityiseasilyfound fromthedispersionrelation
(6.11)
c
o
k c
o
2
Tk c
g
c
o
2
c
g
= _ = = , hence = <1 (6.18)
k
2
+K/T
c c c
2
_ _
16 2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING
Thus the group velocity is always smaller than the phase velocity, and increases with
the wavenumber from 0 for the longest wave to the nite limit c
o
(equal to the phase
velocity) for the shortest waves.
When >
c
where
_
K

c
=c
o
. (6.19)
T
is called the cuto frequency
c
, k is real
_ _
1/2

2
K
k=[k[, with [k[= (6.20)
c
2
T
o
being the real and positive square root. V (x, t) is a propagating wave, with the plus
(minus)signcorrespondingtoright- (left-)goingwave. If <
c
,k=i isimaginary,
with beingthepostiveroot:
_ _
1/2
K
2
= (6.21)
T c
2
o
Then e
ikx
= e
x
. For boundedness one must choose the minus (plus) sign for x > 0
(x < 0). Oscillations are localized orevanecent;there is no wave radiation.
As a simple application, (6.4) is the response in a semi-innite string forced to
oscillate at the left end x = 0,
V (0, t) = 'Ae
it
. (6.22)
If >
c
, then
V (x, t) = ' Ae
i|k|xit
(6.23)
where [k[ is dened by (6.20). The requirement that waves due to a local disturbance
can only radiate outwards is called the radiation condition. More will be said on it in
AppendixB.
If <
c
, we must require boundedness at innity so that
V (x, t) = 'Ae
xit
, x > 0 (6.24)
At the cuto frequency, k = 0, V is constant in x; the innitely long string would
oscillate in unison. This absurd result signes the breakdown of the the linearized
theory.

_ _ _ _
_
_ _
_ _
2.6. STRING EMBEDDED IN AN ELASTIC SUROUNDING 17
6.2 Energy transport
Along a unit length of the string the kinetic energy density is,
_ _
2
V
KE= (6.25)
2 t
The potential energy in any segment dx of the string is the work needed to deform it
from static equilibrium. The part due to lengthening of the string against tension is
_ _
2
_ _
2
V T V
T(dsdx) = T 1 + dxTdx dx
x 2 x
Adding the part against the springs, the total potential energy per unit length is
_ _
2
T V K
PE= + V
2
(6.26)
2 x 2
Wenowcalculatetheirtimeaverages. Iftwotime-harmonicfunctionsarewritteninthe
complex form:
a=' Ae
it
, b =' Be
it
(6.27)
the time average of their product is given by
2

t+
1 1
ab ab dt= '(AB

) = '(A

B) (6.28)
2
t
2 2
Using this recipe, the period-averaged energy densities are,
KE=

'(iAe
i
)
2
=

2
[A[
2
2 4
PE=
T
'(ikAe

)
2
+
K
('Ae

)
2
=
T
k
2
+
K
[A[
2
2 2 4 4
where [k[xtis the wave phase. Since
T K

2
= k
2
+
[rho T
we have

PE=
2
[A[
2
4
Thus the total energy is equally partitioned between kinetic an potential energies, and
E =KE+PE=

2
[A[
2
(6.29)
2
_ _
_ _
2.7. DISPERSION FROM A LOCALIZED DISTURBANCE 18
Now the averaged rate of energy inux across any station x can be calculated from
the product of transverse component of the tension and the transverse velocity,
V V T Tk
T =T'(ikV)'(iV) = k[A[
2
=
2
[A[
2
=Ec
g
x t 2 2
This power input is transported from left to right by the wave. Hence the speed of
energy transport is the group velocty. This result is quite general for many physical
problems and is not limited to the springs.
7 Dispersion from a localized initial disturbance
Thesolutionformonochromaticwavesalreadyshowsthatwavesofdierentwavelengths
move at dierent velocities. What then is the consequence of an initial disturbance?
Since a general initial disturbance bounded in space can be represented by a Fourier
integral which amounts to the sum of innitely many sinusoids with a wide spectrum,
we shall employ the tools of Fourier transform.
In addition to the governing equation:
1
2
V
2
V KV
= , < x < , t > 0 (7.1)
c
2
t
2
x
2
T
we add the initial (Cauchy) conditions
V(x,0)
V(x,0)=f(x), = 0 (7.2)
t
Let us dene the Fourier transform and its inverse by
_

(k) = e
ikx
f(x)dx, f(x) =
1
e
ikx
f

(k)dk (7.3)
2

The transforms of (7.1) is
1 d
2
V

KV

=k
2
V

, t > 0 (7.4)
c
2
dt
2
T
and of the intitial conditions are,

dV

(k,0)
V(k, 0)=f(k), = 0 (7.5)
dt
_
_ _
19 2.7. DISPERSION FROM A LOCALIZED DISTURBANCE
The solution for the tranform is

V(k, t) = f(k) cos t (7.6)
where
_
K
=c
o
k
2
+ (7.7)
T
The Fourier inversion is
1

V(x, t) = dkf

(k)e
ikx
cost (7.8)
2

Anyrealfunctionf(x)canbeexpressed asthesumofanevenand anodd function

ofx. Forsimplicityletusassumethatf(x) is even in xsothatf(k)isrealand evenin

k, then
1
_

V(x, t) = dkf

(k) cos kxcost

0
which can be manipulated to
V(x, t) =
1
'
_

dkf

(k) e
ikxit
+e
ikx+it
dk (7.9)
2
0
The rst term in the integand represents the right-going wave while the second, left-
going. Eachpartcorrepondstoasuperpositionofsinusoidalwavetrainsovertheentire

rangeofwavenumbers,withinthesmallrange(k, k+dk)theamplitudeisf(k)/2. The

function f(k)/2 is called the Fourier amplitude spectrum. In general explicit evalua-
tion of the Fourier integrals is not feasible. We shall therefore only seek approximate
information. Themethod of stationary phaseisparticularlyusefulhere. Itaimsatthe
asymptotic approximation of the integral
_
b
I(t) = F(k)e
it(k)
dk (7.10)
a
forlarget. Letusrstgiveaquickderivationofthemathematicalresult. Assumethat
F(k), (k) are ordinary functions of k. If t is large, then as k increases along the path
of integration both the real and imaginary parts of the exponential function
cos(t(k))+isin(t(k))
oscillatesrapidlybetween-1to+1,unlessthereisapointofstationaryphasek
o
within
(a,b) so that
d(k
o
)
=

(k
o
) = 0, a < k
o
< b. (7.11)
dk
_ _
_ _
2.7. DISPERSION FROM A LOCALIZED DISTURBANCE 20
Figure 8: Neighborhood of Stationary Phase
ThenimportantcontributiontotheFourierintegralcomesonlyfromtheneighborhood
of k
o
. Near the point of stationary phase, we approximate the phase by
(k) = (k
o
) +
1
(kk
o
)
2

(k
o
) +
2
and the integral by
_
b
I(t) F(k
o
)e
it(ko)
exp
it
(kk
o
)
2

(k
o
) dk
2
a
With an error of O(1/t), we also replace the limits of the last integral by ; the
justication is omitted here. Now it is known that
_

_
e
itk
2
dk=

e
i/4
t

It follows that
I(t) =
_
b
F(k)e
it(k)
dk F(k
o
)e
it(ko)i/4
_
2
_
1/2
+O
_
1
_
, if k
o
(a, b),
t[

(k
o
)[ t
a
(7.12)
where the sign is + (or ) if

(k
o
) is positive (or negative). It can be shown that if
thereis no stationary point in the range(a,b), thenthe integralI(t) is small
1
I(t) = O , if k
o
/ (a, b). (7.13)
t
_
_ _
_ _

21 2.7. DISPERSION FROM A LOCALIZED DISTURBANCE

Let us apply this result to the right-going wave
it(kx/ti)
dk V
+
(x, t) =
1
'
_

dkf

(k)e (7.14)
4
0
For a xed x/t=constant, we have
K
(k) = kx/t(k), with x/t=, =c
o
k
2
+ (7.15)
T
For an observer travelling at the speed x/t less than the maximum

(k), i.e., c
o
, there
is a stationary pointk
o
at the root of
x/t=

(k
o
) =
Tk
o
= _
c
o
k
o
(7.16)
(k
o
)
k
2
+
K
o
T
Therootk
o
increasesfromzeroforlongwavestothemaximumc
o
forveryshortwaves.
If the observer is faster than c
o
, there is no stationary point.
Since

(k
o
) =

(k
o
)<0, (7.17)
the nal result is that
it(ko)i/4
V
+
(x, t)
2
1

' f

(k
o
)
t
2

k
o
)
e
=
1
' f

(k
o
)
2
e
ikox(ko)ti/4
(7.18)
2 t

(k
o
)

Thetransformf(k)dependsonthespecicproleoftheinitialdisturbance. Forexam-
ple, if
Sb
f(x) = (7.19)
(x
2
+b
2
)
which has an area S and characteristic width b, the Fourier transform is
f

(k) = Se
|k|b
(7.20)
To a specic observer identied by the speed of travel x/t < c
o
, the approximate
result can be viewed as a simple harmonic wave train with wave number k
o
, frequency
(k
o
) and amplitude

f(k
o
) 2
/= (7.21)
2 t

(k
o
)
22 2.7. DISPERSION FROM A LOCALIZED DISTURBANCE
Figure 9: Snapshotsof wave eld on the sidex > 0.
Since

(k) increases from 0 with increasing k to the nite maximum T/c

o
= c
o
, an
observer faster than c
o
sees no waves. However any observer slower than c
o
is accom-
panied by a train of progressive sinusoidal waves. The local wavelength k
o
is such that
the group velocty matches the observers speed. The faster the observer, the shorter
the waves. If a snapshot is taken, then the shortest wave, whose phase velocity is the
lowestareseenatthefrontwhichmovesasfastasthecrestsoftheshortestwaves. The

crestsoflongerwavesadvancemuchfasterthanthelocalenvelope. Becausef(k) is the

greatest at k=0, the envelope of the longest waves which stays near the source, is the
biggest. Theenvelopeoftheshortestwavesislowerinamplitudeandspreadsoutwith
the wave front. The entire disturbance attenuates in time as t
1/2
. See gure 7 for an
overview.
Note also that very near the wave front, c
g
c
o
; the second derivative

(k
o
) =

(k
o
)vanishes. Hencetheasymptoticformulabreaksdown. Abetterapproximation
isneeded,butisomittedhere. (SeeC.C.Mei,1989,AppliedDynamicsofOceanSurface
Waves). Finally we examine the propagation of wave energy in this transient problem.
_ _
23 2.8. SCATTERING OF SINUSOIDAL WAVES
Using (7.21) the local energy density is:
1
2
f

(k
o
)
2
E =
2

2
[/[
2
=
4t

(k
o
)
At any given t, the waves between two observers moving at slightly dierent speeds,
c
g
(k
1
) and c
g
(k
2
),i.e.,betweentwopointsx
1
/t=c
g
(k
1
) and x
2
/t=c
g
(k
2
)areessentially
simple harmonic so that the total energy is
x
2
x
2

2
(f

(k
o
))
2
dxE = dx
4t

(k
o
)
x
1
x
1
Since x=

(k
o
)t for xed t, we have
dx
=

(k
o
)dk
o
t
Now for x
2
> x
1
, k
2
> k
1
, it follows that
_
x
2
_
k
2

2
dxE = dk
o
(f

(k
o
))
2
=constant
4
x
1
k
1
Therefore the total energy between two observers moving at the local group velocity
remainsthesameforalltime. Inotherwords,wavesaretransportedbythelocalgroup
velocity even in transient dispersion.
8 Scattering of sinusoidal waves
Ifalongalongrod therearesomelocalized inhomogeneities,anincomingtrainofsinu-
soidal waves will be partly reected and partly transmitted. The scattered signals tell
us something about the scatterer. To determine the scattering properties for a known
scattereriscalled thescatteringproblem. Todeterminethescattererfromthescattering
data is called the inverse scattering problem. We shall only study the former.
Various mathematical techniques are needed for dierent cases: (i) Weak scatterers
characterized by small amplitude relative to the wavelength, or slow variation within a
wavelength,(ii)Strongscatterersiftheirdimensionsarecomparabletothewavelength.
8.1 Weak scattering
Let the long rod have a slightly nonuniform cross section,
S(x) = S
o
(1+a(x)), 1. (8.1)
_
_ _
_ _
_
_
24 2.8. SCATTERING OF SINUSOIDAL WAVES
where a(x) diminishes to zero atx . The wave equation reads

2
u
_
u
_
S(x) =E S(x) (8.2)
t
2
x x
For sinusoidal waves
u(x, t) = ' U(x)e
it
the spatial factor is governed by
d dU
S +
2
SU= 0 (8.3)
dx dx
Then the solution can be sought by the perturbation method
U(x) = U
0
(x) + U
1
(x) +
2
U
2
(x) + (8.4)
Substitutingthis into the governing equation
d d
_
ES
o
(1+a(x)) U
0
(x) + U
1
(x) +
2
U
2
(x) +
dx dx
+
2
S
o
(1+a(x)) U
0
(x) + U
1
(x) +
2
U
2
(x) + = 0 (8.5)
and equating the coecients of like powers of to zero, we get from the terms of order
O(
0
):
ES
o
d
2
U
0
+
2
S
0
U
0
= 0 (8.6)
dx
2
The solution is simply the incident wave
U
0
=Ae
ikx
, with k= /E (8.7)
At the orderO() we get
ES
o
d
2
U
1
+
2
S
0
U
1
+a(x)
_
ES
o
d
2
U
0
+
2
S
0
U
0
_
+ES
o
dadU
0
= 0
x
2
x
2
dx dx
or,
d
2
U
1
dadU
0
+k
2
U
1
= (8.8)
dx
2
dx dx
Thusthescattered waveisatmostoftheorderand isgoverned byaninhomogeneous
equation. Let us dene the fundamental solution (Greens function) by
d
2
G(x, x

)
+k
2
G(x, x

) = (xx

) (8.9)
dx
2
_ _
_ _
_ _ _
_ _ _ _ _
_ _
25 2.8. SCATTERING OF SINUSOIDAL WAVES
G outgoing waves at (8.10)
It will be shown in Appendix A that
G(x, x

) =
e
ik|xx

|
(8.11)
2ik
By straightforward dierentiation, it can be veried that
U
1
(x) =

da(x

)dU
0
(x

)
G(x, x

)dx

=
1

da(x

)dU
0
(x

)
e
ik|xx

|
dx

(8.12)

dx

dx

2ik

dx

dx

Since U
0
(x

) = e
ikx

, we have
=
A
_

da(x

)
e
ikx

e
ik|xx

|
dx

U
1
2

dx

=
A
e
ikx
_
x
da
dx

+e
ikx
_

da
e
2ikx

dx

(8.13)
2

dx

x
dx

=
A
e
ikx
a(x) + e
ikx

da
e
2ikx

dx

2 dx

x
=
A
a(x)e
ikx
+e
ikx

d
_
a(x

)e
2ikx

_
dx

a2ike
2ikx

dx

2 dx

x
_

= ikAe
ikx
a(x

)e
2ikx

dx

(8.14)
x
Far to the right we get
U
1
= 0, x . (8.15)
Thusonthetransmissionsidemodicationoftheincidentwavescanatmostbeofthe
orderO(
2
). Far to the left x ,
_

=Ae
ikx
ik a(x

)e
2ikx

dx

, x . (8.16) =U
1

Let us dene the reection coecient by

_

R=ik a(x

)e
2ikx

dx

, (8.17)

For instance ifa is a hump of unit total dimensionless area and lengthb,
b
a(x) = (8.18)
(x
2
+b
2
)
then
R=ikbe
2kb
(8.19)
26 2.8. SCATTERING OF SINUSOIDAL WAVES
Figure 10: Reection coecient given by ( 8.19).
Hence reection is small if kb is large, i.e., long and gentle obstacles are not eective
in reecting short waves . For very long waves or a short obstacle, reection is also
vanishingly small. The maximum reection coecient is R = ia
o
/2 when kb = 1/2.
Note that far to the right,U
1
(x)0 as x. The transmission vwave is
U

=U
0
+U
1
=Ae
ik
0
x
(1+O(
2
)). (8.20)
The tranmission coecient is T = 1 + O(
2
). This is consistent with the general law of
energy conservation [R[
2
+[T[
2
=1 to be proven later.
Homework: Check if the governing equation atO(
2
) for U
2
is:
ES
o
d
2
U
2
+
2
S
0
U
2
=ES
o
_
a
dadU
0

dadU
1
_
(8.21)
x
2
dx dx dx dx
Carry out the solution to nd the transmission coecient T from the amplitude of
U
0
+U
1
+
2
U
2
at x , and show that energy is conserved to the orderO(
2
).
2.8. SCATTERING OF SINUSOIDAL WAVES 27
8.2 Strong scattering of long water-waves by a shelf
Consider an ocean bottom with a step-wise variation of depth.
_

h
1
, x < a;

_
h=
h
2
, a < x < a;
(8.22)

_
h
3
=h
1
, x > a
How does the step interupt the propagation of an incident wave of unit amplitude
arriving from x ?
In each zone of contant depth (i= 1, 2, 3), the shallow water equations read:

i
u
i
+h
i
=0 (8.23)
t x
u
i

i
+g =0 (8.24)
t x
For sinusoidal waves

i
=
i
e
it
, u
i
=U
i
e
it
(8.25)
we get
U
i
i
i
+h
i
=0 (8.26)
x

i
iU
i
+g =0 (8.27)
x
or
d
2

+k
i
2

i
= 0, where k
i
= (8.28)
dx
2
gh
i
At a junction, the pressure and the uxmust be equal, hence
d
1
d
2

1
=
2
, and h
1
=h
2
, x =a; (8.29)
dx dx
d
2
d
3

2
=
3
, and h
2
=h
1
, x =a. (8.30)
dx dx
The forms of the solutions in each zone of constant depth are:

1
=e
ik
1
(x+a)
+Re
ik
1
(x+a)
, x < a; (8.31)

2
=Ae
ik
2
x
+Be
ik
2
x
, a < x < a (8.32)

3
=Te
ik
1
(xa)
, x > a (8.33)
_
28 2.8. SCATTERING OF SINUSOIDAL WAVES
Thereectionand transmissioncoecientsRandT aswellasAandBareyetunknown.
Applying the matching conditions at the left junction, we get two relations
1 + R=Ae
ik
2
a
+Be
ik
2
a
(8.34)
k
1
h
1
(1R) = k
2
h
2
(Ae
ik
2
a
Be
ik
2
a
). (8.35)
Similarly the matching conditions atx=a gives
Ae
ik
2
a
+Be
ik
2
a
=T (8.36)
k
2
h
2
(Ae
ik
2
a
Be
ik
2
a
) = k
1
h
1
T. (8.37)
These four equations can be solved to give
4s
T =
(1 +s)
2
e
2ik
2
a
(1s)
2
e
2ik
2
a
(8.38)
(1s
2
)(e
2ik
2
a
e
2ik
2
a
)
R=
(1 +s)
2
e
2ik
2
a
(1s)
2
e
2ik
2
a
(8.39)
A=
T
e
ik
2
a
(1+s) (8.40)
2
T
B = e
ik
2
a
(1s) (8.41)
2
where
_
k
1
h
1
h
1
c
1
s= = = (8.42)
k
2
h
2
h
2
c
2
The energy associated with the tranmsitted and reected waves are :
4s
2
[T[
2
=
4s
2
+ (1 s
2
)
2
sin
2
2k
2
a
(8.43)
(1s
2
) sin
2
2k
2
a
[R[
2
=
4s
2
+ (1 s
2
)
2
sin
2
2k
2
a
(8.44)
It is evident that [R[
2
+[T[
2
= 1.
Over the shelf the free surface is given by
2s (1+s)e
ik
2
(xa)
+ (1 s)e
ik
2
(xa)

2
= (8.45)
(1+s)
2
e
ik
2
a
(1s)
2
e
ik
2
a
Recallingthetimefactore
it
, we see that the free surface over the shelf consists of two
wave trains advancing in oppopsite directions. Therefore along the shelf the two waves
_
2.8. SCATTERING OF SINUSOIDAL WAVES 29
caninterfereeachotherconstructively,withthecrestsofonecoincidingwiththecrests
of the other at the same moment. At other places the interference is destructive, with
the crests of one wave train coinciding with the troughs of the other. The envelope of
energy on the shelf is given by
4s
2
cos
2
k
2
(xa) + s
2
sin
2
k
2
(xa)
[[
2
=
4s
2
+ (1 s)
2
sin
2
2k
2
a
(8.46)
At the downwave edge of the shelf, x=a, the envelope is
4s
2
[[
2
=
4s
2
+ (1 s)
2
sin
2
2k
2
a
(8.47)
Note that the reection and transmission coecients are oscillatory in k
2
a. In par-
ticulatr for 2k
2
a = n, n = 1,2,3..., that is 4a/ = n, [R[ = 0 and [T[ = 1. The
shelfistransparenttotheincidentwaves,correspoinidngtothemostconstructiveinter-
ference and the strongest transmission Mininum transmision and maximum reection
occur when 2k
2
a = (n 1/2), or 4a/ = n 1/2, when the interference is the most
destructive. The corresponding transimssion and reection coecients are
min[T[
2
=
4s
2
, max[R[
2
=
(1s
2
)
2
. (8.48)
(1+s
2
)
2
(1+s
2
)
2
See Figure 11
The features of interference can be explained physically. The incident wave train
consists of periodic crests and troughs. When one of the crests rst strikes the left
edge at x=a, part of the it is transmitted onto the shelf and part is reected towards
x . After reaching the right edge at x = a, the tranmitted crest has a part
reected to the left and re-reected by the edge x=a to the right again. When the
re-reected crest arrives at the right edge the second time, its total travel distance is
4a. If 4a is an integral multple of the wave length
2
, the crest is in phase with all
the other crests entering the shelf either before (the third, fourth, fth, ... time) or
after. Thus all the crests reinforce one another at the right edge. This is constructive
interference, leading to the strongest tranmission to the right x . On the other
hand if 2k
2
a= (n1/2) or 4a/=n1/2, some crests will be in opposite phase to
some other crests, leading to the most destructive interference at the right edge, and
smallest transmission.
30 2.9. GENERAL IDENIITIES ON SCATTRERING
0
0.2
0.4
0.6
0.8
1
Max|R|
2
Min|T|
2
0 2 4 6 8 10
s
2
0 2 4 6 8 10
2k a
2
Figure 11: Transmission and reection coecients for a rectangular shelf. From Mei,
C.C., Applied Dynamics of Ocean Surface Waves.
From the results in the following exercise, the reected wave suers a reversal of
phase at the discontinuity (a crest turns into a trough and vice versa). Hence at the
reection edge a wave crest tranmitted onto the shelf returns from x = a as a trough.
Itisthentransmitted totheleftofx =a astrouigh. Ontheotherhand,acrestthat
isreected atx =a appearsasacrest. Thereforetheinterferenceisdestructive. The
resulting R, which is the sum of multiple reections, is small.
Homework: Scattering by a depth discontinuity: Consider the scattering of long
water waves by a depth discontinuity at x = 0, i.e., h = h
1
, < x < 0 and h =
h
2
, 0< x < . Show that the tranmssion and reection coecients are:
2 1(h
2
/h
1
)
1/2
T = , R = (8.49)
1 + (h
2
/h
1
)
1/2
1 + (h
2
/h
1
)
1/2
Thusthereected wavesuersareversalofphasewhenh
2
/h
1
> 1,i.e,approachingthe
discontinuity from the shallow side.
0
0.2
0.4
0.6
0.8
1
|R|
2
= 2 / 2

= 1 / 2
= 1 / 4 s
s
s
_ _
2.9. GENERAL IDENIITIES ON SCATTRERING 31
9 Generalidentitiesinscatteringofmonochromatic
waves by arbitrary inhomogeneities
Scattering due to inhomogeneities, caused by nonuniformities either in geometry or in
materialproperties,requiresthesolutionofordinarydierentialequationswithvariable
coecients. In general one must resort to numerical means. For the sake of checking
numerical accuracy and for gaining physical insight, identities which must be true are
useful. They are often deduced by arguments typical in the derivation or the use of
Greens theorem.
We illustrate these identities by the example of an innitely long rod with variable
cross section S(x),

_
u
_
S
2
u
S = (9.1)
x x E t
2
E and are taken to be constants. S(x) is nonuniform only in a nite neighborhood
aroundx= 0. Elsewhere S =S
o
=constant. Consider monochromatic waves
u(x, t) = U(x)e
it
(9.2)
so that U(x) is governed by the ordinary dierential equation

(SU

+k
2
SU= 0, k
2
= (9.3)
E
The boundary conditions depend on the source of the incident waves.
Let us consider two problems: one with the incident wave from x ; the other
with the incident wave from x .
Iftheincidentwavesarrivefromx ,theasymptoticformofthesolutionmust
be
U
1
(x) A

e
ikx
+B

e
ikx
=A

e
ikx
+R
1
e
ikx
, kx (9.4)
and
U
1
(x) A
+
e
ikx
=A

T
1
e
ikx
kx (9.5)
whereT
1
andR
1
arethetransmissionand reectioncoecientswhichareapartofthe
unknown solution of the left-incidence problem. Let us dene the Jost function by
U
1
f
1
(x) = (9.6)
A

T
1
_ _
32 2.9. GENERAL IDENIITIES ON SCATTRERING
then
1
ikx
R
1
ikx
f
1
(x) e + e kx1 (9.7)
T
1
T
1
and
f
1
(x) e
ikx
kx1 (9.8)
Ontheotherhand,iftheincidentwavesarrivefromx ,theasymptoticformof
the solution must be
U
2
(x) b

e
ikx
=b
+
T
2
e
ikx
, kx (9.9)
and
U
2
(x) a
+
e
ikx
+b
+
e
ikx
=b
+
R
2
e
ikx
+e
ikx
, kx (9.10)
where T
2
and R
2
are the unknown transmission and reection coecients of the right-
incidence problem. Similarly we dene another Jost function
U
2
f
2
= (9.11)
b
+
T
2
so that it behaves as
f
2
(x) e
ikx
, kx (9.12)
and
f
2
(x)
1
e
ikx
+
R
2
e
ikx
, kx (9.13)
T
2
T
2
Since both f
1
and f
2
satisfy (9.3), we have
(Sf
1

+k
2
Sf
1
= 0, and (Sf
2

+k
2
Sf
2
= 0, (9.14)
Multiplyingtherstbyf
2
and thesecond byf
1
,thentakingthedierence,wegetafter
partial integration,
d
(S(f
1
f

f
2
f
1

))=0,
dx
2
implying the following Wronskian identity
W(f
1
, f
2
) = (f
1
f
2

f
2
f
1

) =
C
(9.15)
S(x)
_ _ _ _
_ _ _ _
_ _
_
_ _ _ _
33 2.9. GENERAL IDENIITIES ON SCATTRERING
TheconstantC canberelated totheasymptoticlimitsatk[x[ 1 where S S
0
. Far
on the incident side, the Wronskian is
(f
1
f
2

f
2
f
1

)
x
=
1
e
ikx
+
R
1
e
ikx
(ik)e
ikx
e
ikx
ik
e
ikx

ikR
1
e
ikx
T
1
T
1
T
1
T
1
2ik
=
T
1
Far on the transmission side, the Wronskian is
(f
1
f
2

f
2
f
1

)
x
ikx
ik
ikx
ikR
2
ikx
1
ikx
R
2
ikx ikx
=e e + e e + e (ik)e
T
2
T
2
T
2
T
2
2ik
=
T
2
Since the two Wronskians are equal, we conclude that
T
1
=T
2
(9.16)
Thus, the complex transmission coecients are equal for incident waves coming from
either side, no matter how asymmetric the geometry may be in the near eld!
Next we consider the left-incidence problem and omitting the subscript 1. Multi-
plyingtherstof(9.3)bythecomplexconjugateofU,i.e.,U

,and takingthedierence
of the result with its complex conjugate, we get
US(U

(SU

=
d
S(UU

) =0 (9.17)
dx
hence
S(UU

)=constant (9.18)
By using the asymptotic expression (9.4) on the incidence side, we get
S(UU

SU

)
x
=S
o
A

e
ikx ikx
ikA
ikx
+ikB
ikx
+B

e
S
o
_
A
ikx
+B
ikx
_ _
ikA

e
ikx
ikB

e
ikx
_

e
=2ik[A

[
2
+ 2ik[B

[
2
_
_ _ _ _ _ _ _ _
2.10. REFRACTION IN A SLOWLY VARYING MEDIUM 34
Similarly by using the asymptotic expression (9.5) on the transmission side, we get
S(UU

)
x
=S
o
A
+
e
ikx
(ik)A

+
e
ikx
S
o
A
+

e
ikx
(ik)A
+
e
ikx
=2ikS
o
[A
+
[
2
Equating the two limits we conclude that
[A

[
2
=[B

[
2
+[A
+
[
2
(9.19)
or
[R
1
[
2
+[T
1
[
2
= 1. (9.20)
This is merely a statement of conservation ux : the ux rate of total scattered energy
(reected and transmitted) is equal to the ux rate of the incident wave energy; the
speed of energy tranport being the same on both sides of the scatterer. Clearly the
same energy conservation must hold for the right-incidence problem.
[R
2
[
2
+[T
2
[
2
=1 (9.21)
Ifthescatteringproblemissolved numericallyby,say,niteelements,itisnecesssary
thatthecomputed scatteringcoecientsRandT satisfytheidentities(9.16)(9.20)and
(9.21).
Homework: Extend theidentities(9.16)and (9.20)whenS(x)S
1
askx
and S(x)S
2
as kx, where S
1
and S
2
are two dierentconstants.
10 Refraction in a slowly varying medium
Letusconsidertime-harmoniclongwavesinshallowwaterwaves. Thegoverningequa-
tion is
d
_
d
_

2
h + =0 (10.1)
dx dx g
Consider a sea depth which varies slowly within a wavelength, i.e.,
1 dh
=O()1 (10.2)
khdx
35 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
Earlieranalysissuggeststhatreectionisnegligiblysmall. Thusthesolutionisexpected
tobealocallyprogressivewavewithboththewavenunmberand amplitudevaryingmuch
more slowly than the wave phase inx . Hence we try thesolution
=A(x)e
i(x)
(10.3)
where (x)tis the phase function and
d
k(x) = (10.4)
dx
is the local wave number and A is the complex amplitude. Note the spatial rate of
variationofthephasefunction,i.e.,thewavenumberisingneralnotsmall,therefore
itself is not a slowly varying function of x. Let us calculate the rst derivative:
d
dx
=
_
ikA+
dA
dx
_
e
i
and assume
dA
dx
=O(kL)
1
1
kA
In fact we shall assume each derivative of h, A or k is times smaller than kh, kA or
k
2
. Futhurmore,
d
_
d
_

2
_ _
dA
_
d
_
dA
_
d(khA)
2
A
_
h + = ik ikh+h + h +i + e
i
= 0
dx dx g dx dx dx dx g
Note that the complex amplitudeA can be written as
A=[A(x)[e
i
A
(x)
(10.5)
The phase of A,
A
, is a slowly varying function of x. It can be considered as a part of
the wave phase,although its spatial gradient is muchsmaller thank.
Now lets expand
A=A
0
+A
1
+A
2
+ (10.6)
with A
1
/A
0
=O(), A
2
/A
0
=O(
2
), . From O(
0
) the dispersion relation follows:

2
=ghk
2
, or k= (10.7)
gh
_
_ _
36 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
Thus the local wave number and the local depth are related to frequency according
to the well known dispersion relation for constant depth. As the depth decreases, the
wavenumber increases. Hence the local phase velocity

c= = gh (10.8)
k
also decreases.
From O() we get,
dA
0
d(khA
0
)
ikh +i = 0
dx dx
or, after multiplication byA
0
,
d
(khA
2
0
) = 0
dx
which implies
d
(kh[A
0
[
2
)=0 (10.9)
dx
hence
kh[A
0
[
2
=C
2
=constant
or,
gh[A
0
[
2
=constant= gh

[A

[
2
(10.10)
Since in shallow water the group velocity equals the phase velocity, the above result
means that the rate of energy ux is the same for all x and is consistent with the
original assumption of unidirectional propagation. Furthermore, the local amplitude
increases with depth as
A
0
(x)
_
h

_
1/4
= (10.11)
A

h
This result is called Greens law.
In summary, the leading order solution is
_ _
1/4
_ _
1/4
_ _
x
_
=A

e
iit
=A

exp i k(x

)dx

i (10.12)
h h
Homework: Study the refraction of a train of obliquely incident waves towards a
sloping beach. The sea depth varies slowly only inx, h=h(x).
_ _
_ _ _ _
37 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
A Solution for the Green function
Letusdividethedomainintotwoparts,correspondingtotwosidesoftheconcentrated
forcing. On the right side x > x
0
, we dene G(x, x
0
) G
+
(x, x
0
) which must be
governed by :
d
2
G
+k
2
G
+
= 0, x > x
0
(A.1)
dx
2
0
To satisfy the radiation condition we take the sol,ution to beTo satisfy the radiation
condition we take
G
+
=Ce
ik
0
x
(A.2)
On the left side we deneG(x, x
0
) = G

(x, x
0
), which must be governed by
d
2
G

+k
2
G

= 0, x < x
0
(A.3)
dx
2
0
To satisfy the radiation condition we take
G
+
=De
ik
0
x
(A.4)
At x=x
0
we require continuity:
G
+
(x
0
, x
0
) = G

(x
)
, x
)
) (A.5)
Hence
Ce
ikox
0
=De
ik
0
x
0
(A.6)
In addition we get another matching condition at x=x
0
by integrating eq(8.8) across
the concentrated forcing
_
x
0
+
_
x
0
+
d
2
G
+k
0
2
G dx= (xx
0
)dx= 1.
dx
2
x
0
x
0

The left-hand side can be integrated to give the second matching codition,
dG
+
dG

= 1 (A.7)
dx dx
x
0
x
0
implying that
ik
0
Ce
ik
0
x
0
+ik
0
De
ik
0
xo
= 1 (A.8)
_ _
2.10. REFRACTION IN A SLOWLY VARYING MEDIUM 38
Thus
ik
0
x
0
ik
0
x
0
e e
C
0
= , D
0
= , (A.9)
2ik
0
2ik
0
The nal solution is
ik
0
(xx
0
) ik
0
(xx
0
)
e e
G
+
= , x x
0
>0; G

= , x x
0
<0.
2ik
0
2ik
0
or more compactly,
e
ik
0
|xx
0
|
G(x, x
0
) = (A.10)
2ik
0
More on one-dimensional waves will be discussed later.
B On the radiation condition
B.1 Forms of the radiation condition
In the theory of diraction and/or radiation of time-harmonic waves, an important
boundary condition is the so-calledradiation conditiondue to A. Sommerefeld :
The disturbance due to either diraction by a nite scatterer, or radiation from a nite
region, can only behave as outgoing waves at innity.
Let us rst write the unknown as
(x, t) = (x)e
it
(B.1)
In a scattering (diraction) problem we let
=
I
+

(B.2)
where
I
represents the incident wave (e.g.,
I
= A
o
e
ikx
), and

denotes the distur-

bance (scattered wave). In a radiation problem,

represents the radiated wave due to

motion in the nite region. The radiation condition can be stated mathematically in
two equivalent forms, as follows.
1. One-dimensional propagation:

(x)A
0
B

e
ikx
, as kx; (B.3)
or,

lim ik

0 (B.4)
kx x
_ _
_ _
39 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
2. Two-dimensional propagation:
ikr

(r, , z)A
0
B(, z)
e
, as kr; (B.5)
kr
or,

lim kr ik

0 (B.6)
kr r
3. Three-dimensional propagation:
ikr

(r, , )A
0
(, )
e
, as kr; (B.7)
kr
or,

lim kr ik

0 (B.8)
kr r
Note that the radiation condition is meant only for waves that is strictly sinusoidal in
timefromancientpast. Fortransientdisturbances,theconditionshouldbechangedto

(x, t)0, as [x[ , t < . (B.9)

Physically. ifthesourceofthewavedisturbanceisinaniteregion,itisnotpossible
for wave energy to ow inward from innity. Mathematically, a sinuoidal wave is the
limit of a transient wave where the local and sinusouidal forcing begins at some time
in the past, say t = 0. For all nite t, the boundary condition (B.9) holds. In order
to derive the behavior at very large time (the quasi steady state), one can solve an
initial-bounday value problem under the initial conditions,

, = 0, t = 0, x (B.10)
t
and the additional boundary condition in the source region, e.g.,

it
=n Ve , t > 0. (B.11)
n
and thentakethelimitoflarget. TheexplicitsolutionfortheIBVPisnottrivial. We
shall demonstrate for a one-dimensional example where the local energy supply grows
slowly in comparison with the wave period.
2.10. REFRACTION IN A SLOWLY VARYING MEDIUM 40
B.2 Transient motion of an elastically supported string
Consider a semi-inntestring,
V
tt
TV
xx
+KV = 0, , x >0, t > 0 (B.12)
subject to the boundary condition,
V(0, t) = A
0
(t)e
it
, t > 0 (B.13)
where 1 and A
0
(t) increases slowly from A
0
() = 0 to A
0
() = A

=constant
withthetimeranget
1
=O(1). Thusattheboundearythereissinusoidalforcingwith
a slowly increasing amplitude. At innity the boundary condition is
V(, t)0, t < . (B.14)
Let us use the perturbation method of mulitple scales. In vieew of (B.13), we in-
troduce two time coordinates t and t
1
= t. Suggested by the elementary example of
wave groups, we expect that the envelope to be slowly varying in space as well. Let
us therefore also introduce two space coordinates x, x
1
= x
1
. We now pretend that
V =V(x, x
1
;t, t
1
) so that the original time derivatives become, according to the chain
rule,
V V
+ , (B.15)
t t t
1

2
V
_

__
V V
_

2
V
2
V
2
V
+ + = + 2 +
2
. (B.16)
t
2
t t
1
t t
1
t
2
tt
1
t
2
1
Likewise the space derivatives become
V V
+ , (B.17)
x x x
1

2
V
_

__
V V
_

2
V
2
V
2
V
+ + = + 2 +
2
. (B.18)
x
2
x x
1
x x
1
x
2
tx
1
x
2
1
We next substitute these into (B.12)
V =V
0
(x, x
1
;t, t
1
) + V
1
(x, x
1
;t, t
1
) +
2
V(x, x
1
;t, t
1
) + (B.19)
and separate dierent orders. At order O(
0
) we have the following perturbation equa-
tion :
O(
0
):

2
V
0

2
V
0
T +KV
0
=0 (B.20)
t
2
x
2
_ _
_ _
41 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
At order O(), we get:

2
V
1

2
V
1

2
V
0

2
V
0
T +KV
1
=2 + 2T (B.21)
t
2
x
2
tt
1
xx
1
Let us take the leading-order solution to be a sinusoidal progressive wave
V
0
=A(x
1
, t
1
)e
i(kxt)
(B.22)
where
_

2
K
k= . (B.23)
T T
Let the square root be real so that there is propagation. However, we do not know the
sign of k, hence the direction of propagation.
At the orderO() Eqn. (B.21) can be written

2
V
1

2
V
1
_
A A
_
ikxit
T +KV
1
=2i kT + e . (B.24)
t
2
x
2
x
1
x
1
The forcing term on the right is a homogeneous solution of the left, hence unbounded
resonance of V
1
must occur unless
A A
kT + = 0, (B.25)
x
1
x
1
or
A Tk A
+ =0 (B.26)
t
1
k x
1
This is the condition of secularity (solvability) and is a rst order hyperbolic equation.
The general solution is
x
1
A(x
1
, t
1
) = A t
1
(B.27)
Tk/
Now we apply the boundary condition (B.13) atx
1
= 0,
A(t
1
) = A
0
(t
1
) (B.28)
Hence the envelope is
x
1
A(x
1
, t
1
) = A
0
t
1
(B.29)
Tk/
For this transient problem we expect that A(, t
1
) approaches zero when x
for nite t
1
, according to (B.14). Since for large x
1
, A(x
1
, t
1
) A
0
(x
1
/Tk/), we
must take the positive sign for k in view of the given behaviour ofA
0
. Thus
k > 0 (B.30)
42 2.10. REFRACTION IN A SLOWLY VARYING MEDIUM
implyingthatthesolutionisarightwardpropagatingwave. Thustheradiationcondition
is derived.
Note further that
Tk
=c
g
(B.31)

is the group velocity in the string, thus

A(x
1
, t
1
) = A(t
1
x
1
/c
g
) (B.32)
or,Aisaslowlyvaryingfunctionofxc
g
t. Thusresultistrueforalldispersivewaves.