ECE 6382 y

12 1
x

Fall 2023
David R. Jackson

Notes 8
Analytic Continuation

Notes are adapted from D. R. Wilton, Dept. of ECE

1
 Analytic Continuation of Functions

 We use analytic continuation to extend a function off of the real axis

and into the complex plane such that the resulting function is
analytic.
 More generally, analytic continuation extends the representation of a
function in one region of the complex plane into another region,
where the original representation may not have been valid.

For example, consider the Bessel function Jn (x).

 How do we define Jn (z) so that it is computable in some region
and agrees with Jn (x) when z is real?

2
 Analytic Continuation of Functions (cont.)

 One approach to extend the domain of a function is to use Taylor series.

 We start with a Taylor series that is valid in some region.

 We extend this to a Taylor series that is valid in another region.

Note:
This may not be the easiest way in practice, but it always works in
theory, and it illustrates the principle of analytic continuation.

3
 Analytic Continuation of Functions (cont.)
two alternative
representations
    
Example Original geometric series

f z    z n 
1
, z 1
y
n 0 1  z
But we'll pretend that we don't know the closed - form expression 
3
Rc 
2
• Expand about z  12 . Since both series are
x
valid there, find coefficients of a new series : 12 1

f  z   bm  z  12 
m
, z  12  32
m 0

by using
use the above series
representation
    
m m   n
1 d f 1 d 1   n m 
bm    z    n n  1n  2  n  m  1 z 
m! dz m z 
1 m! dz m  n0  z  1 m!  nm  z  1
2 2 2

1   n! 1 n m
  dm n 
 
m!  nm n  m !
 2   Note m z 0 , n  m 
dz
  
 The coefficients of the new series --- with extended region of convergence --- are determined
from the coefficients of the original series, even though that series did not converge in the
extended region. The information to extend the convergence region is contained in
the coefficients of the original series --- even if it was divergent in the new region!
4
 Analytic Continuation of Functions (cont.)
Example (cont.)
Another way to get the Taylor series expansion:
(This assumes that we know the closed-form expression!)


1 m 1 dm  1 
f  z    bm  z  2  , z  12  23 bm   
m 0 m! dz m  1  z  z  1/ 2

so that
m 1 m 1
1 1 1  2  2
bm  12 3 m   m!    
m! 1  z m 1 z  1 / 2
m!  3   3
 m 1
 2 m
 f z      z  12  , z  12  23
m 0  3 

1 1 2 2 3
Note : f  z   ; f z   ; f  z   ; f   z  
1 z 1  z 2 1  z 3 1  z 4
5
 Analytic Continuation of Functions (cont.)
Example (cont.)


f  z   z n
n 0

Here we show the continuation

of f  z  from its power series y
representation in the region
z  1 into the entire complex
plane using Taylor series.
12 1
x

We have a “leapfrogging” of the circles.

 If the singularities are isolated, we can continue any function

into the entire complex plane via a sequence of continuations
using Taylor and / or Laurent series !
6
 Theorem
The Zeros of an Analytic Function are Isolated
(The zeros cannot be arbitrarily close together.)
Proof of theorem :
 Assume that f ( z ) is analytic in a connected region A , and suppose f ( z0 ) 0 (simple zero).
Then f ( z ) has a Taylor series f ( z0 )  a1 ( z  z0 )  a2 ( z  z0 )2   with a0  f ( z0 ) 0 at z0 .

 More generally, we may have a zero with multiplicity N ( N  ) such that f ( z ) has
a Taylor series as :
f ( z ) aN ( z  z0 ) N  aN 1 ( z  z0 ) N 1  
 1 (N) 
( z  z0 ) N  aN  aN 1 ( z  z0 )1   a
 N  f ( z0 ) 0 
 N! 
( z  z0 ) N g ( z ), g ( z0 ) 0, and g ( z ) is analytic in A
(it is represented by a converging Taylor series).

Since g ( z ) is analytic it is also continuous. Since g ( z0 ) 0 , g ( z ) cannot vanish within a

sufficiently small neighborhood of z0 . That is, the zeros of f ( z ) must be isolated .

The only exception to having isolated zeros is if the function f (z) is

identically zero in a neighborhood of z0 (N = ).
7
 The Zeros of an Analytic Function are Isolated
Example

1
The function f ( z ) sin 1 z  has zeros at z  , n 1, 2,
n

This function cannot be analytic at z = 0 since the zeros accumulate there and
hence are not isolated there.

x
1/ 
The origin is an “accumulation point” for the zeros.

8
 Analytic Continuation Principle
Theorem of analytic continuation:
Assume that f (z) and g (z) are analytic in a connected region A, and f (zn) = g(zn)
on a set of points zn in A that converge to a point z0 in A.

Then g(z) = f (z) in A.

In other words, there can be only one function
that is analytic in A and has a defined set of
values at the converging points zn.

z0 Note:
A If f and g agree on a contour
inside A or in a region that is
zn g(z) inside of A, this will make the
functions agree everywhere in A.

Note: These points or contour can be on the real axis as a special case (e.g., a line segment).
9
 Analytic Continuation Principle

Proof of theorem
z0
A
zn g(z)

Proof:
 Construct the difference function F(z) = f (z)-g(z), which in analytic in A. This
function must have a Taylor series at z0.

 This Taylor series for F(z) has all zero coefficients and thus is zero in its
region of convergence about z0; otherwise, the function must have isolated
zeros – which it does not, by assumption.

 By continuing (“leapfrogging”) the Taylor series that has zero coefficients

(analytic continuation), the difference function must be zero throughout A.
10
 Analytic Continuation Principle (cont.)

Corollary (extending a domain from A to A  B)

Assume that f (z) is analytic in A and g(z) is analytic in B, and the two
domains overlap in a region A  B, and f (z) = g (z) in A  B.

 f  z  , z  A
Define h  z  
 g  z  , z  B
Then h(z) is the only analytic function in A  B that equals f (z) on A.

f z  g z  AB The function h(z) uniquely

extends the domain of f (z)
from A to A  B.
A
A B
f z  B
g z 

11
 Analytic Continuation Principle (cont.)

Corollary: h(z) is the only analytic function in A  B that equals f (z) on A.

Proof of corollary:
 The function h is analytic in the region A  B and also equals f (zn) on any set of
converging points in the intersection region.
 The theorem of analytic continuation thus ensures that h is unique in A  B.

Converging points

AB

A
f z  A B B
g z 

12
 Analytic Continuation Principle (cont.)
Two common ways to use analytic continuation.

y
1) Continue off of the real axis.

x
Line segment

y 2) Continue from one region (blue) to

another larger region (blue + red)
with a region of overlap (purple).

x Example:
Blue: 1st and 2nd quadrants
Red: 2nd and 3rd quadrants
Overlap 2nd quadrant
Larger region: 1st, 2nd, 3rd quadrants

13
 Analytic Continuation Principle (cont.)

Example (continuation from real axis)

The function sin (x) is continued off the real axis.

eiz  e iz
y g  z  sin  z  
2i

x
Line segment

eix  e  ix
f  x  sin  x  
2i

The function g(z) is the only one that is analytic in the blue region of the
complex plane and agrees with sin (x) on any segment of the real axis.
14
 Analytic Continuation Principle (cont.)

Example (continuation from real axis)

The Bessel function Jn(x) is continued off the real axis.

k
g  z   J n  z  

  1  z
n 2 k

 
y k 0 k! n  k !  2

x
Line segment

k
f  x  J n  x  

  1  x
n 2 k

 
k 0 k!n  k !  2 

The function g(z) is the only one that is analytic in the blue region of the
complex plane and agrees with Jn (x) on any segment of the real axis.
15
 Analytic Continuation Principle (cont.)
Example (continuation to a larger region)
The function Ln (z) (principal branch) is continued beyond a branch cut.

f  z  Ln  z  ln r   i ,    

z rei y Note: “Ln” denotes the principal branch.

Region of overlap
(second quadrant)

Branch cut for f

x
Define:
g  z  ln  z  ln r   i
  / 2    3 / 2

Branch cut for g

16
 Analytic Continuation Principle (cont.)
Example (cont.)

The function h(z) is the only function that is analytic in the first
three quadrants and agrees with Ln(z) in the second quadrant.

h  z  Ln  z  h  z  Ln  z 

h  z  Ln  z   2 i

17
 Analytic Continuation Principle (cont.)
Example (continuation to a larger region)
The function Yn (z) (Bessel function of the second kind) is continued beyond a
branch cut.
y

Branch cut
x

Ln  z  ln r   i
    

 1 n  1 n  k  1!  z 
2k  n
2   z
Yn ( z )  J n  z   Ln         
   2   k 0 k!  2
1 
1  z
2 k n  0.577216
   1   k    n  k 
k
 
 k 0 k ! n  k !  2  (Euler’s constant)

1 1 1
  p  1      ( p  0)
2 3 p
18
 Analytic Continuation Principle (cont.)
Example (cont.)
The new Bessel function using ln(z) with the new angle range is analytic within the red region.

y Note: In the third quadrant, this new Bessel

In the secondquadrant : function will not be the same as the usual
Ynnew ( z ) Yn ( z ) Bessel function there.

x
In the third quadrant :
Ynnew ( z ) Yn ( z ) New Bessel function Ynnew:
2 ln  z  ln r   i
 J n  z 2 i 

  / 2    3 / 2

 1 n  1 n  k  1!  z 
2k  n
2   z
Yn ( z )  J n  z   ln       
new
 
   2   k 0 k!  2
 2 k n
1 1  z
   1   k    n  k 
k
 
 k 0 k !n  k !  2 
19
 Analytic Continuation Principle (cont.)
Example (cont.)

The extended function h(z) is analytic within the first three quadrants.

h  z  Yn ( z ) h  z  Yn ( z )

2
h  z  Yn ( z )  J n  z 2 i 


20
 Analytic Continuation Principle (cont.)
Example

How do we extend F(x) to arbitrary z?


F  x   e  xt J 0 t  dt, x 0
0
y
Note: The integral does not converge for x < 0.

Original domain: x  0
x

Identity :

1
 J 0 t dt 
 xt
e , x 0
2
0 x 1

21
 Analytic Continuation Principle (cont.)
Example (cont.)
1
Here we define: F z  
z 
2 1/ 2
1

This is defined everywhere in the complex plane (except on the branch cuts).

z 
2 1/ 2
1  x2 1

i
x
i
Note:
The shape of the branch cuts
is arbitrary here.

22
 Analytic Continuation Principle (cont.)
Example (cont.)
1
We next use: F z  
2 Note: Re z 2  1 0
z 1
This corresponds to using vertical branch cuts.

With these branch cuts,

y
 
1/ 2
Re z 2  1 0
z 
2 1/ 2
1  x2 1
The derivation (omitted) is similar to
that of the Sommerfeld branch cuts.
i
x
 z 1
2
1/ 2
 z 2 1 i
(with these branch cuts)

23
 Analytic Continuation Principle (cont.)
Example (cont.)

Analytic continuation onto Riemann surface:

1
y F z  
z 
2 1/ 2
1

 
1/ 2
Top sheet : z 2  1  x2 1

i
x
i

Escalators

24
 Analytic Continuation Principle (cont.)
Example

How do we extend F(x) to arbitrary z?

x
1
F  x   dt (real valued for real x > 0)
Im t  t
1

Original domain: x > 0

 Re t 
1 x

F  x  ln  x   ln 1 Ln  x   Ln 1

Recall : ln  z  = Ln  z   i 2 n  Note : Ln  z  means    arg  z   

F  x  Ln  x 
25
 Analytic Continuation Principle (cont.)
Example (cont.)

Analytic continuation:

F  z  Ln  z 

(This agrees with F(x) on the real axis.)

x
Branch cut

26
 Analytic Continuation Principle (cont.)
Example (cont.)
z
From another point of view: 1
F  z   dt
1
t
Im t 

z
z C
z
 Re t 
z 1

We analytically continue F(z) from the real axis.

We require that the path is varied continuously as z leaves the real axis.
27
 Analytic Continuation Principle (cont.)
Example (cont.)
z
1
Im t  F  z   dt
1
t
 z
C
z
 Re t 
F  z  z 1
z
C z
F 
z 
As z encircles the pole at the origin, we get a different result for the function F(z).

1 1 1
z dz 2 i   z dz   z dz  F 
 z   F 
z  2 i
C C C

This is why we need a branch cut in the z plane, with a branch point at z = 0.
28
 Analytic Continuation Principle (cont.)
Example (cont.)

Analytic continuation onto Riemann surface:

F  z  ln  z 

(The function is defined so that F(z) agrees with F(x) on the real axis.)

y
F  x  Ln  x   0 
(sheet number zero)

x
Escalators

There are an infinite number of sheets!

29
 Analytic Continuation Principle (cont.)
Example (cont.)

F  z  ln  z 

x
Escalators

30
Analytic Continuation Principle (cont.)
Example
How do we interpret cosh-1(z) for arbitrary z?

 
cosh  1  x  Ln x  x 2  1 , x  1 Note : cosh  1 1 0

 
Note : Ln x  x 2  1 is real (and positive) for x  1

Original domain: x > 1

x
1 x

Recall : Ln  z  means    arg  z   

31
 Analytic Continuation Principle (cont.)
Example (cont.)
Analytic continuation:


cosh  1  z  Ln z  z 2  1 
y

z
x
1 x


cosh  1  x  Ln x  x 2  1 
32
 Analytic Continuation Principle (cont.)
Example (cont.)

y
Where would the branch cuts be?

cosh  1  z  Ln z  z 2  1 
f z   z 2  1
Sommerfeld branch cuts
(see note below) 1
x
1
F(z) is analytic in any one quadrant.

Note:

 
1/ 2
f z   z 2  1  Re f  z   0 for all z  Sommerfeld branch cuts for z 2  1

Recall : Re w  0 
33
 Analytic Continuation Principle (cont.)
Example (cont.)
y
Do we need another branch cut
due to the Ln function? 
cosh  1  z  Ln z  z 2  1 
Branch cut due to Ln function

1
x
1
Examine argument of Ln function:

1 1
w z  z 2  1  z   w   derivation omitted
2 w

The branch cut for the Ln(w) function corresponds to w being a negative real
number, i.e, - < w < 0).
Note : w   , 0   z   1,   

w  1
34
 Analytic Continuation Principle (cont.)

Example (cont.)
y

Final Picture 
cosh  1  z  Ln z  z 2  1 
1
x
1

F(z) is an analytic continuation of cosh-1(x) off of the real axis.

F(z) is analytic in any one quadrant.

35
 Analytic Continuation Principle (cont.)

Example (cont.)

Analytic continuation onto Riemann surface:

 
 1/ 2 
cosh  z  ln  z  z  1 
1 2
 

Escalators 1
x
1

cosh  1  x  Ln x  x 2  1 

There are an infinite number of sheets!

36
 EM Example
Assume a radiating phased line source on the z axis.

z Note: j is used here instead of i.

I ( z ) I 0 e  jk z z
kz is real
y

The magnetic vector potential is (ECE 6341):

k  k02  k z2 
1/2

  I 0  (2)  jk z z  k 2  k 2 , k  k
Az  H
 0 ( k   ) e 
0 z z 0

 4 j  2 2
  j k z  k0 , k z  k 0
37
 EM Example (cont.)

 x, y , z  z
We can also write (from ECE 6340):

R I ( z ) I 0 e  jk z z
 jk0 R
 e 
Az  I 0 e  jk z z 
 0  dz 
4 R 0,0, z
  
y
Note: The integral converges for real kz.

Hence
  jk0 R
 e   0 I 0  (2)
Az  I 0 e  jk z z  0  dz   H
 0 ( k   ) e  jk z z

  4 R   4 j 
Exists only for real kz Exists for complex kz

The second form is the analytic continuation of the first form off of the real axis.
38
 EM Example (cont.)

 0 I 0  (2) z
 jk z z
Az  H
 0 ( k   ) e
 4 j 
I ( z ) I 0 e  jk z z
kz is complex

k  k02  k z2  y
1/2

 j k  k
2
z 0 
2 1/2

In order for this to be the analytic continuation off of the real axis of the
integral form, we must chose the branch of the square root function correctly
so that it changes smoothly and it is correct when kz = real > k0.

k   j k z2  k02  negative imaginary number for k z real  k0

1/2

39
 EM Example (cont.)

k   j k  k
2
z 0
2 1/2

Every point in the kz plane is

now a radially decaying wave. Im k z  k   j k z2  k02

Im k   0 Im k   0

k0
Re k z 
 k0
Im k   0 Im k   0

The Sommerfeld branch cuts are a convenient choice.

(but not necessary)
40
 EM Example (cont.)

k   j k z2  k02

The use of the radical sign is equivalent to having Sommerfeld branch cuts.

Im k z 

Im k   0 Im k   0

k0
Re k z 
 k0
Im k   0 Im k   0

41
 EM Example (cont.)

k   j k  k
2
z 0 
2 1/2

k   j k z2  k02
Im k z  (Top sheet)

Riemann surface

k0
Re k z 
 k0
Note:
The function is continuous
on the Riemann surface.

We can now let kz wander anywhere we wish on the Riemann surface, and we know how to
calculate the square root. (We analytically continue to the entire Riemann surface.)
42
 EM Example (cont.)
Example k   j k  k 2
z 
2 1/2
0
What is k at the final
indicated point? k   j k z2  k02
Im k z  (Top sheet)

Riemann surface
Im k   0
k0
Re k z 
 k0 (Bottom sheet)

Final point Im k   0

At the indicated final point, the imaginary part of k is chosen to be positive.

k   j k z2  k02  43
 EM Example (cont.)
Example
k   j k  k 2
z 0 
2 1/2

What type of wave is at

the indicated points?
Im k z  "Surface wave" : k   j  

"Phased array" : k   

1  ,  2 0 
Riemann surface

k0
Re k z 
 k0

"Leaky wave" : k     j  

k   k 
1/2
Note : k   j k z  k0 
1/2
z 0

Note:
The leaky wave field is the analytic continuation of
the phased array field when kz becomes complex.
44
 Schwarz Reflection Principle
Assume that f (z) is the analytic continuation of a real function f (x)
off the real axis (or a segment of the real axis).

f z   f z 
* *
Then within the analytic region, we have

y (proof omitted)

f z 
x
Line segment: f  x 

Examples:
f (z) is assumed analytic in this region.

sin  z , e z , J n  z 

ln  z , Re  z   0 (assuming branch cut on negative real axis)

45