Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

WAVELET FUNCTIONS AND THEIR POLYNOMIALS·

NORMAN RICKERt

ABSTRACT
This paper is a fu~ther extension of the author's wavelet theory of seismogram struc-
ture. The wavelet fune/ions used in the previous papa- are exltended, a modified notation
is adopted, and twenty four of the polynomials are given, together with their roots.
Expressions are given for the displacement, velocity, and acceleration type wavelets.

In a recent paper the author gave a solution of the problem of the

form of the seismic disturbance which travels outward from a sudden
explosion in a homogeneous and isotropic earth. This solution was
developed for the dilatational part of the disturbance but, with modi-
fied constants, may be used to describe the shear part. The complete
equation for the propagation of waves in an elastic earth in which there
is absorption due to internal 'Viscosity' may be written

grad div [(X + 2P.) + ~" ~J X

3 iJt
(1)
-
iJJ
curl curl [ + ,,- X =
Jl.
(J2X
P --
at at 2
wherein X represents the earth particle displacement, X and ~ are the
elastic constants of Lame, p the density, and 'IJ the internal 'Viscosity of
the earth.
X, being a vector point function, may be written as the sum of two
vectors
X = Xl + X2 (2)

where Xl is irrotational and X2 is solenoidal. That is to say

curl Xl = 0 and div X2 = o. (3)
The differential equation thus splits up into two differential equations

* Manuscript received March I, 1944·

t Geophysical Research Laboratory of The Carter Oil Company, Tulsa, Oklahoma.
1 Norman Ricker, Further developments in the wavelet theory of seismogram structure.
Bull. Seismol. Soc. Amer.,Vol. 33 (1943), pp. 197- 228 .
I Loc. cit., p. 197.
 WAVELET FUNCTIONS AND THEIR POLYNOMIALS 315
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

and
(5)

Calling
(6)
and

These equations become

(8)

and

(9)

Calling
Wo = 3pC 2/ 47J (10)
and
(II)
The equations are now

(12)

and

These are the equations of wave propagation of the dilatational

and shear disturbances respectively. The quantities Wo and Wo are con-
stants of the medium as defined by Eqs. 10 and I I above. The
transition frequencies of the medium are given by
Wo _ Wo
fo = - and fo = - . (14)
21T 21T
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

NORM AN RICKER

These transition frequencies are the frequencies above which dis-

persion becomes important for continuous sinusoidal waves. We have
the relation

::= :(~y
We shall carry along for the remainder of the paper using the con-
stants for the dilatational part of the disturbance with the understand-
ing that the shear wave counterpart may be obtained by replacing
Wo by Wo and c by c.
For plane waves moving along the x axis, Eq. 12 becomes

a2 ( I a<I» I a2<I>
ax 2 <I> + Wo at = ~ a;; .
<I> is a generating function from which X is obtained by
<I>a
X =--
at
Introducing dimensionless quantities X and T which we shall call
numerical distance and numerical time respectively, defined as
Wox
X == -- and T == wot, (17)
c
Eq. 16 becomes

A solution of this equation may be written

<I> = L -
00 (2 )(n+llI2
'¥n(U) ,
n-O X
wherein
u == 2 112 ·X- 112(T - X),
and '¥n(U) is the wavelet function of order n. From this solution the
three important forms of seismic disturbance may be obtained. De-
noting by a dot a derivative with respect to T, and by a prime a deriva-
tive with respect to the argument of a function, we have the following:
The displacement function.
• 00 (2)(n+2l/2 I
<I> = L - '¥n (u). (21)
n~O X
 WAVELET FUNCTIONS AND THEIR POLYNOMIALS 317
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The velocity function.

. = L'" (2- )(,,+3)/
<I>
2
\{r,,"(u). (22)
n~O X

Thc accclcration function.

. . = L'" (2- )("+4)/

1>
2
\{rn'''(u).
n~O X

The wavelet function \{rn(U) may be written

wherein
y'; 2
== -.c-u /4
2

and S"O(u) is a polynomial of degree 311. Similarly, the m-th derivative

of \{r,,(u) may be written

wherein Sr.'n(u) is a polynomial of degree 3n+m. These polynomials

are given below for n=o to 5, and for m=o to 3 inclusive.
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3 IS
NORMAN RICKER
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

lows:
WAVELET FUNCTIONS AND THEIR POLYNOMIALS

These polynomials may be expressed in terms of their roots as fol-

319
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

320 NORMAN RICKER

where
lal = 2.4494897 ,al = (.3489532) + i(·347 0 87 8 )
2a l= .4386°57 ,a, = (.3489532) - i(.3480878)
2a2 = 2.°316884 ,a, = 1.7372969
2a , = 4.°976796 .a, = 3.°°39589
,al = .3796°34 .a. = 4.3370220
,a2 = 1.8957716 ,a. = 5.7988°43
,a, = 3.5°8°532 ,a, = 7.5253341
,a, = 5.4II9171
,al = (.2543301) + i(.1537223)
,a2 = (.2543301) - i(.1537223)
,a, = 1.8043057
,a, = 3.2025824
,a. = 4.7320146
.a~ = ti·S530348

(z8a)

Ib) = 1.°492952 ,b l = (.2418248) + i(.224 10 47)

Ib, = 3.3013602 ,b, = (.2418248) - i(.2241047)

2bl = I. 2643058 .b a = 1.3550222
2b 2 = 2.8715097 ,b, = 2.4746194
2 b, = 4.8235956 ,b. = 3.6653883
,b l = .214669° .b. = 4.9464936
ab, = 1.3162830 .b, = 6.3633569
ab, = 2.6799645 .b8 = 8.0460916
(28b)
ab, = 4.2192654
ab• = 6.0473332
,b l = .2080355
,b, = 1.3437754
,b a = 2.5594485
,b, = 3.8863697
,b, = 5.3580087
,h. = 7.1031509
 WAVELET FUNCTIONS AND THEIR POLYNOMIALS 321
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

( 3)0
SO'= -
4
I
'--
2 2 ·0!
(U'-OCI 2)

where
OCI = 1.4142136 .CI = .1218601
ICI = I.9I71449 ,C2 ~ 1.0268945
IC, = 4.0403657 .C, = 2.0187355
2CI = .63I3067 .C, = 3.0954378
,C2 = 2.0037771 .e. = 4.2501778
2C3 = 3.5825660 .C6 = 5.497337 8
2C , = 5.46 46 739 .C7 = 6.88~o771

3CI = .8494710 ,C8 = 8.5294059

3C2 = 2.0184280
3C3 = 2.25'3'24
3C, = 4.8450276
3C, = 6.6205812
,CI = .II88661
,C, = .9592981
,C3 = 2.02II776
'c, = 3.2020235
'c. = 4.49043'5
,C6 = 5.9224984
,C7 = 7.6258907
 3 22 NORMAN RICKER
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

odl = 2·44948<)7 .dl = .069967921

Id l = .8721548 .d, = .74347164
Id z = 2.6716981 .da = 1.61 7 2947
Ida = 4·7012099 .d4 = 2.5969881
2dl = 1.2851870 .d. = 3.6531245
.d~ = 2.6705800 .a. = 4.7 8368°3
2d. = 4.2158628 .a7 = 6.005355 2
2d 4 = 6.0459828 .a. = 7.3627757
adl = .4416455 .as = 0
8.982 75 3
ad2 = 1.4593267
ad. = 2.6442323
- i14 = 3.9534818
ad. = 5.4126795
ad. = 7.1476965
cd, = .6286818
4d 2 = 1.5554091
4d. = 2.6191473
411-, = 3·7795034
411-. = 5.04°3263
411-6 = 6.4419739
411-7 = 8.1119438
The'polynomial S"m(u) satisfies the recurrence formula
 WAVELET FUNCTIONS AND THEIR POLYNOMIALS 323
Downloaded 08/29/13 to 142.150.190.39. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The notation has been changed somewhat from that given in the
paper mentioned in footnote (1) above. The functions P,,(u) and Q,.(u)
of that paper are related to 'l',,(u) by
Pn(u) = - 'l'/(u)
and

This reversal in sign has been made so that in all three forms of
the seismic disturbance, displacement, velocity, and acceleration, the
first motion is positive. The wavelet-form functions may be written

V(u X) -
t E(; ) nI2'l'n"(U)

W(u X) ==
t f
n~O
(-=-)"/2'l'nll/(u).
X

These functions represent the forms of the earth displacement,

velocity, and acceleration type wavelet respectively. Here again the
sign has been reversed~the script 'V; V being the negative of the
italic U; V of the previous paper. In the present notation the solution
with X held constant and the time allowed to vary, describes the con-
ventional motion of the trace of a seismograph, with time advancing
from left to right and the first motion upwards. Also, in the present
notation, the solution with T held constant and X allowed to vary
gives the wave profile in the earth with the wave advancing from left
to right and the first motion positive.
It may be seen from these solutions that for plane waves the wave-
let amplitude decays as (Z/X)2/2; (2/X)3/2; (2/X)4/2 for the displace-
ment, velocity, and acceleration wavelet types respectively. For
spherical waves spreading out from a center the wavelet amplitude
must decay as (Z/R)4/ 2; (Z/R)5/ 2; (2/R)6/ 2 for these respective forms,
wherein R is the numerical radial distance defined by
WOr
R ==~­
c
in which r is the true distance.