Fundamentals of Seismic

Refraction
Theory, Acquisition, and Interpretation
Craig Lippus
Manager, Seismic Products
Geometrics, Inc.
December 3, 2007

Geometrics, Inc.
Owned by Oyo Corporation,
Japan
In business since 1969
Seismographs, magnetometers,
EM systems
Land, airborne, and marine
80 employees

Located in San Jose,

California

Fundamentals of Seismic
Waves
Q. What is a seismic wave?

Fundamentals of Seismic
Waves
Q. What is a seismic wave?
A. Transfer of energy by way of
particle motion.

Different types of seismic waves are

characterized by their particle motion.

Three different types of

seismic waves
Compressional (p) wave
Shear (s) wave
Surface (Love and Raleigh)
wave
Only p and s waves (collectively referred to
as body waves) are of interest
in seismic refraction.

Compressional (p) Wave

Identical to sound wave particle
motion is parallel to propagation
direction.

Animation courtesy Larry Braile, Purdue University

Shear (s) Wave

Particle motion is perpendicular
to propagation direction.

Animation courtesy Larry Braile, Purdue University

Velocity of Seismic Waves

Depends on density elastic moduli

Vp

4
K
3

Vs

where K = bulk modulus, = shear

modulus, and = density.

Velocity of Seismic Waves

Bulk modulus = resistance to
compression = incompressibility
Shear modulus = resistance to
shear = rigidity
The less compressible a material is, the
greater its p-wave velocity, i.e., sound
travels about four times faster in water
than in air. The more resistant a material
is to shear, the greater its shear wave
velocity.

Q. What is the rigidity of

water?

Q. What is the rigidity of

water?
A. Water has no rigidity. Its shear
strength is zero.

Q. How well does water

carry shear waves?

Q. How well does water

carry shear waves?
A. It doesnt.

Fluids do not carry shear waves. This

knowledge, combined with earthquake
observations, is what lead to the
discovery that the earths outer core is
a liquid rather than a solid shear
wave shadow.

p-wave velocity vs. s-wave

velocity
p-wave velocity must always
be greater than s-wave
velocity. Why?
K
Vp 2

2
Vs

4
3

K 4

K and are always positive numbers, so Vp is

always greater than Vs.

Velocity density paradox

Q. We know that in practice, velocity
tends to be directly proportional to
density. Yet density is in the
denominator. How is that possible?

Velocity density paradox

Q. We know that in practice, velocity
tends to be directly proportional to
density. Yet density is in the
denominator. How is that possible?
A. Elastic moduli tend to increase with density also,
and at a faster rate.

Velocity density paradox

Note: Elastic moduli are
important parameters for
understanding rock
properties and how they
will behave under various
conditions. They help
engineers assess
suitability for founding
dams, bridges, and other
critical structures such as
hospitals and schools.
Measuring p- and s-wave
velocities can help
determine these
properties indirectly and
non-destructively.

Q. How do we use seismic

waves to understand the
subsurface?

Q. How do we use seismic

waves to understand the
subsurface?
A. Must first understand wave
behavior in layered media.

Q. What happens when a

seismic wave encounters a
velocity discontinuity?

Q. What happens when a

seismic wave encounters a
velocity discontinuity?
A. Some of the energy is
reflected, some is refracted.
We are only interested in refracted energy!!

Q. What happens when a

seismic wave encounters a
velocity discontinuity?

Five important
concepts

Seismic Wavefront
Ray
Huygens Principle
Snells Law
Reciprocity

Q. What is a seismic
wavefront?

Q. What is a seismic
wavefront?
A. Surface of constant phase, like
ripples on a pond, but in three
dimensions.

Q. What is a seismic
wavefront?

The speed at which a wavefront

travels is the seismic velocity of
the material, and depends on the
materials elastic properties. In a
homogenious medium, a
wavefront is spherical, and its
shape is distorted by changes in
the seismic velocity.

Seismic wavefront

Q. What is a ray?

Q. What is a ray?
A. Also referred to as a wavefront
normal a ray is an arrow
perpendicular to the wave front,
indicating the direction of travel at
that point on the wavefront. There
are an infinite number of rays on a
wave front.

Ray

Huygens' Principle
Every point on a wave front can be
thought of as a new point source for
waves generated in the direction the
wave is traveling or being propagated.

Q. What causes
refraction?

Q. What causes
refraction?
A. Different portions of the
wave front reach the
velocity boundary earlier
than other portions,
speeding up or slowing
down on contact, causing
distortion of wave front.

Understanding and
Quantifying How Waves
Refract is Essential

Snells Law
sin i V 1

sin r V 2

(1)

Snells Law
If V2>V1, then as i increases, r
increases faster

Snells Law
r approaches 90o as i increases

Snells Law
Critical Refraction
At Critical Angle of incidence ic, angle of
refraction r = 90o
sin(ic ) V 1

sin 90 V 2
sin(ic )

V1
V2

(2)

V1
V2

(3)

ic sin 1

Snells Law
Critical Refraction
At Critical Angle of incidence ic, angle of
refraction r = 90o

Snells Law
Critical Refraction
At Critical Angle of incidence ic, angle of
refraction r = 90o

Snells Law
Critical Refraction
Seismic refraction makes use of
critically refracted, first-arrival
energy only. The rest of the wave
form is ignored.

Principal of
Reciprocity
The travel time of seismic energy
between two points is independent of
the direction traveled, i.e.,
interchanging the source and the
geophone will not affect the seismic
travel time between the two.

Using Seismic Refraction

to Map the Subsurface
Critical Refraction Plays a Key
Role

T 1 x /V 1
T2

ac cd df

V1 V 2 V1

ac df

h
cos(ic )

bc de h tan(ic )
cd x bc de x 2h tan(ic )
T2

2h
x 2h tan(ic )

V 1 cos(ic )
V2

T2

2h
2h tan(ic ) x

V 1 cos(ic )
V2
V2

1
sin(ic )
x

V2
V
1
cos
(
i
c
)
V
2
cos(
i
c
)

T 2 2h


V2
V 1 sin(ic )
x

T 2 2h

V 1V 2 cos(ic ) V 1V 2 cos(ic ) V 2
V 2 V 1 sin(ic )
x

T 2 2h
V2
V
1
V
2
cos(
i
c
)

V2

sin(ic )

x

T 2 2hV 1 V 1
V 1V 2 cos(ic ) V 2

sin ic

V1
V2

sin(ic )
x
sin(ic )

T 2 2hV 1

V 1V 2 cos(ic) V 2

(Snells Law)

T 2 2hV 1

1 sin 2 (ic )
x

V 1V 2 sin(ic ) cos(ic ) V 2

cos 2 (ic )
x

T 2 2hV 1
V 1V 2 sin(ic ) cos(ic ) V 2
cos(ic )
x

T 2 2h
V 2 sin(ic) V 2

From Snells Law,

V 1 V 2 sin(ic )

T2

2h cos(ic ) x

V1
V2

(4)

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Using Seismic Refraction

to Map the Subsurface

Depth

Depth

Xc V 2 V 1
2 V 2 V1

(5)

Using Seismic Refraction

to Map the Subsurface

Depth

Xc V 2 V 1

2 V 2 V1

T iV 1

V1
2 cos(sin
)
V
2
For layer parallel to
surface

Depth

(6)

Summary of Important
Equations
sin i V 1

sin r V 2

(1) Snells Law

For refractor
parallel to surface
T2

sin(ic )

ic sin 1

V1
V2
V1
V2

(2)
h

(3)

2h cos(ic) x

V1
V2

(4)

Xc V 2 V 1
2 V 2 V1

(5)

T iV 1

1 V 1
2 cos sin

V
2

(6)

h1

Ti 2V 1
V1
2 cos(sin
)
V2
1

cos(sin 1 V 1 / V 3)
Ti 3 Ti 2
V2

cos(sin 1 V 1 / V 2)
h1

1
2 cos(sin V 2 / V 3)

h2

cos(sin 1 V 1 / V 4) 2h 2 cos(sin 1 2 / V 4)
V3
Ti 4 Ti 2 cos(sin 1 V 1 / V 2)

V2

h3
h1 h 2
1
2 cos(sin V 3 / V 4)

Crossover Distance vs.

Depth

Depth/Xc vs. Velocity

Contrast

Important Rule of Thumb

The Length of the Geophone
Spread Should be 4-5 times
the depth of interest.

Dipping Layer
Defined as Velocity Boundary
that is not Parallel to Ground Surface
You should always do a minimum
of one shot at either end the
spread. A single shot at one end
does not tell you anything about
dip, and if the layer(s) is dipping,
your depth and velocity calculated
from a single shot will be wrong.

Dipping Layer
If layer is dipping (relative to ground
surface), opposing travel time curves
will be asymmetrical.
Updip shot apparent velocity > true velocity
Downdip shot apparent velocity < true velocity

Dipping Layer

Dipping Layer
V 1md sin(ic )

V 1mu sin(ic )
ic sin 1 V 1md
ic sin 1 V 1mu

1
ic (sin 1 V 1md sin 1 V 1mu )
2
1
(sin 1 V 1md sin 1 V 1mu )
2

Dipping Layer
From Snells Law,

V1
V2
sin(ic )
V 1Tiu
2 cos(ic )
Du
cos
V 1Tid
2 cos(ic )
Dd
cos

Dipping Layer
The true velocity V2 can also be calculated
by multiplying the harmonic mean of the updip and down-dip velocities by the cosine of
the dip.

2V 2UV 2 D
V2
cos
V 2U V 2 D

What if V2 < V1?

What if V2 < V1?

Snells Law

sin i V 1

sin r V 2

What if V2 < V1?

Snells Law

sin i V 1

sin r V 2

What if V2 < V1?

If V1>V2, then as i increases, r
increases, but not as fast.

If V2<V1, the energy

refracts toward the normal.
None of the refracted energy
makes it back to the surface.
This is called a velocity inversion.

Seismic Refraction requires

that velocities increase
with depth.
A slower layer beneath a
faster layer will not be
detected by seismic refraction.
The presence of a velocity inversion can
lead to errors in depth calculations.

Delay Time Method

Allows Calculation of Depth
Beneath Each Geophone
Requires refracted arrival at each
geophone from opposite directions
Requires offset shots
Data redundancy is important

Delay
Time
Method
x
V1

V2

Delay
Time
Method
x
V1

V2
TAB

hA
AB hA tan(ic ) hB tan(ic )
hB

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )

Delay
Time
Method
x
V1

V2
TAB

hA
AB hA tan(ic ) hB tan(ic )
hB

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )

TAP

hA
AP hA tan(ic ) hP tan(ic )
hP

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )

Delay
Time
Method
x
V1

V2
TAB

hA
AB hA tan(ic ) hB tan(ic )
hB

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )

hA
AP hA tan(ic ) hP tan(ic )
hP
TAP

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )
hB
BP hB tan(ic ) hP tan(ic )
hP
TBP

V 1 cos(ic ) V 2
V2
V2
V 1 cos(ic )

Delay
Time
Method
x
V1

V2
Definition:

t0 T

AP

BP

AB

(7)

t 0 TAP TBP TAB

hB
BP hB tan(ic ) hP tan(ic )
hP
hA
AP hA tan(ic ) hP tan(ic )
hP

t0

V
1 cos(ic )
V
2
V
2
V
2
V
1 cos(ic )
V
1 cos(ic )
V
2
V
2
V
2
V
1 cos(ic )

hA
AB hA tan(ic ) hB tan(ic )
hB

V
1 cos(ic )
V
2
V
2
V
2
V
1 cos(ic )

t0

AP BP AB
2hp
2hP tan(ic )

V2
V 1 cos(ic )
V2

But from figure above, AB AP BP. Substituting, we get

t0

AP BP AP BP
2 hp
2hP tan(ic )

V2
V 1 cos(ic )
V2
or

t0

2 hp
2hP tan(ic )

V 1 cos(ic )
V2

1
sin(ic )

t 0 2hp

V 1 cos(ic )
V 2 cos(ic )

V2
V 1 sin(ic )

t 0 2hp

V 1V 2 cos(ic )
V 1V 2 cos(ic )

V2

sin(
i
c
)

V
1

t 0 2hpV 1

V
1
V
2
cos(
i
c
)
V
1
V
2
cos(
i
c
)

Substituting from Snells Law, sin ic V 1

V2

sin(
i
c
)

sin
i
c

t 0 2hpV 1

V
1
V
2
cos(
i
c
)
V
1
V
2
cos(
i
c
)

sin(
i
c
)

sin ic

t 0 2hpV 1

V 1V 2 cos(ic )
V 1V 2 cos(ic )

Multiplying top and bottom by sin(ic)

1
sin 2 (ic )

t 0 2hpV

V 1V 2 sin(ic) cos(ic ) V 1V 2 sin(ic ) cos(ic )

cos 2 (ic )

t 0 2hpV
V 1V 2 sin(ic ) cos(ic )

cos(ic )

t 0 2hp
V 2 sin(ic )

cos(ic )

t 0 2hp
V 2 sin(ic )

Substituting from Snells Law,

sin ic

V1

V2

We get

2hp cos(ic )
t0
V1

(8)

to 2hp cos(ic ) hp cos(ic )

Delay time at point P DTP

2
2V 1
V1

(9)

Reduced Traveltimes
x

Definition:
TAP = Reduced Traveltime at point P for a source at A
TAP=TAP
Reduced traveltimes are useful for determining V2. A
plot of T vs. x will be roughly linear, mostly unaffected
by changes in layer thickness, and the slope will be
1/V2.

Reduced Traveltimes
x

From the above figure, TAP is also equal to TAP minus the
Delay Time. From equation 9, we then get

T ' AP TAP DTP TAP

to
2

Reduced Traveltimes
x

Earlier, we defined to as

t0 T

AP

BP

AB

(7)

Substituting, we get

T ' AP TAP

to
TAP TBP TAB
TAP
2
2

(10)

Reduced Traveltimes
Finally, rearranging yields

T ' AP

AB

T AP T BP
2

(11)

The above equation allows a graphical determination of the T

curve. TAB is called the reciprocal time.

Reduced Traveltimes
T ' AP

AB

T AP T BP
2

The first term is represented by the dotted line below:

Reduced Traveltimes
T ' AP

AB

T AP T BP
2

The numerator of the second term is just the difference in the

traveltimes from points A to P and B to P.

Reduced Traveltimes
T ' AP

AB

T AP T BP
2

Important: The second term only applies to refracted arrivals. It

does not apply outside the zone of overlap, shown in yellow
below.

Reduced Traveltimes
T ' AP

AB

T AP T BP
2

The T (reduced traveltime) curve can now be determined graphically

by adding (TAP-TBP)/2 (second term from equation 9) to the TAB/2 line
(first term from equation 9). The slope of the T curve is 1/V 2.

We can now calculate the delay time at point P. From Equation 10,
we see that

to
T ' AP TAP
2

(10)

to hp cos(ic )

2
V1

(8)

According to equation 8

So

T ' AP TAP

t0
hp cos(ic )
TAP
2
V1

(12)

Now, referring back to equation 4

2h cos(ic ) x
T2

V1
V2

(4)

Its fair to say that

2hp cos(ic ) x
TAP

V1
V2

(13)

Combining equations 12 and 13, we get

hp cos(ic ) 2hp cos(ic ) x hp cos(ic )

T ' AP TAP

V1
V1
V2
V1
Or

T ' AP

hp cos(ic ) x

V1
V2

(14)

Referring back to equation 9, we see that

hp cos(ic )
DTp
V1

(9)

Substituting into equation 14, we get

hp cos(ic ) x
x
T ' AP

DTp
V1
V2
V2
Or

x
DTp T ' AP
V2

(15)

Solving equation 9 for hp, we get

D TPV 1
hP
c o s (ic)

(16)

We know that the incident angle i is critical when r is 90o.

From Snells Law,

sin i V 1

sin r V 2
sin ic V 1

sin 90 V 2
V1
sin ic
V2
V1
ic sin

V2
1

Substituting back into equation 16,

DTpV 1
hp
cos(ic )

(16)

we get

hp
cos

DTpV 1
1 V 1
sin

V2

(17)

In summary, to determine the

depth to the refractor h at any
given point p:

1.Measure V1 directly from the

traveltime plot.

2.Measure the difference in traveltime

to point P from opposing shots (in
zone of overlap only).

3.Measure the reciprocal time TAB.

4. Per equation 11, T ' A P

AB

T AP T BP
2

divide the reciprocal time TAB by 2.

T AP T BP

,
5. Per equation 11, T ' A P 2
2
add the difference time at each
point P to TAB/2 to get the reduced
traveltime at P, TAP.
AB

6. Fit a line to the reduced

traveltimes, compute V2 from slope.

7. Using equation 15,

x
DTp T ' AP
V2

(15)

Calculate the Delay Time DT at

P1, P2, P3.PN

8. Using equation 17,

hp
cos

DTpV 1
1 V 1
sin

V2

(16)

Calculate the Depth h at P1, P2,

P3.PN

Thats all there is to it!

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less

More Data is Better Than Less