GEOPHYSICS 224 – Winter 2008

CHAPTER 3 – SEISMIC WAVES

(Textbook: chapter 3)

3.1 Introduction to waves

A wave is a periodic disturbance that transmits energy through a medium, but does not
cause permanent deformation of the material. Seismic waves are parcels of elastic strain
energy that propagate outward from a seismic source (e.g., earthquake, explosion).

There are two complementary ways to look at waves:

View 1. A snapshot of the displacement in the material at a given time.

 2

View 2. A plot of the displacement as a function of time for a given point in the material.

Some general characteristics of waves:

Amplitude (A) – the maximum amount of displacement caused by the wave – i.e., from
the rest position (0) to a peak or trough. Note that the peak-to-trough displacement is 2A.

Wavelength (λ) – the distance between two points on a wave that have the same
displacement (or phase). For example, the distance between two peaks or two troughs.
The units are distance (m, km, etc.)

Period (T) – the time taken for a wave to go through one cycle (e.g., peak, trough, peak).
The units are time (s).

Frequency (f) – the number of wave cycles that occur in one second. This is the inverse
of the period (f = 1/T). The units are Hertz (Hz). 1 Hz = 1 s-1.

Angular frequency (ω) – sometimes frequency is reported as an angular frequency,

which reflects the number of rotations per second: ω = 2πf. The units are radians per
second (rad s-1).

Velocity (v) – the rate at which the wave passes through the material. v = λf
 3

Seismic exploration uses recordings of seismic waves that arrive at a seismic station
(view #2). To interpret the recordings, we must determine how the waves have
propagated through the subsurface – how their velocity and direction have changed, how
their amplitude has changed, etc.

We will study wave propagation by thinking about wavefronts and rays. Consider a
seismic wave that is generated by an explosion (point source). The resulting waves
propagate outward from the source:

• The wavefront is the locus of points

where the pulse has reached at a given
time. All points on the wavefront have
the same phase (e.g., peak, trough)

• The ray is a vector which shows the

direction that the wave is travelling.

• The wavefront and ray are perpendicular

to one another

Over time, the wavefront will expand spherically. The velocity at which the wavefront
travels from the source is the seismic velocity of the material.

Propagation of the wavefront over time can be visualized using Huyghens Principle:
This states that every point on a wavefront acts as a secondary point source generating
spherical waves. The secondary waves propagate outwards and at a later time, the
envelope of secondary waves gives the overall wavefront.
 4

3.2 Types of seismic waves

3.2.1 Body waves

Body waves are seismic waves that travel through the interior of the Earth. There are two
types of body waves:

• P-waves are compressional waves where the particle motion is in the direction of
wave propagation.

• S-waves are shear waves where the particle motion is perpendicular to the
direction of propagation.

Fig 3.3 from Kearey et al., (2002):

The passage of seismic energy through a material exerts a stress, which causes the
material to deform (strain). The deformation is elastic – once the stress is removed, the
material returns to its undeformed state. The velocity at which a seismic wave travels
through a material is given by the elastic parameters and density of the material.

Elastic parameters (moduli) – parameters that provide a measure of the strength of a

material (how much stress is needed to produce a given strain).
 5

Longitudinal (axial) modulus (ψ). Consider a block of length L and cross-sectional

area A. If a force (F) is applied to one end, it will shorten by an amount ΔL. The
relationship between the F and ΔL is given by the longitudinal modulus:

ψ = longitudinal stress (F/A) / longitudinal strain (ΔL/L)

The bulk modulus (K) determines how the volume of a material will change when a
(three-dimensional) pressure is applied:

K = volume stress (P) / volume strain (ΔV/V)

The shear modulus (μ) gives the relationship between the shear stress and resulting
shear strain:

μ = shear stress (τ) / shear strain (tan θ)

4
It can be shown that the three moduli are related: ψ = K + μ
3

Body wave velocities

P-waves involve uniaxial compressional strain. Therefore, the P-wave velocity (VP)
depends on the longitudinal modulus:
1
⎛ 4 ⎞2
⎜K + μ⎟
1
⎛ψ ⎞ 2
3 ⎟
V P = ⎜⎜ ⎟⎟ =⎜
⎝ρ⎠ ⎜
⎜
ρ ⎟
⎟
⎝ ⎠

S-waves involve shearing and therefore their velocity (VS) depends on the shear modulus:
1
⎛ μ ⎞2
VS = ⎜⎜ ⎟⎟
⎝ρ⎠

Note that:
• VP is always larger than VS

• the moduli will be larger for stronger materials (more stress is needed for elastic
deformation), therefore VP and VS will be larger

• VP and VS decrease as density increases

• VP and VS do not depend on frequency – they are non-dispersive

• in fluids, μ=0 (and K>0) – only P-waves can travel through fluids (VS=0)
 6

VP/VS and Poisson’s ratio

When a column of material is stretched, it will get longer (longitudinal strain), but its
thickness will decrease (lateral strain). This is described by the Poisson’s ratio (σ):

σ = lateral strain / longitudinal strain

1
VP ⎡ 2(1 − σ ) ⎤ 2
It can be shown that: =
VS ⎢⎣ (1 − 2σ ) ⎥⎦

For consolidated crustal material, σ ~ 0.25. Therefore, VP/VS ~1.7. If there is fluid
present, VS decreases. Thus, a large VP/VS ratio (or large σ) can be indicative of fluids.

3.2.2 Surface waves

Surface waves travel along interfaces, such as the ground surface. There are two types:

• Rayleigh waves are generated by coupling between P-waves and S-waves at an

interface. Rayleigh waves produce an elliptical retrograde particle motion in the
plane perpendicular to the surface and parallel to the direction of wave
propagation. Particle motion decreases exponentially with depth. Rayleigh wave
velocity is less than the shear wave velocity. Because the Earth has a layered
structure (body wave velocities increase with depth), Rayleigh waves are
dispersive – their velocity depends on frequency. Lower frequency Rayleigh
waves have a higher velocity because they extend deeper into the Earth. In
seismic exploration, the Rayleigh waves are sometimes called ground roll.

• Love waves occur when a near-surface layer has a lower shear wave velocity than
the underlying layer. Shear waves become trapped in the near-surface layer,
producing Love waves. Love waves have horizontal particle displacements,
perpendicular to the direction of wave travel. Love waves are dispersive and have
a velocity that is intermediate between the shear wave velocities of the two layers.

Fig 3.4 from Kearey et al. (2002):

 7

3.3 Seismic velocities of rocks

As shown above, seismic velocity is given by the material properties (elastic parameters,
density), which depend primarily on composition. Typical P-wave velocities (km s-1) are:

Air 0.3
Water 1.4-1.5
Petroleum 1.3-1.4
Sand (dry) 0.2-1.0
Water-saturated sand 1.5–2.0
Sandstones
Tertiary 2.0-2.5
Carboniferous 4.0-4.5
Limestones
Cretaceous 2.0-2.5
Jurassic 3.0-4.0
Carboniferous 5.0-5.5
Salt 4.5-5.0
Granite 5.5-6.0
Gabbro 6.5-7.0
Ultramafic rocks 7.5-8.5

Other factors that affect seismic velocity are:

Porosity – a porous rock contains a rock matrix and pores. The porosity (Φ) is the
fractional space that is occupied by pores. Since pores are generally filled with a low
velocity material (e.g., air, water, petroleum), the overall seismic velocity of the rock will
decrease with increasing porosity. In general, the overall rock properties will be the
average of the matrix and pore fluid properties, weighted by the porosity. For example, if
the matrix has a density ρm and the pore fluid has a density ρf, the overall (bulk) density
of the rock will be:
ρ = ρf Φ + (1- Φ) ρm

The bulk seismic velocity is calculated by weighting the matrix and pore velocities by the
amount of time that a seismic wave spends in each component. The time is inversely
proportional to the velocity of each material. It can be shown that the bulk VP of the rock
is given by the time-average equation:
1 Φ (1 − Φ )
= +
VP V f Vm

Depth – at larger depths, compaction results in a reduction of porosity, increasing

seismic velocities. In addition, the elastic moduli tend to increase at higher pressures.
Thus, even for a constant lithology, seismic velocities will be higher at greater depths.

Age – velocities tend to increase with age of sedimentary rocks, owing to cementation.
As a rock becomes more cemented, its rigidity (strength) increases.
 8

Questions

1. A rock has 25% porosity. The P-wave velocity for the rock matrix and the pore fluid
are 3.2 and 1.5 km s-1, respectively. What is the overall P-wave velocity of the rock?

2. Consider a gas reservoir which has 10%

porosity. From well-logs, the following P-wave
velocities were determined:
Sandstone 4.3 km s-1
Water 1.5 km s-1
Gas 0.3 km s-1
Shale 2.4 km s-1

What is the average P-wave velocity in the water-saturated reservoir? What is VP in the
gas-filled reservoir?

3. In a similar reservoir, VP in the gas-filled region was measured to be 2.2 km s-1. What
porosity does this imply?

3.4 Factors that affect seismic wave amplitude

In seismic exploration, the travel time and amplitude of the recorded seismic waves are
used to determine the distribution of seismic velocities in the subsurface and hence map
subsurface geology. A number of factors can cause the amplitude of seismic waves to
decrease as they travel from a source to the receiver:

1. Geometrical spreading. Seismic waves originate from a point source (e.g.,

explosion) and carry energy away from the source as a spherical wavefront of expanding
radius. Conservation of energy requires that the total amount of energy in the wavefront
is constant at all times. Therefore, as the sphere expands, the energy will be spread out
over a larger area, so that the energy at each point on the sphere surface will decrease. If
the energy at the source is E0, the energy per unit area at any point on a sphere of radius r
will be:
E
E (r ) = 0 2 where 4πr2 is the surface area of the sphere.
4πr

The energy of seismic waves is proportional to their amplitude (A) squared: E(r) α A2.

1 1 1
Since E (r )α : A 2α or Aα
r2 r2 r

Thus, the amplitude of seismic waves will decrease by 1/r as the sphere expands. This is
known as geometrical spreading or spherical divergence.
 9

2. Attenuation. As a seismic wave travels through a material, the deformation is not

perfectly elastic. Some energy is converted into heat due to friction associated with
movement along grain boundaries. The loss of energy results in a decrease in the
amplitude of the seismic waves. This is called attenuation or absorption.

Attenuation is characterized by the quality factor (Q), which indicates the fractional loss
of energy per cycle of the seismic wave:
2πE
Q=
ΔE

where E is the total elastic energy in the wave and ΔE is the energy lost in one cycle
(oscillation). High values of Q mean that little energy is lost.

It can be shown that attenuation results in an exponential decrease in amplitude of the

seismic wave as it travels. If the seismic wave has an initial amplitude A0, its amplitude
after a time t will be:
− πft

A( t ) = A0 e Q

where e is the exponential constant (e = 2.718) and f is the frequency of the seismic wave.

This equation can also be expressed in terms of the distance (x) travelled by the wave:
− πx
Qλ
A( x ) = A0 e
where λ is the wavelength of the wave. (to derive this, remember that v=fλ and v=x/t).

The above equations indicate that the amplitude decays more rapidly for low values of Q
or for waves that have a high frequency (or short λ).
 10

Seismic sources generally release energy over a wide range of frequencies. The pulse of
energy can be decomposed into a set of sine and cosine curves with different wavelengths
(Fourier’s Theorem). As the wave travels, the higher frequency (short wavelength)
components will be attenuated more strongly. Over time, the pulse will become
dominated by the longer wavelength (low frequency) components, and will be smoother
and have a longer duration (see Kearey et al., Fig 3.7).

Average values of Q are 100-200 for the lithosphere. Q tends to be larger for P-waves
than S-waves (QP > QS) – S-waves are attenuated more than P-waves. This may be
because the conversion of seismic energy into frictional heat may be primarily associated
with shear deformation.

3. Scattering. Most rocks contain small heterogeneities, such as grains with different
compositions and the edges of minerals or pores. As a seismic wave encounters these
heterogeneities, some seismic energy will be scattered. As a result, the seismic wave will
lose some energy, and its amplitude will decrease.

3.5 Seismic sources

Most seismic exploration uses artificially generated seismic waves (active technique).
Natural signals can also be used (passive) – e.g., seismic waves generated by
earthquakes. The passive approach is common in regional/global tectonic studies.

The artificial seismic source for seismic exploration should satisfy four requirements:
- safe, efficient and environmentally acceptable
- economical
- generate sufficient energy over a frequency range suitable for the study
- repeatable – should always produce the same waveform

3.5.1 Land sources

Explosives – explosives can be detonated in shallow boreholes (6-30 m depth). Using
boreholes improves coupling to the solid ground and reduces surface damage.

Advantages: - produce a sharp, impulsive signal of mostly P-wave energy (why?)

- can produce a high amplitude signal
- reasonably cheap (although borehole drilling can be expensive)

Disadvantages: - signal is not repeatable

- slow (especially borehole drilling)
- can be destructive and generally can not be used in urban areas

Impulsive signals can also be generated by other methods:

- striking a hammer on a metal plate (hammer seismic survey)
- dropping a weight from a truck or helicopter
- using shotguns, buffalo guns, or rifles
 11

Vibroseis method (“Thumper trucks”) – a heavy truck (30,000 pounds) lowers a base
plate onto the ground. By lifting its wheels, the weight of the truck is concentrated onto
the base plate. Low amplitude vibrations are generated for 10-20 seconds. The signal
varies in frequency between 10 and 100 Hz. This is known as a sweep. For higher
amplitude signals, multiple trucks are used and they vibrate in phase with one another.

Advantages: - well-defined, repeatable signal

- little to no disturbance to the surrounding land
- trucks can also produce horizontal shaking, creating S-waves

Disadvantages: - need a hard surface

- confined to areas that are accessible to the trucks
- trucks are expensive (>$500,000)

Processing of Vibroseis data

The long sweep time of the Vibroseis source means that reflections from subsurface
layers are spread out over time. Since the original source signal is well-known, cross-
correlation can be used to isolate the individual reflections. This works well, even when
reflections over-lap with one another and where there is significant noise in the data.
When multiple trucks are used or multiple recordings are made at a site, the individual
sweeps can be added together (stacked) in order to improve the signal-to-noise ratio.
The assumption is that at each point in time, the recorded data contains a coherent signal
from subsurface reflections (which is the same for all recordings) and noise (which is
incoherent or random among the recordings). By stacking multiple data sets, the noise
will cancel and the signal will be enhanced.

Example 1 – two overlapping reflections in Vibroseis data

2
Amplitude

Reflection from base of layer 1

0

-2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2
Amplitude

Reflection from base of layer 2

0

-2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2
Amplitude

Reflection 1 + Reflection 2
0

-2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
0.1
Amplitude

0
Cross-correlation with source signal
-0.1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
 12

Example 2 – two overlapping reflection and noise in Vibroseis data

2
Amplitude Reflection from base of layer 1
0

-2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2
Amplitude

Reflection from base of layer 2

0

-2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
2
Amplitude

0
-2 Reflection 1 + Reflection 2 + noise
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
0.1
Amplitude

0
Cross-correlation with source signal
-0.1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)

Example 3 – stacking noisy data to increase the signal to noise ratio

2
Amplitude

Vibroseis signal
0

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
2
signal + noise
Amplitude

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
2
Amplitude

5-fold stack
0

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
2
Amplitude

20-fold stack
0

-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (s)
 13

3.5.2 Marine sources

Air guns – the air gun is lowered under the water and high-pressure, compressed air is
pumped into a chamber. Through electrical triggering, the air is released, creating a high-
pressure bubble. Rapid expansion of the bubble creates seismic waves with a frequency
of 10-100 Hz.

Advantages - reliable, repeatable source

- inexpensive

Disadvantages - the bubble pulse oscillates, producing a long pulse length (can be
improved by using an airgun array – see Fig 3.18 in Kearey et al., 2002)
- only P-waves are created (why?), but S-waves may be generated once
the signal travels through the seafloor
- there has been concern that air guns may be harmful to marine animals

Sparkers, pingers, boomers – a signal is produced electrically, with a frequency of 100-

100,000 Hz. These are most often used for mapping the shallow seafloor and for
geotechnical surveying (e.g., mapping sites for submarine pipelines).

Advantages - well-known source signal

- reliable and inexpensive

Disadvantages - the high frequency signal attenuates very quickly, and thus these
devices can only image the upper ~1 m of the seafloor

3.6 Seismic detectors

Seismic detectors are devices that sense the motion associated with seismic waves and
turn it into an electrical signal that is recorded for later analysis. Devices can measure the
displacement, velocity or acceleration associated with the seismic wave motion. Seismic
waves at the surface produce three-dimensional motions. For seismic exploration, much
of the signal has a nearly vertical incidence angle, and thus, it is usually sufficient to only
record vertical motions. The recording system itself has to be timed accurately relative to
the seismic source, with timing errors less than 1 millisecond.

3.6.1 Land detectors

Geophones - most of these instruments only measure a single component of motion
(vertical). The most common design is the moving-coil geophone. A cylindrical coil is
suspended vertically from a spring support. It is surrounded by a permanent magnet,
which is attached to the instrument casing. The geophone must be installed so that it is
solidly coupled to the ground. Vertical motion of the ground causes the instrument and
magnet to move up and down, while the coil is stationary. This produces an electric
 14

voltage through the coil, which is recorded as an

electric signal. The unit is calibrated so that the
voltage is proportional to the velocity of the coil.
Care must be taken to ensure that behaviour of the
coil and magnet do not distort the signal. Things to
consider are damping due to interference between
the magnetic fields of the magnet and coil, and
resonance of the unit itself.
Fig 3.19 in Kearey et al., 2002

Seismometers – these are similar to geophones, but they are installed in permanent
housings (vaults). Both one and three-component seismometers exist, and they can be
tuned to record within certain frequency bands. These are most common for earthquake
monitoring.

3.6.2 Marine detectors

Hydrophones – detect a change in pressure in the water and piezoelectric elements
change it into an electrical signal. Only P-waves are recorded by the hydrophones
(why?). Hydrophones are usually attached to a streamer that is towed behind a boat. The
streamer is an oil-filled plastic tube that is neutrally buoyant. Streamer lengths can
exceed 6 km, and in modern surveys, it is common for a ship to tow multiple streamers in
parallel, to obtain 3D coverage. Pressure sensors, compasses and tail buoys are used to
monitor the position of the streamer, and fins are used to adjust the streamer depth.

Ocean-bottom Seismometer (OBS) – these are similar to land seismometers, but are
housed in an airtight, metal container. The container is lowered overboard and weights
on the OBS cause it to sink to the seafloor. The OBS can be recovered through a remote-
control device which releases the weights, enabling the OBS to float to the surface.
OBS’s can contain three-component seismometers, enabling the recording of the full 3D
motion of the seafloor. However, it is often difficult to position them accurately (they
may be dropped onto soft sediment, rather than hard seafloor), and OBS’s generally have
short lives on the seafloor, due to limited battery power and limited disk drive space for
recording data.