Wavelets Lecture - Clean

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 31
At a glance
Powered by AI
The key takeaways are that a wavelet represents the seismic response at a boundary, and can be described by its amplitude and phase spectra. Parameters like amplitude, frequency, and phase of a wavelet impact the shape and properties of the wave.

A wavelet is a limited-duration time series that approximates the seismic response at a boundary. It can be described by its amplitude and phase spectra, with the amplitude spectrum determined from autocorrelation but phase requiring well ties. A wavelet is like a seismic reflection convolved with the reflection coefficient.

The amplitude determines the strength of the wave, with higher amplitudes resulting in taller waves. Frequency is a measure of the number of wave cycles that pass a point per unit of time. Phase refers to where the wave is in its cycle, such as at its peak or trough.

Wavelets

Chapter
7

You guys arent drilling that hard streak, are


you?

Introduction
The objective of this section is to discuss and
understand wavelets. This will include:
The amplitude and phase spectrum
Zero, minimum, mixed and reverse phase
wavelets
How wavelets can affect the seismic response

Chapter 7: Wavelets

What is a wavelet?
A wavelet is a limited-duration
time series that approximates
the response of seismic energy
at a lithologic boundary. A
wavelet convolved with a
reflection coefficient should
approximate an actual seismic
reflection.
A wavelet can be approximated by a sum of cosine waves.
A wavelet can be described by its amplitude and phase spectra.

Chapter 7: Wavelets

Waves
To start our review of wavelets, lets first consider waves.
The cosine function is defined as the x coordinate of a
particle moving around a circle whose radius is 1.
y
+1

+1

+360

+90
+90

+270
+180 +270 +360

+180

Here the horizontal axis is angle . In future


graphs, the horizontal axis will be time.
Chapter 7: Wavelets

Waves: Frequency

1 sec.

+1 sec.

1 sec.

+1 sec.

The top wave has a frequency of 2 Hz.


The lower wave has a frequency of 4 Hz.
Both waves have amplitudes of 1.
Chapter 7: Wavelets

Amplitude

How would we plot the amplitude spectra


of these cosine waves?

2
1

Frequency
5

Waves: Amplitude

+1 sec.

1 sec.

+1 sec.

How would we plot the amplitude


spectra of these cosine waves?

The top wave has an amplitude of 1.


The lower wave has an amplitude of 2.
Both waves have frequencies of 2 Hz.
Chapter 7: Wavelets

Amplitude

1 sec.

2
1

Frequency
6

Waves: Phase
0

Zero Phase
1 sec.

+360

180

90

+90

90 Phase
1 sec.

+180

+270

+1 sec.

180

90

+360
+90

+180

+1 sec.

+270

A normal cosine wave has its peak amplitude at time 0.


If the peak amplitude is at a different time, we say that the phase is shifted.
A wavelets phase refers to its peak, trough, or position elsewhere on the wave.
A waves phase at time zero is called the phase of the wave.
What is the phase of the upper wave?
The upper wave has zero phase.
What is the phase of the lower wave?
The lower wave has 90 phase.
Chapter 7: Wavelets

Waves: Phase
0

Zero Phase

180

90

+90

90 Phase
1 sec.

+270

+1 sec.

180

90

How would we plot the phase


spectra of these waves?
The upper wave has zero phase.
What is the phase of the lower wave?
The lower wave has 90 phase.
Both waves have frequencies of 2 Hz.
Chapter 7: Wavelets

+180

+360
+90

+180

+1 sec.

+270

+180
+90

Phase

1 sec.

+360

90
180
2
4
Frequency
8

Waves: Amplitude and Phase


What are the amplitude and phase spectra of the following cosine
wave?

1 sec.

+1 sec.

+90

Phase

Amplitude

+180

90
180

Chapter 7: Wavelets

2
4
Frequency

2
4
Frequency

Summing Cosine Waves

0
Chapter 7: Wavelets

Sum

Lets see what happens when we add cosine waves of different frequencies
together. The sum of three cosine functions (upper left) provides a rather oddlooking waveform (lower left). However the sum of cosine functions from 5 to 65
Hz (upper right), incrementing by 5 Hz, looks more like a seismic wavelet!

0
10

Amplitude and Phase Spectra

Amplitude Spectrum
Chapter 7: Wavelets

The wavelet is decomposed into the


amplitude and phase components. Since
the amplitude of each cosine is 1, the
amplitude spectrum consists of a series of
spikes or dots of amplitude 1. The spikes
are located at the frequency corresponding
to that of each cosine function. The phase
of each cosine is 0 and is illustrated in the
phase spectrum by a point at 0 at each
frequency.

Phase Spectrum
11

Summing Cosine Waves

0
Chapter 7: Wavelets

Sum

As in a previous slide, each component cosine in the wavelet on the left has the
same amplitude. On the right, we have varied the amplitudes of each component
cosine. The 30-Hz cosine (in bright green) has the highest amplitude. What has
happened to the wavelet amplitudes away from the central peak?

0
12

Amplitude Spectra

Here we compare the sum of waves of equal amplitude with waves of varying amplitude. Note
the decreased side-lobe energy in the latter. In both cases, the wavelets are zero-phase.
Chapter 7: Wavelets

13

Phase
Similarly, right, the sum of several 90 cosine waves of varying amplitudes results
in a +90 wavelet.

0
Sum

Zero-Phase

+90 Phase

Phase-shifting a cosine wave by 90 is equivalent to a time shift.


Phase-shifting a wavelet by 90, however, gives a very different curve!
Chapter 7: Wavelets

14

Zero-Phase Wavelet
Any wavelet can be expressed in terms of amplitude and phase. Below are the
amplitude and phase spectra of a 30-Hz. Ricker wavelet. Note that the wavelet is
displayed in time, while amplitude and phase are displayed in terms of frequency.

Symmetric about
Zero Time.
This wavelet has
Zero Phase.
Negative Time

Phase

Increasing Frequency
Chapter 7: Wavelets

Positive Time

Amplitude

Increasing Frequency
15

15

The Ricker Wavelet


The Ricker wavelet (or Mexican-hat wavelet) is the
negative of the 2nd derivative of the Gaussian function:
1
0.9
0.8

Gaussianfunction

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0

-5

-4

-3

-2

-1

0.8
0.6

1st derivative

0.4
0.2
0
-5

-4

-3

-2

-1
-0.2
-0.4
-0.6
-0.8

0.6

2ndderivative

0.4

0.2

0
-5

-4

-3

-2

-1

0
-0.2

-0.4

-0.6

-0.8

Norman Hurd Ricker


developed the Ricker
wavelet while working
for Western Electric (a
US telephone company).
He later applied this
work to the seismic
industry, first at Carter
Oil and later at the
University of Oklahoma.

-1

Chapter 7: Wavelets

16

Minimum-Phase Wavelet
Below is a 30 Hz. minimum phase Ricker Wavelet. Note that the phase
has a linear shift, but the amplitude spectrum is the same as for the zerophase wavelet. Note that the wavelet is not symmetric.
No energy before Zero
Time (causal).
Phase varies
with frequency.
Negative Time

Phase

Positive Time

Amplitude

This type of wavelet is generated from an explosive source such as dynamite or air gun.
Chapter 7: Wavelets

17

Minimum phase: Whats in a name?


Minimum phase really means minimum delay
that is, the wavelet rapidly builds to its maximum
energy, then quickly dies away.
The word phase means different things to different
people and in different contexts. Christopher Liner
wrote a rather coherent summary of possible
meanings in the May 2002 Leading Edge, pp. 456457.
http://scitation.aip.org/getabs/servlet/GetabsServlet?
prog=normal&id=LEEDFF000021000005000456000001&idtype=cvips&gifs=yes

Chapter 7: Wavelets

18

Summing Cosines: Minimum Phase


Each phase
is different!

All phases
are +90.

0
Sum

Minimum-Phase

+90 Phase

Here we see the cosines summed to create the minimum-phase wavelet,


compared to a +90 wavelet.
Chapter 7: Wavelets

19

Rotating the Phase


The phase of a wavelet can be rotated between 0 and 360. Note that in each
case the amplitude spectrum remains the same, but the wavelet shape changes
dramatically. Note where the peak energy is located relative to zero time.

+90 Degrees

Symmetric wavelets

180 Degrees
or Reversed Polarity

270 Degrees
or 90 Degrees
0

Antisymmetric wavelets

0 Degrees

For interpretation and synthetic seismograms, which wavelet is the best?


Chapter 7: Wavelets

20

Changing the Amplitude Spectrum


Each of the wavelets below is zero-phase. Only the amplitude spectra differ. As the
amplitude spectrum becomes broader, the wavelet narrows.

10 Hz

30 Hz
0

Narrow amp
spectrum

20 Hz

40 Hz
0

Broad amp
spectrum

For interpretation and synthetic seismograms which wavelet is the best?


Chapter 7: Wavelets

21

A few words about Polarity


Polarity is related to phase. SEG-Y format includes a header
location in which to specify the polarity, but you, the
interpreter, wont normally see this. Even if someone tells
you that your data are zero phase, should you believe it?
In this school, we follow the polarity convention that a
positive impedance contrast (a soft shale over a hard
limestone, for instance) is shown in a zero-phase display as a
peak (black). This is sometimes called the American
convention.
Many companies follow the opposite convention.
The company that operates a given block usually decides the
phase and polarity.
Even if the phase is claimed to be zero, you should verify!
Sometimes, even within the same office, different polarity
conventions will be in use, depending on the project.
Chapter 7: Wavelets

22

Polarity: A Sad Case History


Consider the following example:
Earth
Model
Interpretation

Assumed Polarity Convention


Observed
Response

Acoustic
Impedance
+

Hard
Hard
Soft
High Porosity Gas Sand

A pretty basic concept, but...


Observed
Response

Hard
Hard
Earth
Model
Interpretation

this happened with a deep marine


exploration well circa 2000.
Actual Polarity Convention
Acoustic
Impedance

Hard
Soft
Hard
Low Porosity Brine Sand
Hard
Soft

Chapter 7: Wavelets

23

Using the Sea Floor to Determine Polarity


The polarity convention can be checked by comparing the sea-floor reflection to the reservoir
section. In general, the sea floor is an increase in impedance and in this deep-water-Nigeria
example, it is displayed as a black peak. In this region, sand is generally low in acoustic impedance
compared to encasing shale. The top of the sand should be a trough, and the base a peak. The
zone of interest appears to contain several low-impedance (possibly sand) events.

Ocean Bottom
Zone of interest

Chapter 7: Wavelets

24

Apparent Phase and Tuning Effects.


What is the Phase?
Tuning effects can cause ambiguity in the seismic response. This wedge model shows the seismic
response as a function of thickness. The input wavelet is zero-phase. At 40-130 ft in thickness, the
model response appears to be minimum-phase or -90 phase. So be careful how and where
phase is determined!

Stack Response Wedge Model

Input Model
Hard
Soft

0 ft
500 ft

Hard

Seismic
Response

Input
Wavelet
Chapter 7: Wavelets

300 ft thick:
Zero-Phase

70 ft thick: appears to be
Minimum or -90 Phase
25

Tuning Thickness
Here is the same wedge model as we saw in the previous slide. At a thickness of
approximately 75 feet, the trough and peak amplitudes reach their maxima. At this
thickness, we say that the soft layer is tuned.

Stack Response Wedge Model


0 ft

Input Wavelet
To first order, tuning in zero-phase data occurs when
the peak of the positive reflection aligns with the sidelobe peak of the negative reflection.
Tuning Thickness =

Vint
4(Dominant Frequency)

Here, Vint = 6000 ft/sec and the dominant frequency is


20 Hz.
Chapter 7: Wavelets

Top
Reflection

Base
Reflection

Tuned
Response
26

Wavelet Summary
A wavelet is the response of seismic energy reflecting
from a rock interface.
We can determine the amplitude spectrum but not
any phase information of a wavelet by autocorrelating seismic data.
To estimate phase information, we need to tie at least
one well to seismic data. SynTool, Jason, Hampson
Russell, and Wavex include tools to do this.
In geophysical reservoir characterization, wavelets are
important for:
Forward modeling
Inversion

Chapter 7: Wavelets

27

Review
Draw a 40 Hz Ricker wavelet in each graph.

Zero Phase

+90 Phase

-180 Phase

-90 Phase

Chapter 7: Wavelets

28

Why should I care about wavelet phase?

Vp = 3 km/s Vs = 1.5 km/s = 2.2 g/cc


AI = 6.60 km/s*g/cc
Vp = 2.8 km/s Vs = 1.4 km/s = 2.1 g/cc
AI = 5.88 km/s*g/cc
Vp = 3 km/s Vs = 1.5 Km/s = 2.2 gm/cc
AI = 6.6 km/s*g/cc

What are the Acoustic Impedances?


What are the Reflection Coefficients?
Convolve these RCs with a
zero-phase wavelet.
Chapter 7: Wavelets

29

Why should I care about wavelet phase?

Vp = 3 km/s Vs = 1.5 km/s = 2.2 g/cc


AI = 6.60 km/s*g/cc
Vp = 2.8 km/s Vs = 1.4 km/s = 2.1 g/cc
AI = 5.88 km/s*g/cc
Vp = 3 km/s Vs = 1.5 Km/s = 2.2 gm/cc
AI = 6.6 km/s*g/cc

Convolve these RCs with a


90 wavelet.
Chapter 7: Wavelets

30

Does 90 reflectivity look like a log


curve?

Vp = 3 km/s Vs = 1.5 km/s = 2.2 g/cc


AI = 6.60 km/s*g/cc
Vp = 2.8 km/s Vs = 1.4 km/s = 2.1 g/cc
AI = 5.88 km/s*g/cc
Vp = 3 km/s Vs = 1.5 Km/s = 2.2 gm/cc
AI = 6.6 km/s*g/cc

GR

Convolve these RCs with a


90 wavelet.
Chapter 7: Wavelets

31

You might also like