Marine Geosciences in Bretagne
Brest, May 24 - June 7, 2005

International Training Course in Marine Geosciences, UBO - Purdue 2005
Stage international de Formation en Géosciences marines, UBO – Purdue 2005

UBO
-Purdue
Earth Science
Exchange
Programme
Multichannel Seismic Processing
By: Tabrez Syed Ali, PhD, 1st Year, Vincent Riboulot, Master 1 SML-GO
Introduction
Part I: Seismic wave propagation and data acquisition 3I-1. Seismic wave propagation
I-2. Data acquisition

Part II: Seismic processing
II-1. CMP gather
II-2. Normal Move Out correction (NMO)
II-3. Migration
II-4 CMP stack (summed trace)
II-5. Deconvolution
II-6. Statics correction
II-7. Bandpass filter
II-8. Gain -
Conclusion




INTRODUCTION

The objective is to improve the signal to noise ratio. The data treatment consists of many processing steps on the seismic trace so as to exclude unwanted signals (noise, multiples, etc...).


Figure 1:
Sparker section from the Strataforme cruise (Stsp045). Section A. is the raw seismic data. Section B shows the same thing after processing (G. Jouet, 2003).

Figure 1 shows the necessity of the seismic processing. It compares two seismic sections, the first is raw data obtained by seismic acquisition and the second is the same after processing.
In a first part we introduce the subject, discuss data acquisition and the seismic wave propagation and then we discuss seismic processing.


PART I: SEISMIC WAVE PROPAGATION AND DATA ACQUISITION

I-1. Seismic wave propagation
The wave paths followed can be categorized as.
The phenomenon is more complicated due to multiple reflections:
- The peg-leg is a resonance phenomenon resulting due to a geological source.
- First multiple: Energy is reflected at the geological interface.
- Ghost: The energy given off by the source propagates towards the ocean surface, and then gets reflected. Thus, there is a late wave which creates a signal deformation.
- Diffraction: The wave is scattered in all directions when it meets a discontinuity.


PART II: SEISMIC PROCESSING
Seismic processing can be done with many software such as SITHERE, SISBISE, SPW etc… But, in this part, we discuss the different steps involved in the processing.The processing is divided into two parts:§ Data reduction
- Raw seismic trace gather (CMP gather)
- Normal Move Out correction (Might also use Dip move out to further enhance)
- CMP stack or summed trace (reflection for a single reflector is enhanced)
- Steps repeated for all points to get§ Data Enhancement
- Statics correction
- Deconvolution (removing source signature)
- Bandpass filter (Getting rid of noise outside frequency band)
- Mute, Gain recovery

Migration
II-1. CMP gather

Reflection data transmitted from the same point (reflector) is collected by each of the hydrophones.


Figure 4: Pattern shows the trace gather for the first CDP. It’s the same for the other CDP.


 

II-3. Migration
Reflection data is corrected for travel time and position relative to shot points which can arise due to geologic structures such as synclines and is seen as a bow tie on the stacked time series data. Migration is an inverse wave scattering calculation that relocates seismic reflections and diffractions to the location of their origin. Various methods of migration are DMO and frequency domain, ray trace and wave equation migration.


Figure 6:
Migration representation. A: Real sea floor topology with wave paths. B: Sea floor representing on a seismic section.

 


Figure 7: Seismic section example before (a) and after (b) migration.






II-5. Deconvolution
Seismic trace is basically a convolution of source function and earths response. So we can use deconvolution to remove the source signature which is known. But this isn’t as easy as it sounds and if successful, each seismic trace is transformed into a time series of impulses/spikes with arrival times that represent primary reflection times to the reflectors, and amplitudes that represent the reflection coefficients of the reflectors.


Figure 9: First: reflection signal, then: source signal, and convolved signal


II-7. Bandpass filter
A frequency filter allows the exclusion of noise corresponding to lower and higher frequencies i.e. data which lies outside the reflected frequency domain.


Figure 11:
Frequency filtration pattern.




II-8. Gain application
Attenuation due to the head-on wave propagation results in a wave amplitude loss which is inversely proportional to ‘r’ (wave distance covered)
Wave attenuation = higher frequency absorption. = ratio reflection/transmitting of the wave.

The noise amplitude increases with the signal as we go deep.


Figure 12:
It shows a trace without the gain (a) and the same trace after the correction.

CONCLUSION
Correct processing of raw seismic data is very important as it allows the correct interpretation of the seismic section under study.

 

 


Figure 2: Different models of wave propagation

Pass 1: Direct wave.
Pass 2: Reflected wave.
Pass 3: Head wave.The reflection point is called mirror point or CDP (common depth point)



I-2. Data acquisition

A strong impulsive source, S (air-gun, sparker etc.), is generally used (Figure 3). Receivers R (streamer composed of many hydrophones) aim to record reflection signals from geologic interfaces which result fromacoustic impedance contrasts. Reflection signal (time series) data are collected for each shot recorded (distance between shot points is kept constant)

Figure 3: Pattern shows successive shots.


II-2. Normal Move Out correction (NMO)

It’s the process to cancel the travel time delay caused by the separation between a source and receiver in case of a flat reflector.The total distance travelled by wave increases with successive shot.


The change in travel time with distance away from shot is called move out (Hyperbolic function).


Figure 5
: Above, the record of the three shots and then the corrected signal.



II-4. CMP stacks (summed trace)
Reflection data transmitted from the same point is weighted thereby increasing the amplitude of signal and reduces noise which is generally random and cancels out.An example is presented in figure X. After the operation, noise which is out of phase cancels out. Whereas reflective signals, which are in phase, add up.


Figure 8
: Pattern of stack.


 

II-6. Statics correction
Static correction: To correct for variations in topography as the source is not held at the same level.



Figure 10: Seismic section to show difference before and after static correction.