Mitigation of Noise in Marine Multicomponent Seismic Data through the Relationship between Wavefield Components at the Free Surface

• US 20080259726A1
• Filed: 04/17/2007
• Published: 10/23/2008
• Est. Priority Date: 04/17/2007
• Status: Active Grant

• Current Assignee: WesternGeco LLC (Schlumberger NV)
• Original Assignee: WesternGeco LLC (Schlumberger NV)
• Inventors: van Manen, Dirk-Jan, Robertsson, Johan Olof Anders
• Granted Patent: US 8,077,543 B2
• Time in Patent Office: Days
• US Class Current: 367/24
• CPC Class Codes: G01V 1/36 Effecting static or dynamic corrections on records, e.g. correcting spread; Correlating seismic signals; Eliminating effects of unwanted energy
• CPC Class Codes: G01V 1/38 specially adapted for water-covered areas G01V1/28 takes precedence
• CPC Class Codes: G01V 2210/32 Noise reduction
• CPC Class Codes: G01V 2210/56 De-ghosting; Reverberation compensation

Abstract

A method and apparatus implementing a technique for mitigating noise in seismic data are presented. The technique includes estimating a noise component in a set of multicomponent seismic data representing a recorded wavefield from the relationship at the free surface of an upgoing component and a downgoing component decomposed from the recorded wavefield; and removing the estimated noise component from the seismic data. The noise component can be estimated by decomposing a recorded wavefield into an upgoing component and a downgoing component and isolating a noise component in the recorded wavefield from the relationship of the decomposed upgoing and downgoing components at the free surface.

Claims

1. A software-implemented method for mitigating noise in multicomponent, marine seismic data, comprising:

estimating a noise component in a set of multicomponent seismic data representing a recorded wavefield from the relationship at the free surface of an upgoing component and a downgoing component decomposed from the recorded wavefield; and

removing the estimated noise component from the seismic data.

2. The software-implemented method of claim 1, wherein estimating the noise component includes:

decomposing a recorded wavefield into an upgoing component and a downgoing component; and

isolating a noise component in the recorded wavefield from the relationship of the decomposed upgoing and downgoing components at the free surface.

3. The software-implemented method of claim 2, wherein isolating the noise component includes:

processing at least one of the decomposed upgoing and downgoing components such that the upgoing and downgoing components are at a common depth and propagating in the same direction;

taking the difference between the processed upgoing and downgoing components to find the noise component; and

solving the found noise component for the noise.

4. The software-implemented method of claim 3, wherein the processing includes:

forward propagating the upgoing component from its collection depth to the free surface;

applying a reflection coefficient to the forward propagated upgoing component at the free surface to obtain a reflected equivalent; and

forward propagating the reflected equivalent to the common depth.

5. The software-implemented method of claim 4, further comprising propagating the downgoing component to the common depth.

6. The software-implemented method of claim 4, wherein the reflection coefficient is represented by e^iω

t^refl, wherein

ω

≡

the frequency of the acoustic signal; and

t_refl≡

the vertical two-way traveltime of the wavefield through the water layer.

7. The software-implemented method of claim 3, wherein the processing includes:

backward propagating the downgoing component from its collection depth to the free surface;

applying an inverse reflection coefficient to the backward propagated downgoing component at the free surface to obtain a reflected equivalent; and

backward propagating the reflected equivalent to the common depth.

8. The software-implemented method of claim 7, further comprising propagating the upgoing component to the common depth.

9. The software-implemented method of claim 7, wherein the reflection coefficient is represented by (e^iω

t^refl)<sup>−

1</sup>, wherein

ω

≡

the frequency of the acoustic signal; and

t_refl≡

the vertical two-way traveltime of the wavefield through the water layer.

10. The software-implemented method of claim 3, wherein the found noise component is represented by:

$N=\frac{2<br><br>\left({\hat{P}}^{\mathrm{down}}-{\tilde{P}}^{\mathrm{down}}\right)}{1-{\mathrm{<br><br>}}^{\mathrm{<br><br>\omega <br><br>}<br><br><br><br>t_{\mathrm{refl}}}}$

wherein;

{circumflex over (P)}^down≡

a second estimate of the downgoing component of the recorded wavefield;

{tilde over (P)}_down≡

the estimated downgoing component of the composite wavefield;

e<sup>iω

trefl</sup>≡

the reflection coefficient;

ω

≡

the frequency of the acoustic signal; and

t_refl≡

the vertical two-way traveltime of the wavefield through the water layer.

11. The software-implemented method of claim 2, wherein isolating the noise component includes:

redatuming one of the upgoing and downgoing components to the free surface;

reflecting the redatumed component at the free surface to obtain a reflected equivalent;

redatuming the reflected equivalent to the common depth.

12. The software-implemented method of claim 11, further comprising redatuming the other component to the common depth.

13. The software-implemented method of claim 2, wherein isolating the noise component includes:

redatuming the decomposed upgoing and downgoing components to the free surface;

adding the redatumed upgoing and downgoing components to find the noise component; and

solving for the found noise component.

14. The software-implemented method of claim 13, wherein the found noise component is represented by:

$N=\frac{2<br><br>\left({\hat{P}}^{\mathrm{down}}-{\tilde{P}}^{\mathrm{down}}\right)}{1-{\mathrm{<br><br>}}^{\mathrm{<br><br>\omega <br><br>}<br><br><br><br>t_{\mathrm{refl}}}}$

wherein;

{circumflex over (P)}_down≡

a second estimate of the downgoing component of the recorded wavefield;

{tilde over (P)}_down≡

the estimated downgoing component of the composite wavefield;

e<sup>iω

refl</sup>≡

the reflection coefficient;

ω

≡

the frequency of the acoustic signal; and

t_refl≡

the vertical two-way traveltime of the wavefield through the water layer.

15. The software-implemented method of claim 1, wherein the decomposed upgoing and downgoing components are represented by:

${\tilde{P}}^{\mathrm{down}}=\frac{1}{2}<br><br>\left(P-Z\right)=P^{\mathrm{down}}-\frac{1}{2}<br><br>N,\mathrm{and}$ ${\tilde{P}}^{\mathrm{up}}=\frac{1}{2}<br><br>\left(P+Z\right)=P^{\mathrm{up}}+\frac{1}{2}<br><br>N$

wherein;

{tilde over (P)}^down≡

the estimated downgoing component of the composite wavefield;

{tilde over (P)}^up≡

the estimated upgoing component of the composite wavefield;

P≡

the sensed composite pressure wavefield;

P^up≡

the sensed upgoing component of the composite wavefield;

P^down≡

the sensed downgoing component of the composite wavefield;

Z≡

the sensed vertical velocity; and

N≡

the noise component of the sensed vertical velocity.

16. A method for mitigating noise in seismic data, comprising:

acquiring a set of multicomponent seismic data in a marine survey; and

removing a noise component in the seismic data estimated from the relationship of the decomposed upgoing and downgoing components at the free surface.

17. The method of claim 16, wherein acquiring the multicomponent seismic data in the marine survey includes acquiring the multicomponent seismic data with an over and under array.

18. The method of claim 16, further comprising estimating the noise component.

19. The method of claim 18, wherein estimating the noise component includes:

decomposing a recorded wavefield into an upgoing component and a downgoing component; and

isolating a noise component in the recorded wavefield from the relationship of the decomposed upgoing and downgoing components at the free surface.

20. The method of claim 19, wherein isolating the noise component includes:

processing at least one of the decomposed upgoing and downgoing components such that the upgoing and downgoing components are at the same depth and propagating in the same direction;

taking the difference between the processed upgoing and downgoing components to find the noise component; and

solving the found noise component for the noise.

21. The method of claim 19, wherein isolating the noise component includes:

redatuming one of the upgoing and downgoing components to the free surface;

reflecting the redatumed component at the free surface to obtain a reflected equivalent;

redatuming the reflected equivalent to a predetermined depth.

22. The method of claim 19, wherein isolating the noise component includes:

redatuming the decomposed upgoing and downgoing components to the free surface;

adding the processed upgoing and downgoing components to find the noise component; and

solving for the found noise component.

23. A apparatus, comprising:

a marine survey system capable of acquiring multicomponent seismic data; and

a computing apparatus capable of removing an estimated a noise component in the seismic data determined from the relationship at the free surface of an upgoing component and a downgoing component decomposed from the recorded wavefield.

24. The apparatus of claim 23, wherein the marine survey system includes a towed-array marine survey system.

25. The apparatus of claim 24, wherein the towed-array marine survey system comprises an over/under acquisition system.

26. The apparatus of claim 23, wherein the marine survey system includes a survey vessel.

27. The apparatus of claim 23, wherein the marine survey system includes a seabed survey system.

28. The apparatus of claim 23, wherein the marine survey system includes an over/under acquisition system.

29. The apparatus of claim 23, wherein the computing apparatus is a centralized computing apparatus.

30. The apparatus of claim 23, wherein the computing apparatus is a distributed computing system.

31. The apparatus of claim 23, wherein the computing apparatus is capable of decomposing the recorded wavefield.

32. The apparatus of claim 23, wherein the computing apparatus is capable of estimating the noise component.