Wavelet based intercept attribute for seismic exploration

• US 7,616,524 B1
• Filed: 08/27/2007
• Issued: 11/10/2009
• Est. Priority Date: 08/25/2006
• Status: Active Grant

• Current Assignee: Fairfield Industries Incorporated, University of Tulsa
• Original Assignee: Fairfield Industries Incorporated, University of Tulsa
• Inventors: Gersztenkorn, Adam
• Primary Examiner(s): Bolda; Eric
• Assistant Examiner(s): BREIER, KRYSTINE E
• Application Number:
  US11/895,709
• Time in Patent Office: 806 Days
• Field of Search: 367/25, 367/26, 367/27, 367/31, 367/32, 367/40, 367/41, 367/42, 367/38, 702/17
• US Class Current: 367/38
• CPC Class Codes: G01V 1/30 Analysis G01V1/50 takes precedence

Abstract

A wavelet-based method for improving the quality of seismic data utilizing the intercept determined by applying least squares regression to the wavelet transform of a seismic trace. The intercept is calculated for every time point of the wavelet transform for each seismic trace. The intercepts are then plotted versus time or depth. These plots are used in place of seismic traces themselves to create two dimensional and three dimensional seismic section images. In one embodiment, the real and imaginary portions of the selected wavelet transform are weighted to generate a finer representation of the intercept. In another embodiment, a minimum amplitude value is utilized to establish a noise floor, thus stabilizing the regression calculation. In yet another embodiment, a taper of the amplitude is applied to wavelet enhance the resolving power of the wavelet.

Claims

1. A wavelet-based method for analysis of seismic data comprising:

acquiring a plurality of seismic traces;

applying a wavelet transform to said plurality of seismic traces, said wavelet transform having a basis wavelet and producing wavelet coefficients for localized scales and localized time or depth points; and

calculating an intercept by regression for at least one of said localized time or depth points, wherein said intercept is used to form images of the subsurface.

2. The method of claim 1 wherein said basis wavelet is selected from the group consisting of a Harr wavelet, a Morlet wavelet, a Ricker wavelet or a Daubechies wavelet.

3. The method of claim 1 wherein said at least one seismic trace has undergone any stage of seismic data processing including raw field data to final migrated data.

4. The method of claim 2 wherein said basis wavelet is said Morlet wavelet.

5. The method of claim 1 wherein said wavelet transform is selected from the group consisting of a continuous wavelet transform, a dyadic wavelet transform, a discreet wavelet transform or a fast wavelet transform.

6. The method of claim 1 wherein said calculating an intercept by regression for at least one localized time or depth point comprises performing a least squares linear regression upon a log-log plot of said wavelet coefficients at said localized scales.

7. The method of claim 1 further comprising choosing an appropriate range of scales in order to improve the fit of said intercept by regression prior to performing said regression.

8. The method of claim 1 wherein said calculating an intercept for at least one localized time or depth point comprises calculating intercepts for a plurality of localized time points.

9. The method of claim 8 further comprising forming an intercept trace from the intercepts for a plurality of localized time or depth points.

14. The method of claim 1 wherein said wavelet transform comprises a weighted sum of real and imaginary parts of the wavelet transform.

15. The method of claim 14, wherein the real and the imaginary parts of the wavelet transform are weighted differently.

16. The method of claim 1, further comprising the step of establishing a noise floor for the wavelet transform.

21. The method of claim 1 wherein the wavelet used is the Morlet wavelet.

22. The method of claim 1 wherein the wavelet is the Daubechies wavelet.

23. The method of claim 1 wherein a weighted sum of real and imaginary parts of the wavelet transform is generated.

24. The method of claim 1 wherein a noise floor is established for the wavelet transform.

10. A method for analysis of seismic trace data comprising:

performing a continuous wavelet transform on a seismic trace, said transform producing wavelet coefficients across a range of scales for each localized time point;

calculating the intercept across a range of scales for each localized time or depth point of said transform by a regression means, thereby producing an intercept trace; and

applying image processing to the intercept trace, thereby producing an intercept trace section.

11. The method of claim 10 wherein said seismic trace has undergone data processing comprising at least one of deconvolution, NMO, migration and stacking.

12. The method of claim 10 wherein said continuous wavelet transform comprises a basis wavelet and said basis wavelet is a Morlet wavelet.

13. The method of claim 10 wherein said regression means comprises a least-squares linear regression of a log-log plot.

17. The method of claim 10 wherein said wavelet transform comprises a weighted sum of real and imaginary parts of the wavelet transform.

18. The method of claim 17, wherein the real and imaginary parts of the wavelet transform are weighted differently.

19. The method of claim 10, further comprising the step of establishing a noise floor for the wavelet transform.

20. A method of processing seismic data, said method comprising:

providing original seismic signal data;

putting said seismic data through preprocessing;

putting said seismic data through a wavelet transform;

determining the intercept; and

using the characteristics of the intercept function to create an output value.

25. The method of claim 20, wherein the preprocessing is to attenuate noise.

26. A wavelet based method for analysis of seismic data comprising:

acquiring at least one seismic trace;

applying a wavelet transform to said at least one seismic trace, said wavelet transform having a basis wavelet and producing wavelet coefficients for localized scales and localized time or depth points;

calculating intercepts by regression for a plurality of localized time or depth points; and

forming an intercept trace from the intercepts for a plurality of localized time or depth points.