Method for removing non-geologic amplitude variations from seismic data

Abstract

A method for removing non-geologic amplitude variations from seismic data. The method is based on the premise that lateral amplitude variations with spatial wavelengths shorter than the width of the Fresnel zone, a known wave propagation effect, for the target cannot be caused by changes in the reflectivity of the target and must, therefore, be due to non-geologic causes. The method permits these non-geologic amplitude effects to be identified and removed from the seismic data. The method may be implemented either manually or automatically by a computer.

We claim: 
     
       1. A method for removing non-geologic amplitude variations from a set of seismic data traces, said method comprising the steps of: (a) dividing said set of seismic data traces into one or more time windows containing segments of said seismic data traces;   (b) determining an amplitude value for each of said trace segments;   (c) selecting a seismic data trace from said set of seismic data traces and for each time window along said selected seismic data trace (i) determining the width of the Fresnel zone for said time window,   (ii) determining a Fresnel zone average amplitude value for all trace segments within said Fresnel zone, and   (iii) determining a scalar that will scale the amplitude value for the segment of said selected seismic data trace within said time window to match said Fresnel zone average amplitude value;     (d) interpolating said scalars between adjacent time windows to determine a prorated gain value for each data sample on said selected seismic data trace;   (e) multiplying each data sample on said selected seismic data trace by said prorated gain value for said data sample to create a scaled seismic data trace; and   (f) repeating steps (c) through (e) for each of said seismic data traces.   
     
     
       2. The method of claim 1, wherein said seismic data traces are pre-stack divergence-corrected common offset seismic data traces. 
     
     
       3. The method of claim 1, wherein said seismic data traces are post-stack seismic data traces. 
     
     
       4. The method of claim 1, wherein said time windows are between about 200 milliseconds and about 600 milliseconds in length. 
     
     
       5. The method of claim 1, wherein said method is implemented on a digital computer. 
     
     
       6. The method of claim 1, wherein said amplitude value for each of said trace segments is the average absolute amplitude value for said trace segment. 
     
     
       7. The method of claim 1, wherein said amplitude value for each of said trace segments is the root-mean-square amplitude value for said trace segment. 
     
     
       8. The method of claim 1, wherein said Fresnel zone average amplitude value is the median of said amplitude values for all trace segments within said Fresnel zone. 
     
     
       9. The method of claim 1, wherein said Fresnel zone average amplitude value is the mean of said amplitude values for all trace segments within said Fresnel zone. 
     
     
       10. A method for suppressing offset consistent noises in a set of unstacked seismic data traces, said method comprising the steps of: (a) obtaining a preliminary common midpoint (CMP) stack of said seismic data traces;   (b) dividing said CMP stack into one or more time windows;   (c) calculating Fresnel zone average amplitude values for each time window and a every CMP location;   (d) dividing said set of unstacked seismic data traces into CMP gathers;   (e) scaling each of said CMP gathers to match said Fresnel zone average amplitude values for said CMP location;   (f) stacking said scaled CMP gathers to create a new CMP stack; and   (g) using said new CMP stack to repeat steps (b) through (f) until no new improvements are made in the stacked section.   
     
     
       11. The method of claim 10, wherein said time windows are between about 200 milliseconds and about 600 milliseconds in length. 
     
     
       12. The method of claim 10, wherein said method is implemented on a digital computer. 
     
     
       13. A method for removing non-geologic amplitude variations from a set of seismic data traces, said method comprising the steps of: (a) dividing said set of seismic data traces into one or more time windows containing segments of said seismic data traces;   (b) calculating Freznel zone average amplitude values for each time window and at every data trace location; and   (c) scaling each of said seismic data traces to match said Fresnel zone average amplitude values for said trace.   
     
     
       14. The method of claim 13, wherein said seismic data traces are pre-stack, divergence-corrected common offset data traces. 
     
     
       15. The method of claim 13, wherein said seismic data traces are post-stack seismic data traces. 
     
     
       16. The method of claim 13, wherein said time windows are between about 200 milliseconds and about 600 milliseconds in length. 
     
     
       17. The method of claim 13, wherein each Fresnel zone average amplitude value is the median of the average absolute amplitude for each trace segment within said Fresnel zone. 
     
     
       18. The method of claim 13, wherein each Fresnel zone average amplitude value is the main of the average absolute amplitude for each trace segment within said Fresnel zone. 
     
     
       19. The method of claim 13, wherein said method is implemented on a digital computer. 
     
     
       20. A method for removing non-geologic amplitude variations from a set of seismic data traces, said method comprising the steps of: (a) dividing said set of seismic data traces into one or more time windows and determining the width of the Fresnel zone for each of said time windows;   (b) estimating the geologic portion of the seismic amplitude field of said seismic data traces by removing lateral amplitude variations which have spatial wavelengths shorter than the width of the Fresnel zone for the relevant time window; and   (c) scaling said seismic data traces to match said geologic portion of the seismic amplitude field.   
     
     
       21. The method of claim 20, wherein said seismic data traces are pre-stack, divergence-corrected common offset data traces. 
     
     
       22. The method of claim 20, wherein said seismic data traces are post-stack seismic data traces. 
     
     
       23. The method of claim 20, wherein said time windows are between about 200 milliseconds and about 600 milliseconds in length. 
     
     
       24. The method of claim 20, wherein said method is implemented on a digital computer.