Signal-to-Noise Enhancement Methods and Multiple Elimination



The occurrence of noise is a pervading problem in seismic data. As there are many types of noise (random, organized in various ways) there are also many ways to reduce its impact. Moreover, while removing the noise, it is essential that the desired signal should be preserved.

A representative overview will be given of the various types of noise together with methods to eliminate each type. A special type of organized of noise are multiples. It may even be argued that multiples are useful energy that should be used, after its separation from the primaries, for e.g. imaging. Depending on the environment, a specific type of multiple might be dominant, e.g. in a water layer with arbitrary waterbottom geometry or free-surface generated multiples. Such multiples require specific techniques for prediction and subsequent subtraction.

An overview of all multiple types will be given together with methods to eliminate these. 





Geophysicists who deal with seismic data processing and who especially want to familiarize themselves with the occurrence of noise and methodologies of handling of the large variety of noise phenomena.






Participants will get a full understanding of the various types of noise. They will be able for each dataset to assess the type of noise, apply the appropriate method with the correct choice of parameters and use the proper diagnostics before and after the noise elimination.



Continuing Professional Development






Signal-to-Noise Enhancement Methods


1.  Straight stack

2.  Weighted stack

3.  Diversity stack

4.  Velocity stack, Parabolic Radon Transform and Non-uniform Fourier Transform

5.  Median based methods

6.  Wiener filters

7.  Matched filter and Output energy filter

8.  Karhunen Loeve transform

9.  Despiking

10. f-x prediction filtering

11. Fan filtering

12. Ground roll filtering

13. Arrays

14. Acquisition perturbations

15. Power line noise filtering

16. Tau-p filtering

17. Trace interpolation


Multiple elimination


1. Predictive deconvolution:

   - spiking deconvolution

   - gapped deconvolution

2. Differential moveout:

        - stack

    - weighted stack

    - (k,f)-filtering

        - velocity stack

    - parabolic tau,p transform

3. Convolution methods (the feedback loop) (free surface multiples)

4. Methods based on reciprocity  (the wave equation) (free surface multiples)

5. Dereverberation with the wave equation and with redatuming

6. Image processing




2 days




London, United Kingdom



16-17 October 2017


Houston, Texas, United States



6-7 November 2017