The Reconstruction of Missing Marine Seismic Data

The Reconstruction of Missing Marine Seismic Data

Van Leeuwen, L.P.

Hegge, R.F. (mentor)
Wapenaar, C.P.A. (mentor)
Bosch, F. (mentor)

Civil Engineering and Geosciences

Geoscience & Engineering

Section Applied Geophysics and Petrophysics

2012-08-23

To evaluate the subsurface, seismic surveys are carried out. A requirement for the processing of these data is that it is dense and regularly sampled and it should also include the near-offsets. Because these data cannot be acquired in practice for towed marine surveys, it should be obtained by inter- and extrapolation. The first objective of this report was to give an overview of the methods which already exist for the inter- and extrapolation of seismic data. Only the most important methods are described, because the total number of methods is too large. They are first explained and finally classified into the three main categories: Wave-equation based methods, Domain transform methods and Prediction-Error-Filter methods. The other objective of this thesis was to explain and test the Mixed domain reconstruction method. It is based on the Polya-Plancherel theorem, which states that band-limited data in one domain has infinite support in the other domain. This theorem makes it possible to reconstruct the missing seismic data. A conjugate gradient method is used in the optimization part of the implementation. The method was tested for its interpolating qualities and it was shown that it works very well. For extrapolation, the data should be transformed into a ‘split-spread’ configuration instead of the ‘end-on’ configuration which is typical for towed marine surveys. With this transformation, the near-offset gap is filled accurately. It was also shown that the error increases for larger gaps and that the method is limited by the offset and not by the number of traces. The method works in the frequency domain, but a time domain implementation of the algorithm was generated which gave promising results. Finally, the method can handle a wide variety of datasets: low quality data, conflicting dip data, dual-sensor data and real data. The conclusion of this report is that the Mixed domain reconstruction method works very well and can be applied to many types of seismic data. It was demonstrated that a dense and regularly sampled dataset can be obtained which contains also the near-offset.

seismic data reconstruction
extrapolation
interpolation
Mixed domain reconstruction method

http://resolver.tudelft.nl/uuid:73c0ddc1-351c-43ae-b536-59829a0f9d07

Student theses

master thesis

(c) 2012 Van Leeuwen, L.P.