Выпуск: 4 , Том: 29 , Год издания: 2011

There is discussed a new method for seismic data processing and technology of its use in solving geological and field tasks. This method is based on the Prony decomposition and selection of components of this decomposition in order to obtain images of seismic data corresponding to a narrow band of frequencies. Therefore, the method can be called Prony filtration. Its closest analogue is a bandpass filtration. However, Prony filtration method provides higher resolution of signal image with respect to the time variable. Also, the method allows to obtain a damping parameter corresponding to the frequency of filtration. This makes it possible to investigate changes in the damping parameter as a function of the frequency. Special attention is paid to three points: (i) theoretical foundations of the Prony decomposition to obtain stable estimates of the Prony spectra for signals similar to seismic ones; (ii) use of Prony spectra in the signal filtration processing, i.e., construction of the Prony filtration algorithms; and (iii) description of the Prony filtration technology and examples demonstrating its capabilities in practice. Using simple examples, we introduce basic ideas of the technology, illustrating the main points that are important for the application of this method in the real seismic data processing. The results show good stability of the algorithm for different waveforms and in the presence of other signals and noise. We hope that this approach attracts the attention of specialists to the nonlinear Prony filtration method, as well as to aspects of its implementation in geophysical spectral analysis