Signals and noises are decomposed by hybrid filtering of norm L 1 and L 2 from seismic data obtained along the hyperbolic curve determined by t0 times and stack velocity.
3. A new seismic phase identification method which can reflect the characteristics of amplitude increase and frequency changing of earthquake signal is developed.
It recounts the fundamental, system structure and function of mixed inversion based on genetic algorithm and introduces the main characteristics of the system. The correctness of f-x noise-eliminating, target function's wavelet pick-up and the mixed inversion disposal is detected by the numerical value model and real earthquake signal.
The analysis of deep-layer reflection seismic signal energy shows that low-frequency seismic signals are capable of both penetrating the thin high-velocity basalt layer and reducing the diffraction noise caused by the rough surfaces.
The wavelet length can be determined according to the wavelet transform results and be eliminated thereafter so that we are able to detect thin bed seismic signal with resolution of 1/32 wavelength.
A new method for calculating the ellipticity parameters of a wave in the form of a time-frequency spectrum is proposed, which offers wide possibilities for filtering seismic signals in order to suppress or extract the Rayleigh components.
Published seismic data were considered in combination with the results of lithogeochemical and electrochemical data on soils above oil reservoirs and subvertical dynamic anomalies at the territory of Tatarstan.
We present numerical implementations of the B-spline algorithm for an earthquake signal and compare the numerical performance of this approach with that given by the standard empirical mode decomposition.