Fourier transform and inverse Fourier transform build up mapping for signal between time domain and frequency domain. Although Fourier transform can analyze signals in terms of time-domain or frequency-domain respectively, can not integrate with both directly. For non-stationary seismic signal, it is very important to understand frequency characteristic of different time around, so time-variant signals can be described using time-frequency joint denotation, then 1D time-domain signals can be mapped into 2D time-frequency profile, time-frequency joint characteristic of seismic signals can be expressed entirely. Because of the limitation of traditional time-frequency analysis methods, a newer time-frequency analysis method, S-transform is introduced into this paper.Traditional time-frequency analysis methods are summarized firstly in the paper, the definition, derivation and characteristic of S-transform are expatiated, 2D S-transform and generalized S-transform are introduced. S-tranform can be deduced from short-window Fourier transform or continuum wavelet transform and combine two kinds of time-frquency analysis methods advantage overcoming their disadvantage: S-transform has a direct relation with Fourier transform, the characteristics such as lossless invertibility, linearity, multi-resolution of wavelet. S-transform is a kind of linear time-frequency represent as short-time fourier transform and wavelet transform, so it has not cross-term ; S-transform has muti-resolution overcoming single resolution of short-time fourier transform. Otherwise, S-transform contains phase factor which wavelet transform has not. In order to understand the character of S-transform, theory models are designed, S-transform, short-time fouier transform and Wigner-Ville distributing are compared and analysised.Based on above theory, S-transform is firstly applied to some processing and interpretation of seismic prospecting.On the seismic processing, first-break picking is its first application due to high time-frequency resolution. different S/N ratios are added in theory seismic data, first-breaks are detected using S-transform, then which is applied to first-break picking of tomographic static in hilly data with lower S/N ratios; second, seismic surface wave has low frequency and regional property in time-frequency surface and S-transform has lossless invertibility. surface wave suppression technology based on S-transform is applied to theory data and practical data.On the seismic interpretation, the geological cyclothem bodies identification is its first...
|