The Method Study On The High Resolution Seismic Data Processing Based On The VSP Records

Along with high resolution seismic prospecting needs, more and more people cared about the problems of attenuation and dispersion. The actual subsurface materials are viscoelastic and inhomogeneous. When elastic waves propagate in the medium, some parts of energy are converted into heat because of the friction effects among particles. Due to the attenuation of seismic energy by the earth, the amplitude dissipation is an exponential function of the time and frequency. This effect decreased the seismic dominant frequency, reduced seismic resolution and distorted amplitude spectrum and phase spectrum of wavelets.To study and compensate the earth Q effects is the key of seismic processing, which is important to improve seismic imaging precision, reasonably interpret AVO effect and properly inverse medium parameters. VSP seismic records, due to particular observation mode of ground excitation and well reception, have higher quality and more high frequency components compared to seismic data. So we can combine seismic data with VSP records to improve seismic records quality.Considering the seismic attenuation and the particular predominance of VSP records, in this paper we give two methods: (1) Inverse Q filtering on seismic data with Q extracted from VSP downgoing arrivals; (2) High frequency restoration to seismic data based on VSP downgoing arrivals.Often the earth attenuation effects are measured as Q effect. The quality factor Q is a valuable parameter, which contain many useful information concerning lithology and reservoir, such as press, porosity, permeability and saturation. Studying Q can also forecast stratum lithology, fluid types, fluid saturation, press and permeability, besides accurately Q estimation is the key for inverse Q filtering.Combined with the domestic and foreign research situation of Q analysis algorithms. In this paper I greatly present fundamentals and modeling examples on spectrum ratio algorithm, centroid frequency shift algorithm and analytic algorithm. Q analysis refers to the procedure for estimating Q directly from a reflection seismic trace. Conventional Q analysis method compares two seismic wavelets selected from different depth (or time) levels, but picking"clean"wavelets without interferences from other wavelet and noise from a reflection seismic trace is really a problem. Therefore, using the Gabor transform spectrum, I propose two Q analysis methods based on the attenuation function and compensation function. The robust Q analysis methods use information from the whole seismic trace, rather than comparing individual wavelets. Because wavelet analysis has better time-frequency localization. combined with seismic wave propagating equation and energy attenuation formula in viscoelastic medium, we study seismic wave attenuation characteristic in wavelet scaling domain and compare energy attenuation attributes to routine instantaneous attributes. At last we indicate that attenuation characters in wavelet scaling domain have more sensitivity.For accurate Q estimation, inverse Q filtering to seismic data can minimize dissipation and dispersion. Combined with the domestic and foreign research situation of inverse Q filtering algorithms. In this paper I states routine inverse Q filtering fundamentals, and furthermore derive a fast inverse Q filtering algorithm and a stable inverse Q filtering algorithm. Both of algorithms are based on the theory of wavefield downward continuation, assuming a depth-dependent layered-earth Q model, they are implemented in a layered manner. For each individual constant Q layer, inverse Q filtering is accomplished in two step: (1) the surface-recorded wavefield is extrapolated to the top of the current layer and (2) a constant Q inverse filtering is performed across that layer. For the fast algorithm, in step two we make variable substitution to extrapolated wavefield, resample and interpolate using spline function in the frequency domain. We make spectrum analysis to entire seismic trace and make least square fit to variation of center frequency along with time. The amplitude operator of the inverse Q filtering, which is a 2-D function of traveltime and frequency, is approximated optimally as the product of two 1-D functions depending, respectively, on time and frequency. The function depending on time relate to center frequency which can be determined by function concerning time. This method is fast due to decomposition of amplitude compensation operator, which insures the application of FFT algorithm. For the stable inverse Q filtering algorithm, the surface-recorded wavefield is extrapolated to the top of the current layer by solving an inversion system to the downward continuation within the overburden. We pick time-variable maximum trimming frequency using gabor transform and according to this frequency determine time-variable gain-limited amplitude and corresponding time-variable gain-limited frequency to implement the stabilization of inverse Q filtering.For another method to compensate the earth Q effects, that is high frequency restoration algorithm. In this paper I also expatiate the fundamentals: this method analyzes the frequency decay of direct arrivals at different VSP depth levels in a well, further more picks up the earth filtering operater, and then compensates the surface seismic data for that decay with HFR by designing inverse earth filtering operator. By high-frequency restoration, this method enhances predominant frequency components, stretches frequency-band and compensates frequency decay and amplitude variation via earth effects, furthermore improves seismic resolution.At last, I verify the inverse Q filtering algorithm and high frequency restoration algorithm by processing the field data. I select a surveying line across well from 3-D section for example. Using zero-offset VSP records to pick up Q operator and earth filtered operator, and I make inverse Q filtering and high frequency restoration to surface seismic data. By respectly comparison among shallow layer, intermediate layer and deep layer, we draw conclusion that both inverse Q filtering algorithm and high-frequency restoration algorithm compensate earth Q effects, enhance dominant frequency and broaden bandwidth of seismic records. Moreover the robust methods improve events continuity and seismic records resolution and recover the frequency components that in principle are recoverable. The application of high-resolution algorithms lays the foundation for the later data processing and interpretation.