Seismic Wave Velocity To Strike A Number Of Issues For Study

In seismic data processing, the effects of dynamic correction, pre-stack andpost-stack migration, residual static correction are all based on the accuracy ofmedium velocity, including the resolution of seismic data. So it is the key ofinterpretation and processing of seismic data to get the precise velocity correctly.Seismic exploration in more and more complex geologic condition causes moredifficulty in data processing and brings us more challenge. It is more difficult thanbefore for velocity analysis because that the signal-to-noise becomes lower, thereflection in CMP gather can't be found clearly, and there are no remarkable focusesin velocity spectrums.Anisotropic medium and data collection by long spreads break up theassumption of the standard velocity analysis, which based on short spreads and layerisotropic medium. So using the standard hyperbolic velocity analysis to process theseismic data of these data must cause big error. Focusing on these problems, thispaper analyzes the existing theories by models, explores their industrial applicationsand made some progress.In order to offer corrective research data for velocity analysis, this paperpresented a new reflection forward method. Based on fitting complex interface bytriple spline function, we combined global searching with Quasi-Newton method.Firstly, using global searching way get the approximate solution of linearminimization, then realize Quasi-Newton method byBroyden-Fletcher-Goldfarbo-Shanno method.For the low signal-to-noise data, this paper presents a new distinguish principleof velocity analysis—the principal component analysis. In this way, the reflection canbe obtained though comparing the ratio of the first principle component to the others.Using weighting coefficient can improve the sensitivity of principal component whilededuce the influence of noise to principle component. The standard hyperbolic approximation for reflection moveouts in layered mediais accurate only for relatively short spreads, even if the layers are isotropic. Velocityanisotropy may significantly enhance deviations from hyperbolic moveout. Foranisotropic medium, this paper analyzes the character of moveout curves of reflectedwave by vertical transverse isotropy, and analyzes the theory of two kinds ofnonhyperbolic reflection moveout by models. It proves that the two nonhyperbolicmoveout methods have more accurate than standard hyperbolic way. At last, thispaper points out the nonhyperbolic way of anisotropic media to analyze the velocityof anisotropic media and long-spread data.These conclusions in this paper have been verified by program. These methodspresented in this paper have been tested against by field data. The data processingresults prove that these methods have more effects than traditional way.